mirror/hdf5
2
0
Fork 0
mirror of https://github.com/HDFGroup/hdf5.git synced 2024-12-09 07:32:32 +08:00
Code Issues Packages Projects Releases Wiki Activity
608a71200d
hdf5/fortran/test/tH5A_1_8.F90

2778 lines
92 KiB
Fortran
Raw Normal View History

[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
!****h* root/fortran/test/tH5A_1_8.f90
!
! NAME
! tH5A_1_8.f90
!
! FUNCTION
! Basic testing of Fortran H5A APIs introduced in 1.8.
!
! COPYRIGHT
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
! * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! Copyright by The HDF Group. *
! Copyright by the Board of Trustees of the University of Illinois. *
! All rights reserved. *
! *
! This file is part of HDF5. The full HDF5 copyright notice, including *
! terms governing use, modification, and redistribution, is contained in *
Merge pull request #426 in HDFFV/hdf5 from ~LRKNOX/hdf5_lrk:hdf5_1_10 to hdf5_1_10 * commit '54957d37f5aa73912763dbb6e308555e863c43f4': Commit copyright header change for src/H5PLpkg.c which was added after running script to make changes. Add new files in release_docs to MANIFEST. Cimmit changes to Makefile.in(s) and H5PL.c that resulted from running autogen.sh. Merge pull request #407 in HDFFV/hdf5 from ~LRKNOX/hdf5_lrk:hdf5_1_10_1 to hdf5_1_10_1 Change copyright headers to replace url referring to file to be removed and replace it with new url for COPYING file.
2017-04-18 03:32:16 +08:00
! the COPYING file, which can be found at the root of the source code *
! distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases. *
! If you do not have access to either file, you may request a copy from *
! help@hdfgroup.org. *
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
! * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
! CONTAINS SUBROUTINES
! attribute_test_1_8, test_attr_corder_create_compact, test_attr_null_space,
! test_attr_create_by_name, test_attr_info_by_idx, attr_info_by_idx_check,
! test_attr_shared_rename, test_attr_delete_by_idx, test_attr_shared_delete,
! test_attr_dense_open, test_attr_dense_verify, test_attr_corder_create_basic,
! test_attr_basic_write, test_attr_many, attr_open_check,
!
!*****
[svn-r24967] Maintenance: Reorganized and cleaned the code to remove compiler warnings in the Fortran test code and examples. Platforms tested: Manual testing in place and using srcdir on jam, platypus, and emu with default and PGI, Intel and new GNU compilers. ifort compiler was also tested with -i8 and -r8 flags on jam. CMake tested on jam.
2014-04-06 23:56:21 +08:00
MODULE TH5A_1_8
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
USE HDF5 ! This module contains all necessary modules
USE TH5_MISC
USE TH5_MISC_GEN
[svn-r24967] Maintenance: Reorganized and cleaned the code to remove compiler warnings in the Fortran test code and examples. Platforms tested: Manual testing in place and using srcdir on jam, platypus, and emu with default and PGI, Intel and new GNU compilers. ifort compiler was also tested with -i8 and -r8 flags on jam. CMake tested on jam.
2014-04-06 23:56:21 +08:00
CONTAINS
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
SUBROUTINE attribute_test_1_8(cleanup, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! This subroutine tests following 1.8 functionalities:
! h5acreate_f, h5awrite_f, h5aclose_f,h5aread_f, h5aopen_name_f,
! h5aget_name_f,h5aget_space_f, h5aget_type_f, H5Pset_shared_mesg_nindexes_f,
! H5Pset_shared_mesg_index_f
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
!
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
LOGICAL, INTENT(IN) :: cleanup
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER, INTENT(INOUT) :: total_error
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
!general purpose integer
!
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: i, j
INTEGER :: error ! Error flag
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! NEW STARTS HERE
INTEGER(HID_T) :: fapl = -1, fapl2 = -1
INTEGER(HID_T) :: fcpl = -1, fcpl2 = -1
INTEGER(HID_T) :: my_fapl, my_fcpl
LOGICAL, DIMENSION(1:2) :: new_format = (/.TRUE.,.FALSE./)
LOGICAL, DIMENSION(1:2) :: use_shared = (/.TRUE.,.FALSE./)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
INTEGER :: ret_total_error
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! ********************
! test_attr equivelent
! ********************
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
! WRITE(*,*) "TESTING ATTRIBUTES"
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_FILE_ACCESS_F,fapl,error)
CALL check("h5Pcreate_f",error,total_error)
CALL h5pcopy_f(fapl, fapl2, error)
CALL check("h5pcopy_f",error,total_error)
CALL H5Pcreate_f(H5P_FILE_CREATE_F,fcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pcopy_f(fcpl, fcpl2, error)
CALL check("h5pcopy_f",error,total_error)
CALL H5Pset_shared_mesg_nindexes_f(fcpl2,1,error)
CALL check("H5Pset_shared_mesg_nindexes_f",error, total_error)
CALL H5Pset_shared_mesg_index_f(fcpl2, 0, H5O_SHMESG_ATTR_FLAG_F, 1, error)
CALL check(" H5Pset_shared_mesg_index_f",error, total_error)
DO i = 1, 2
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF (new_format(i)) THEN
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
WRITE(*,'(1X,A)') "Testing with new file format:"
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
my_fapl = fapl2
ELSE
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
WRITE(*,'(1X,A)') "Testing with old file format:"
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
my_fapl = fapl
END IF
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_basic_write(my_fapl, ret_total_error)
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
CALL write_test_status(ret_total_error, &
' - Tests INT attributes on both datasets and groups', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format(i)) THEN
DO j = 1, 2
IF (use_shared(j)) THEN
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
WRITE(*,*) " - Testing with shared attributes:"
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
my_fcpl = fcpl2
ELSE
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
WRITE(*,*) " - Testing without shared attributes:"
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
my_fcpl = fcpl
END IF
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_dense_open(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error, &
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing INT attributes on both datasets and groups', &
total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_null_space(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error, &
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing storing attribute with "null" dataspace', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_many(new_format(i), my_fcpl, my_fapl, ret_total_error)
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
CALL write_test_status(ret_total_error, &
' - Testing storing lots of attributes', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_corder_create_basic(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error, &
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing creating objects with attribute creation order', &
total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_corder_create_compact(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error, &
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing compact storage on objects with attribute creation order', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
[svn-r20033] Porblem: A typo caused compilation failures with gfortran 4.6 on fred. Fixed. Platforms tested: fred and jam (minor change)
2011-02-03 09:24:52 +08:00
CALL test_attr_info_by_idx(new_format(i), my_fcpl, my_fapl, ret_total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
CALL write_test_status(ret_total_error, &
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing querying attribute info by index', &
total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
[svn-r20033] Porblem: A typo caused compilation failures with gfortran 4.6 on fred. Fixed. Platforms tested: fred and jam (minor change)
2011-02-03 09:24:52 +08:00
CALL test_attr_delete_by_idx(new_format(i), my_fcpl, my_fapl, ret_total_error)
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
CALL write_test_status(ret_total_error, &
' - Testing deleting attribute by index', &
total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_create_by_name(new_format(i), my_fcpl, my_fapl, ret_total_error)
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
CALL write_test_status(ret_total_error, &
' - Testing creating attributes by name', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! More complex tests with both "new format" and "shared" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF( use_shared(j) ) THEN
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_shared_rename(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error,&
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing renaming shared attributes in "compact" & "dense" storage', &
total_error)
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ret_total_error = 0
CALL test_attr_shared_delete(my_fcpl, my_fapl, ret_total_error)
CALL write_test_status(ret_total_error,&
[svn-r15065] Description: Broke-up longer lines over multiple lines Removed unused variables.
2008-05-22 04:08:24 +08:00
' - Testing deleting shared attributes in "compact" & "dense" storage', &
total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
END IF
END DO
END IF
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
ENDDO
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pclose_f(fcpl, error)
CALL CHECK("H5Pclose", error,total_error)
CALL H5Pclose_f(fcpl2, error)
CALL CHECK("H5Pclose", error,total_error)
IF(cleanup) CALL h5_cleanup_f("tattr", H5P_DEFAULT_F, error)
CALL check("h5_cleanup_f", error, total_error)
RETURN
END SUBROUTINE attribute_test_1_8
SUBROUTINE test_attr_corder_create_compact(fcpl,fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
!**
!** test_attr_corder_create_compact(): Test basic H5A (attribute) code.
!** Tests compact attribute storage on objects with attribute creation order info
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! Needed for get_info_by_name
IMPLICIT NONE
[svn-r24967] Maintenance: Reorganized and cleaned the code to remove compiler warnings in the Fortran test code and examples. Platforms tested: Manual testing in place and using srcdir on jam, platypus, and emu with default and PGI, Intel and new GNU compilers. ifort compiler was also tested with -i8 and -r8 flags on jam. CMake tested on jam.
2014-04-06 23:56:21 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! - - - arg types - - -
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid
INTEGER :: error
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
CHARACTER(LEN=8) :: DSET3_NAME = "Dataset3"
INTEGER, PARAMETER :: NUM_DSETS = 3
INTEGER :: curr_dset
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER(HID_T) :: dset1, dset2, dset3
INTEGER(HID_T) :: my_dataset
INTEGER :: u
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=7) :: attrname
CHARACTER(LEN=2) :: chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER(HID_T) :: attr !String Attribute identifier
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
data_dims = 0
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
! WRITE(*,*) " - Testing Compact Storage of Attributes with Creation Order Info"
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataset creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
CALL H5Pset_attr_creation_order_f(dcpl, IOR(H5P_CRT_ORDER_TRACKED_F, H5P_CRT_ORDER_INDEXED_F), error)
CALL check("H5Pset_attr_creation_order",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the attribute creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dset1, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dset2, error, dcpl )
CALL check("h5dcreate_f",error,total_error)
CALL h5dcreate_f(fid, DSET3_NAME, H5T_NATIVE_CHARACTER, sid, dset3, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
END SELECT
DO u = 0, max_compact - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(my_dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error)
CALL check("h5acreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, u, data_dims, error)
CALL check("h5awrite_f",error,total_error)
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
END DO
END DO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dset1, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset2, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset3, error)
CALL check("h5dclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl, error)
CALL check("h5pclose_f",error,total_error)
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error, fapl)
CALL check("h5open_f",error,total_error)
CALL h5dopen_f(fid, DSET1_NAME, dset1, error)
CALL check("h5dopen_f",error,total_error)
CALL h5dopen_f(fid, DSET2_NAME, dset2, error)
CALL check("h5dopen_f",error,total_error)
CALL h5dopen_f(fid, DSET3_NAME, dset3, error)
CALL check("h5dopen_f",error,total_error)
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
CASE DEFAULT
WRITE(*,*) " WARNING: To many data sets! "
END SELECT
DO u = 0,max_compact-1
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve information for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aget_info_by_name_f(my_dataset, ".", attrname, &
f_corder_valid, corder, cset, data_size, error, lapl_id = H5P_DEFAULT_F ) !with optional
CALL check("H5Aget_info_by_name_f", error, total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify creation order of attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_name_f", f_corder_valid, .TRUE., total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL verify("H5Aget_info_by_name_f", corder, u, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve information for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aget_info_by_name_f(my_dataset, ".", attrname, &
f_corder_valid, corder, cset, data_size, error) ! without optional
CALL check("H5Aget_info_by_name_f", error, total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify creation order of attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_name_f", f_corder_valid, .TRUE., total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL verify("H5Aget_info_by_name_f", corder, u, total_error)
END DO
END DO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dset1, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset2, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset3, error)
CALL check("h5dclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
END SUBROUTINE test_attr_corder_create_compact
SUBROUTINE test_attr_null_space(fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
!**
!** test_attr_null_space(): Test basic H5A (attribute) code.
!** Tests storing attribute with "null" dataspace
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
[svn-r24967] Maintenance: Reorganized and cleaned the code to remove compiler warnings in the Fortran test code and examples. Platforms tested: Manual testing in place and using srcdir on jam, platypus, and emu with default and PGI, Intel and new GNU compilers. ifort compiler was also tested with -i8 and -r8 flags on jam. CMake tested on jam.
2014-04-06 23:56:21 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: sid, null_sid
INTEGER(HID_T) :: dataset
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
INTEGER :: error
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: value_scalar
INTEGER, DIMENSION(1) :: value
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER(HID_T) :: attr !String Attribute identifier
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER(HID_T) :: attr_sid
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER(HSIZE_T) :: storage_size ! attributes storage requirements
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
LOGICAL :: equal
! test: H5Sextent_equal_f
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error)
CALL check("h5open_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "null" dataspace for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_NULL_F, null_sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create a dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dataset, error)
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Add attribute with 'null' dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, "null attr", H5T_NATIVE_INTEGER, null_sid, attr, error)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Try to read data from the attribute
! (shouldn't fail, but should leave buffer alone)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
value(1) = 103
[svn-r21475] Description: fixed C_LOC(X) non-standard argument X for multi-dimensional arrays detected using xlf compiler. Tested: jam (intel, gnu, pgi) F2003 compilers.
2011-10-05 12:51:08 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aread_f(attr, H5T_NATIVE_INTEGER, value, data_dims, error)
CALL check("h5aread_f",error,total_error)
CALL verify("h5aread_f",value(1),103,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Try to read data from the attribute again but
! for a scalar
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
value_scalar = 104
data_dims(1) = 1
CALL h5aread_f(attr, H5T_NATIVE_INTEGER, value_scalar, data_dims, error)
CALL check("h5aread_f",error,total_error)
CALL verify("h5aread_f",value_scalar,104,total_error)
CALL h5aget_space_f(attr, attr_sid, error)
CALL check("h5aget_space_f",error,total_error)
CALL H5Sextent_equal_f(attr_sid, null_sid, equal, error)
CALL check("H5Sextent_equal_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Sextent_equal_f",equal,.TRUE.,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_storage_size_f(attr, storage_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_storage_size_f",INT(storage_size),0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_f(attr, f_corder_valid, corder, cset, data_size, error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
CALL check("h5aget_info_f", error, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check the attribute's information
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f.corder",corder,0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f.cset", cset, H5T_CSET_ASCII_F, total_error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
CALL h5aget_storage_size_f(attr, storage_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f.data_size", INT(data_size), INT(storage_size), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr,error)
CALL check("h5aclose_f",error,total_error)
CALL H5Sclose_f(attr_sid, error)
CALL check("H5Sclose_f",error,total_error)
CALL H5Dclose_f(dataset, error)
CALL check("H5Dclose_f", error,total_error)
CALL H5Fclose_f(fid, error)
CALL check("H5Fclose_f", error,total_error)
CALL H5Sclose_f(sid, error)
CALL check("H5Sclose_f", error,total_error)
CALL H5Sclose_f(null_sid, error)
CALL check("H5Sclose_f", error,total_error)
END SUBROUTINE test_attr_null_space
SUBROUTINE test_attr_create_by_name(new_format,fcpl,fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
!**
!** test_attr_create_by_name(): Test basic H5A (attribute) code.
!** Tests creating attributes by name
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
INTEGER(SIZE_T), PARAMETER :: NAME_BUF_SIZE = 7
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: new_format
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
INTEGER :: max_compact,min_dense,u
CHARACTER (LEN=NAME_BUF_SIZE) :: attrname
CHARACTER(LEN=8) :: dsetname
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
CHARACTER(LEN=8) :: DSET3_NAME = "Dataset3"
INTEGER, PARAMETER :: NUM_DSETS = 3
INTEGER :: curr_dset
INTEGER(HID_T) :: dset1, dset2, dset3
INTEGER(HID_T) :: my_dataset
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
CHARACTER(LEN=2) :: chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL, DIMENSION(1:2) :: use_index = (/.FALSE.,.TRUE./)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: Input1
INTEGER :: i
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset & attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataset creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the attribute creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over using index for creation order value
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO i = 1, 2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Print appropriate test message
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
IF(use_index(i))THEN
WRITE(*,*) " - Testing Creating Attributes By Name w/Creation Order Index"
ELSE
WRITE(*,*) " - Testing Creating Attributes By Name w/o Creation Order Index"
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set attribute creation order tracking & indexing for object
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format)THEN
IF(use_index(i))THEN
Input1 = H5P_CRT_ORDER_INDEXED_F
ELSE
Input1 = 0
ENDIF
CALL H5Pset_attr_creation_order_f(dcpl, IOR(H5P_CRT_ORDER_TRACKED_F, Input1), error)
CALL check("H5Pset_attr_creation_order",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dset1, error, dcpl_id=dcpl )
CALL check("h5dcreate_f2",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dset2, error, dcpl_id=dcpl )
CALL check("h5dcreate_f3",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET3_NAME, H5T_NATIVE_CHARACTER, sid, dset3, error, dcpl_id=dcpl )
CALL check("h5dcreate_f4",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
dsetname = DSET1_NAME
CASE (1)
my_dataset = dset2
dsetname = DSET2_NAME
CASE (2)
my_dataset = dset3
dsetname = DSET3_NAME
! CASE DEFAULT
! CALL HDassert(0.AND."Toomanydatasets!")
END SELECT
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attributes, up to limit of compact form
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_compact - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL H5Acreate_by_name_f(fid, dsetname, attrname, H5T_NATIVE_INTEGER, sid, &
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
attr, error, lapl_id=H5P_DEFAULT_F, acpl_id=H5P_DEFAULT_F, aapl_id=H5P_DEFAULT_F)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Acreate_by_name_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, u, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify information for NEW attribute
[svn-r20033] Porblem: A typo caused compilation failures with gfortran 4.6 on fred. Fixed. Platforms tested: fred and jam (minor change)
2011-02-03 09:24:52 +08:00
CALL attr_info_by_idx_check(my_dataset, attrname, INT(u,HSIZE_T), use_index(i), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! CALL check("FAILED IN attr_info_by_idx_check",total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Test opening attributes stored compactly
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL attr_open_check(fid, dsetname, my_dataset, u, total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
dsetname = DSET1_NAME
CASE (1)
my_dataset = dset2
dsetname = DSET2_NAME
CASE (2)
my_dataset = dset3
dsetname = DSET3_NAME
END SELECT
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create more attributes, to push into dense form
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = max_compact, max_compact* 2 - 1
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL H5Acreate_by_name_f(fid, dsetname, attrname, H5T_NATIVE_INTEGER, sid, &
attr, error, lapl_id=H5P_DEFAULT_F)
CALL check("H5Acreate_by_name",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, u, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dset1, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset2, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset3, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl, error)
CALL check("h5pclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_create_by_name
SUBROUTINE test_attr_info_by_idx(new_format, fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
!**
!** test_attr_info_by_idx(): Test basic H5A (attribute) code.
!** Tests querying attribute info by index
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
LOGICAL :: new_format
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
CHARACTER(LEN=8) :: DSET3_NAME = "Dataset3"
INTEGER, PARAMETER :: NUM_DSETS = 3
INTEGER :: curr_dset
INTEGER(HID_T) :: dset1, dset2, dset3
INTEGER(HID_T) :: my_dataset
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
INTEGER(HSIZE_T) :: n
LOGICAL, DIMENSION(1:2) :: use_index = (/.FALSE.,.TRUE./)
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=2) :: chr2
INTEGER :: i, j
INTEGER, DIMENSION(1) :: attr_integer_data
CHARACTER(LEN=7) :: attrname
INTEGER(SIZE_T) :: size
CHARACTER(LEN=80) :: tmpname
INTEGER :: Input1
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
INTEGER(HSIZE_T) :: hzero = 0_HSIZE_T
INTEGER :: minusone = -1
INTEGER(HSIZE_T) :: htmp
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset & attributes
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataset creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the attribute creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over using index for creation order value
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO i = 1, 2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set attribute creation order tracking & indexing for object
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format)THEN
IF(use_index(i))THEN
Input1 = H5P_CRT_ORDER_INDEXED_F
ELSE
Input1 = 0
ENDIF
CALL H5Pset_attr_creation_order_f(dcpl, IOR(H5P_CRT_ORDER_TRACKED_F, Input1), error)
CALL check("H5Pset_attr_creation_order",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create datasets
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dset1, error )
CALL check("h5dcreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dset2, error )
CALL check("h5dcreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET3_NAME, H5T_NATIVE_CHARACTER, sid, dset3, error )
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
END SELECT
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check for query on non-existant attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
n = 0
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS --
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
! 1) call by passing an integer with the _hsize_t declaration
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
CALL h5aget_info_by_idx_f(my_dataset, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, 0_hsize_t, &
f_corder_valid, corder, cset, data_size, error, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",error,minusone,total_error)
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! 2) call by passing an integer with the INT(,hsize_t) declaration
CALL h5aget_info_by_idx_f(my_dataset, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, INT(0,hsize_t), &
f_corder_valid, corder, cset, data_size, error, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",error,minusone,total_error)
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! 3) call by passing a variable with the attribute hsize_t
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL h5aget_info_by_idx_f(my_dataset, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",error,minusone,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_name_by_idx_f(my_dataset, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, &
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
hzero, tmpname, error, size, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f",error,minusone,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attributes, up to limit of compact form
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO j = 0, max_compact-1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') j
attrname = 'attr '//chr2
! check with the optional information create2 specs.
CALL h5acreate_f(my_dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = j
data_dims(1) = 1
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify information for new attribute
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
!EP CALL attr_info_by_idx_check(my_dataset, attrname, INT(j,HSIZE_T), use_index(i), total_error )
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
htmp = j
CALL attr_info_by_idx_check(my_dataset, attrname, htmp, use_index(i), total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!CHECK(ret, FAIL, "attr_info_by_idx_check");
ENDDO
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dset1, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset2, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset3, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
END DO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl,error)
CALL check("h5pclose_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_info_by_idx
SUBROUTINE attr_info_by_idx_check(obj_id, attrname, n, use_index, total_error )
IMPLICIT NONE
INTEGER :: error, total_error
[svn-r14928] Maintenance: Cleaned up the code to make it compile on smirom with the g95 compiler Platforms tested: kagiso with PGI compilers, linew, smirom with GCC and g95 compilers; some tests and function calls are commented out with !EP string; we will be working on it.
2008-05-05 00:48:07 +08:00
INTEGER(HID_T) :: obj_id
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=*) :: attrname
INTEGER(HSIZE_T) :: n
LOGICAL :: use_index
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
INTEGER(SIZE_T) :: NAME_BUF_SIZE = 7
CHARACTER(LEN=7) :: tmpname
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
INTEGER(HSIZE_T) :: hzero = 0_HSIZE_T
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the information for first attribute, in increasing creation order
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5aget_info_by_idx_f",error,total_error)
CALL verify("h5aget_info_by_idx_f",corder,0,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the information for new attribute, in increasing creation order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, n, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the name for new link, in increasing creation order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
! Try with the correct buffer size
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_name_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, &
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
n, tmpname, error, NAME_BUF_SIZE)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5aget_name_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f", INT(NAME_BUF_SIZE), 7, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
IF(attrname.NE.tmpname)THEN
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
error = -1
ENDIF
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f",error,0,total_error)
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Don't test "native" order if there is no creation order index, since
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! * there's not a good way to easily predict the attribute's order in the name
! * index.
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF (use_index) THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the information for first attribute, in native creation order
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_NATIVE_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the information for new attribute, in native creation order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_NATIVE_F, n, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the name for new link, in increasing native order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_name_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_NATIVE_F, &
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
n, tmpname, error) ! check with no optional parameters
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5aget_name_by_idx_f",error,total_error)
IF(TRIM(attrname).NE.TRIM(tmpname))THEN
WRITE(*,*) "ERROR: attribute name size wrong!"
error = -1
ENDIF
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f",error,0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
END IF
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_DEC_F, n, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,0,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_DEC_F, 0_HSIZE_T, &
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS --
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
! 1) call by passing an integer with the _hsize_t declaration
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_DEC_F, 0_HSIZE_T, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! 2) call by passing an integer with the INT(,hsize_t) declaration
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_DEC_F, INT(0,HSIZE_T), &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
! 3) call by passing a variable with the attribute hsize_t
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_DEC_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r14955] Purpose: * test passing integer constant to subroutine Description: ! -- CHECK PASSING AN INTEGER CONSTANT IN DIFFERENT FORMS -- ! 1) call by passing an integer with the _hsize_t declaration ! 2) call by passing an integer with the INT(,hsize_t) declaration ! 3) call by passing a variable with the attribute hsize_t
2008-05-09 01:04:11 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_INC_F, 0_HSIZE_T, &
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_INC_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_INC_F, n, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_DEC_F, n, &
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,0,total_error)
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_DEC_F, 0_HSIZE_T, &
CALL h5aget_info_by_idx_f(obj_id, ".", H5_INDEX_NAME_F, H5_ITER_DEC_F, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_by_idx_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_by_idx_f",corder,INT(n),total_error)
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
END SUBROUTINE attr_info_by_idx_check
SUBROUTINE test_attr_shared_rename( fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_shared_rename(): Test basic H5A (attribute) code.
!** Tests renaming shared attributes in "compact" & "dense" storage
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid, big_sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
INTEGER(HID_T) :: dataset, dataset2
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HID_T) :: attr_tid
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=2) :: chr2
INTEGER, DIMENSION(1) :: attr_integer_data
CHARACTER(LEN=7) :: attrname
CHARACTER(LEN=11) :: attrname2
INTEGER :: u
INTEGER(HID_T) :: my_fcpl
CHARACTER(LEN=5), PARAMETER :: TYPE1_NAME = "/Type"
INTEGER :: test_shared
INTEGER(HSIZE_T), DIMENSION(1) :: adims2 = (/1/) ! Attribute dimension
INTEGER :: arank = 1 ! Attribure rank
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Initialize "big" attribute data
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" dataspace for "large" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_simple_f(arank, adims2, big_sid, error)
CALL check("h5screate_simple_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over type of shared components
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO test_shared = 0, 2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make copy of file creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcopy_f(fcpl, my_fcpl, error)
CALL check("H5Pcopy",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up datatype for attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Tcopy_f(H5T_NATIVE_INTEGER, attr_tid, error)
CALL check("H5Tcopy",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Special setup for each type of shared components
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF( test_shared .EQ. 0) THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make attributes > 500 bytes shared
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_nindexes_f(my_fcpl,1,error)
CALL check("H5Pset_shared_mesg_nindexes_f",error, total_error)
CALL H5Pset_shared_mesg_index_f(my_fcpl, 0, H5O_SHMESG_ATTR_FLAG_F, 500,error)
CALL check(" H5Pset_shared_mesg_index_f",error, total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up copy of file creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_nindexes_f(my_fcpl,3,error)
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make attributes > 500 bytes shared
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
CALL H5Pset_shared_mesg_index_f(my_fcpl, 0, H5O_SHMESG_ATTR_FLAG_F, 500,error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make datatypes & dataspaces > 1 byte shared (i.e. all of them :-)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_index_f(my_fcpl, 1, H5O_SHMESG_DTYPE_FLAG_F, 1,error)
CALL H5Pset_shared_mesg_index_f(my_fcpl, 2, H5O_SHMESG_SDSPACE_FLAG_F, 1,error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, my_fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close FCPL copy
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(my_fcpl, error)
CALL check("h5pclose_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error,fapl)
CALL check("h5open_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Commit datatype to file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(test_shared.EQ.2) THEN
[svn-r14928] Maintenance: Cleaned up the code to make it compile on smirom with the g95 compiler Platforms tested: kagiso with PGI compilers, linew, smirom with GCC and g95 compilers; some tests and function calls are commented out with !EP string; we will be working on it.
2008-05-05 00:48:07 +08:00
CALL H5Tcommit_f(fid, TYPE1_NAME, attr_tid, error, H5P_DEFAULT_F, H5P_DEFAULT_F, H5P_DEFAULT_F)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Tcommit",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up to query the object creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dataset, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dataset2, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve limits for compact/dense attribute storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl,error)
CALL check("h5pclose_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Add attributes to each dataset, until after converting to dense storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, (max_compact * 2) - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Alternate between creating "small" & "big" attributes
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(MOD(u+1,2).EQ.0)THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "small" attribute on first dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, attrname, attr_tid, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" attribute on first dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, attrname, attr_tid, big_sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
data_dims(1) = 1
attr_integer_data(1) = u + 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Alternate between creating "small" & "big" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(MOD(u+1,2).EQ.0)THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "small" attribute on second dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset2, attrname, attr_tid, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" attribute on second dataset
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset2, attrname, attr_tid, big_sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
! CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
! CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create new attribute name
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname2 = 'new attr '//chr2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Change second dataset's attribute's name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Arename_by_name_f(fid, DSET2_NAME, attrname, attrname2, error, lapl_id=H5P_DEFAULT_F)
CALL check("H5Arename_by_name_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on attributes now
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on renamed attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_f(dataset2, attrname2, attr, error, aapl_id=H5P_DEFAULT_F)
CALL check("H5Aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on original attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_f(dataset, attrname, attr, error)
CALL check("H5Aopen",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Change second dataset's attribute's name back to original
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Arename_by_name_f(fid, DSET2_NAME, attrname2, attrname, error)
CALL check("H5Arename_by_name_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on attributes now
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on renamed attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_f(dataset2, attrname, attr, error)
CALL check("H5Aopen",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on original attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check refcount on renamed attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_f(dataset, attrname, attr, error)
CALL check("H5Aopen",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute's datatype
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5tclose_f(attr_tid, error)
CALL check("h5tclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute's datatype
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dataset2, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Unlink datasets with attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Ldelete_f(fid, DSET1_NAME, error, H5P_DEFAULT_F)
CALL check("HLdelete",error,total_error)
CALL H5Ldelete_f(fid, DSET2_NAME, error)
CALL check("HLdelete",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Unlink committed datatype
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(test_shared == 2)THEN
CALL H5Ldelete_f(fid, TYPE1_NAME, error)
CALL check("HLdelete_f",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check size of file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!filesize = h5_get_file_size(FILENAME);
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
!verify(filesize, empty_filesize, "h5_get_file_size");
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspaces
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
CALL h5sclose_f(big_sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_shared_rename
SUBROUTINE test_attr_delete_by_idx(new_format, fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
!**
!** test_attr_delete_by_idx(): Test basic H5A (attribute) code.
!** Tests deleting attribute by index
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r15054] Purpose: Made reporting of the test status global by handling the output via a module. Cleaned-up output to the terminal. Description: Put writing the test status as a call to a subroutine instead of on a per account basis. Added the dependency of compiling in the correct order in the Makefiles for use of the Module.
2008-05-21 23:02:24 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
LOGICAL, INTENT(IN) :: new_format
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
INTEGER(HID_T) :: fid ! HDF5 File ID
INTEGER(HID_T) :: dcpl ! Dataset creation property list ID
INTEGER(HID_T) :: sid ! Dataspace ID
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
CHARACTER(LEN=8) :: DSET3_NAME = "Dataset3"
INTEGER, PARAMETER :: NUM_DSETS = 3
INTEGER :: curr_dset
INTEGER(HID_T) :: dset1, dset2, dset3
INTEGER(HID_T) :: my_dataset
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
LOGICAL, DIMENSION(1:2) :: use_index = (/.FALSE.,.TRUE./)
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=2) :: chr2
INTEGER :: i
INTEGER, DIMENSION(1) :: attr_integer_data
CHARACTER(LEN=7) :: attrname
INTEGER(SIZE_T) :: size
CHARACTER(LEN=8) :: tmpname
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: idx_type
INTEGER :: order
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
INTEGER :: u ! Local index variable
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: Input1
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
INTEGER(HSIZE_T) :: hzero = 0_HSIZE_T
INTEGER :: minusone = -1
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset & attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataset creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the attribute creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over operating on different indices on link fields
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO idx_type = H5_INDEX_NAME_F, H5_INDEX_CRT_ORDER_F
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over operating in different orders
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO order = H5_ITER_INC_F, H5_ITER_DEC_F
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over using index for creation order value
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO i = 1, 2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set attribute creation order tracking & indexing for object
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format)THEN
IF(use_index(i))THEN
Input1 = H5P_CRT_ORDER_INDEXED_F
ELSE
Input1 = 0
ENDIF
CALL H5Pset_attr_creation_order_f(dcpl, IOR(H5P_CRT_ORDER_TRACKED_F, Input1), error)
CALL check("H5Pset_attr_creation_order",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create datasets
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dset1, error, dcpl )
CALL check("h5dcreate_f2",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dset2, error, dcpl )
CALL check("h5dcreate_f3",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET3_NAME, H5T_NATIVE_CHARACTER, sid, dset3, error, dcpl )
CALL check("h5dcreate_f4",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
! CASE DEFAULT
! CALL HDassert(0.AND."Toomanydatasets!")
END SELECT
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check for deleting non-existant attribute
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL H5Adelete_by_idx_f(my_dataset, '.', idx_type, order, 0_HSIZE_T,error, lapl_id=H5P_DEFAULT_F)
CALL H5Adelete_by_idx_f(my_dataset, '.', idx_type, order, hzero,error, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Adelete_by_idx_f",error,minusone,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attributes, up to limit of compact form
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_compact - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(my_dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u
data_dims(1) = 1
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify information for new attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL attr_info_by_idx_check(my_dataset, attrname, INT(u,HSIZE_T), use_index(i), total_error )
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check for out of bound deletions
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, INT(u,HSIZE_T), error, lapl_id=H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Adelete_by_idx_f",error,minusone,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
DO curr_dset = 0, NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
! CASE DEFAULT
! CALL HDassert(0.AND."Toomanydatasets!")
END SELECT
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete attributes from compact storage
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_compact - 2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete first attribute in appropriate order
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, 0_HSIZE_T, error)
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, hzero, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Adelete_by_idx_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the attribute information for first attribute in appropriate order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! HDmemset(&ainfo, 0, sizeof(ainfo));
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL h5aget_info_by_idx_f(my_dataset, ".", idx_type, order, 0_HSIZE_T, &
CALL h5aget_info_by_idx_f(my_dataset, ".", idx_type, order, hzero, &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
f_corder_valid, corder, cset, data_size, error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format)THEN
IF(order.EQ.H5_ITER_INC_F)THEN
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_idx_f",corder,u + 1,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
ELSE
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_idx_f",corder, max_compact-(u + 2),total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the name for first attribute in appropriate order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
size = 7 ! *CHECK* IF NOT THE SAME SIZE
CALL h5aget_name_by_idx_f(my_dataset, ".", idx_type, order,INT(0,hsize_t), &
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
tmpname, error, lapl_id=H5P_DEFAULT_F, size=size)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check('h5aget_name_by_idx_f',error,total_error)
IF(order .EQ. H5_ITER_INC_F)THEN
WRITE(chr2,'(I2.2)') u + 1
attrname = 'attr '//chr2
ELSE
WRITE(chr2,'(I2.2)') max_compact - (u + 2)
attrname = 'attr '//chr2
ENDIF
IF(TRIM(attrname).NE.TRIM(tmpname)) error = -1
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f",error,0,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete last attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
!EP CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, 0_HSIZE_T, error)
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, hzero, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Adelete_by_idx_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
! CASE DEFAULT
! CALL HDassert(0.AND."Toomanydatasets!")
END SELECT
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create more attributes, to push into dense form
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, (max_compact * 2) - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(my_dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u
data_dims(1) = 1
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check for out of bound deletion
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, INT(u,HSIZE_T), error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Adelete_by_idx_f",error,minusone,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Work on all the datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO curr_dset = 0,NUM_DSETS-1
SELECT CASE (curr_dset)
CASE (0)
my_dataset = dset1
CASE (1)
my_dataset = dset2
CASE (2)
my_dataset = dset3
END SELECT
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete attributes from dense storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, (max_compact * 2) - 1 - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete first attribute in appropriate order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, INT(0,HSIZE_T), error)
CALL check("H5Adelete_by_idx_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the attribute information for first attribute in appropriate order
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_by_idx_f(my_dataset, ".", idx_type, order, INT(0,HSIZE_T), &
f_corder_valid, corder, cset, data_size, error)
IF(new_format)THEN
IF(order.EQ.H5_ITER_INC_F)THEN
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_idx_f",corder,u + 1,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
ELSE
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aget_info_by_idx_f",corder, ((max_compact * 2) - (u + 2)), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify the name for first attribute in appropriate order
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! HDmemset(tmpname, 0, (size_t)NAME_BUF_SIZE);
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
size = 7 ! *CHECK* if not the correct size
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_name_by_idx_f(my_dataset, ".", idx_type, order,INT(0,hsize_t), &
[svn-r14996] Purpose: Updated the call using the new interface for the get_name_by_idx_f routine
2008-05-15 02:58:44 +08:00
tmpname, error, size)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(order .EQ. H5_ITER_INC_F)THEN
WRITE(chr2,'(I2.2)') u + 1
attrname = 'attr '//chr2
ELSE
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
WRITE(chr2,'(I2.2)') max_compact * 2 - (u + 2)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attrname = 'attr '//chr2
ENDIF
IF(TRIM(attrname).NE.TRIM(tmpname)) error = -1
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_name_by_idx_f",error,0,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete last attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, INT(0,HSIZE_T), error, lapl_id=H5P_DEFAULT_F)
CALL check("H5Adelete_by_idx_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check for deletion on empty attribute storage again
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_idx_f(my_dataset, ".", idx_type, order, INT(0,HSIZE_T), error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Adelete_by_idx_f",error,minusone,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dset1, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset2, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dset3, error)
CALL check("h5dclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
ENDDO
ENDDO
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl,error)
CALL check("h5pclose_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_delete_by_idx
SUBROUTINE test_attr_shared_delete(fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_shared_delete(): Test basic H5A (attribute) code.
!** Tests deleting shared attributes in "compact" & "dense" storage
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid, big_sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
CHARACTER(LEN=8) :: DSET2_NAME = "Dataset2"
INTEGER(HID_T) :: dataset, dataset2
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HID_T) :: attr_tid
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=2) :: chr2
INTEGER, DIMENSION(1) :: attr_integer_data
CHARACTER(LEN=7) :: attrname
INTEGER :: u
INTEGER(HID_T) :: my_fcpl
CHARACTER(LEN=5), PARAMETER :: TYPE1_NAME = "/Type"
INTEGER :: test_shared
INTEGER(HSIZE_T), DIMENSION(1) :: adims2 = (/1/) ! Attribute dimension
INTEGER :: arank = 1 ! Attribure rank
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Output message about test being performed
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Initialize "big" attribute DATA
! Create dataspace for dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" dataspace for "large" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_simple_f(arank, adims2, big_sid, error)
CALL check("h5screate_simple_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Loop over type of shared components
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO test_shared = 0, 2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make copy of file creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcopy_f(fcpl, my_fcpl, error)
CALL check("H5Pcopy",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up datatype for attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Tcopy_f(H5T_NATIVE_INTEGER, attr_tid, error)
CALL check("H5Tcopy",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Special setup for each type of shared components
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF( test_shared .EQ. 0) THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make attributes > 500 bytes shared
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_nindexes_f(my_fcpl,1,error)
CALL check("H5Pset_shared_mesg_nindexes_f",error, total_error)
CALL H5Pset_shared_mesg_index_f(my_fcpl, 0, H5O_SHMESG_ATTR_FLAG_F, 500,error)
CALL check(" H5Pset_shared_mesg_index_f",error, total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up copy of file creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_nindexes_f(my_fcpl,3,error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make attributes > 500 bytes shared
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_index_f(my_fcpl, 0, H5O_SHMESG_ATTR_FLAG_F, 500,error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Make datatypes & dataspaces > 1 byte shared (i.e. all of them :-)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_shared_mesg_index_f(my_fcpl, 1, H5O_SHMESG_DTYPE_FLAG_F, 1,error)
CALL H5Pset_shared_mesg_index_f(my_fcpl, 2, H5O_SHMESG_SDSPACE_FLAG_F, 1,error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, my_fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close FCPL copy
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(my_fcpl, error)
CALL check("h5pclose_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error,fapl)
CALL check("h5open_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Commit datatype to file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(test_shared.EQ.2) THEN
[svn-r14928] Maintenance: Cleaned up the code to make it compile on smirom with the g95 compiler Platforms tested: kagiso with PGI compilers, linew, smirom with GCC and g95 compilers; some tests and function calls are commented out with !EP string; we will be working on it.
2008-05-05 00:48:07 +08:00
CALL H5Tcommit_f(fid, TYPE1_NAME, attr_tid, error, H5P_DEFAULT_F, H5P_DEFAULT_F, H5P_DEFAULT_F)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Tcommit",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set up to query the object creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dataset, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET2_NAME, H5T_NATIVE_CHARACTER, sid, dataset2, error, dcpl_id=dcpl )
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve limits for compact/dense attribute storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl,error)
CALL check("h5pclose_f", error, total_error)
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Add attributes to each dataset, until after converting to dense storage
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, (max_compact * 2) - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Alternate between creating "small" & "big" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(MOD(u+1,2).EQ.0)THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "small" attribute on first dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, attrname, attr_tid, sid, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" attribute on first dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, attrname, attr_tid, big_sid, attr, error)
CALL check("h5acreate_f",error,total_error)
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Alternate between creating "small" & "big" attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(MOD(u+1,2).EQ.0)THEN
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "small" attribute on second dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset2, attrname, attr_tid, sid, attr, error)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ELSE
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create "big" attribute on second dataset
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset2, attrname, attr_tid, big_sid, attr, error, acpl_id=H5P_DEFAULT_F, aapl_id=H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_integer_data(1) = u + 1
data_dims(1) = 1
CALL h5awrite_f(attr, attr_tid, attr_integer_data, data_dims, error)
CALL check("h5awrite_f",error,total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete attributes from second dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_compact*2-1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Delete second dataset's attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Adelete_by_name_f(fid, DSET2_NAME, attrname,error,lapl_id=H5P_DEFAULT_F)
CALL check("H5Adelete_by_name", error, total_error)
CALL h5aopen_f(dataset, attrname, attr, error, aapl_id=H5P_DEFAULT_F)
CALL check("h5aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute's datatype
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5tclose_f(attr_tid, error)
CALL check("h5tclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Datasets
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
CALL h5dclose_f(dataset2, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Unlink datasets WITH attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5ldelete_f(fid, DSET1_NAME, error, H5P_DEFAULT_F)
CALL check("H5Ldelete_f", error, total_error)
CALL h5ldelete_f(fid, DSET2_NAME, error)
CALL check("H5Ldelete_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Unlink committed datatype
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF( test_shared == 2) THEN
CALL h5ldelete_f(fid, TYPE1_NAME, error)
CALL check("H5Ldelete_f", error, total_error)
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspaces
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
CALL h5sclose_f(big_sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_shared_delete
SUBROUTINE test_attr_dense_open( fcpl, fapl, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_dense_open(): Test basic H5A (attribute) code.
!** Tests opening attributes in "dense" storage
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
[svn-r15417] Description: Changed the INTENT for total_error to be consistantly INTENT(INOUT) as noted in bug 1251
2008-07-29 03:19:11 +08:00
INTEGER, INTENT(INOUT) :: total_error
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
INTEGER :: error
INTEGER(HID_T) :: attr !String Attribute identifier
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
INTEGER :: max_compact ! Maximum # of links to store in group compactly
INTEGER :: min_dense ! Minimum # of links to store in group "densely"
CHARACTER(LEN=2) :: chr2
CHARACTER(LEN=7) :: attrname
INTEGER(HID_T) :: dataset
INTEGER :: u
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error, fapl)
CALL check("h5open_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the group creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Enable creation order tracking on attributes, so creation order tests work
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_attr_creation_order_f(dcpl, H5P_CRT_ORDER_TRACKED_F, error)
CALL check("H5Pset_attr_creation_order",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create a dataset
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dataset, error, &
lcpl_id=H5P_DEFAULT_F, dcpl_id=dcpl, dapl_id=H5P_DEFAULT_F)
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve limits for compact/dense attribute storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_phase_change_f(dcpl, max_compact, min_dense, error)
CALL check("H5Pget_attr_phase_change_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl, error)
CALL check("h5pclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Add attributes, until just before converting to dense storage
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_compact - 1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL h5acreate_f(dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error, aapl_id=H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, u, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify attributes written so far
[svn-r14954] re-instated the test_attr_dense_open routine for debugging
2008-05-08 23:59:38 +08:00
CALL test_attr_dense_verify(dataset, u, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r21248] Mereged the F2003 branch into the trunk. Items merged: fortran directory, src/libhdf5.settings.in configure.in configure MANIFEST Tested: (all platforms used by daily tests, both with --enable-fortran and --enable-fortran2003)
2011-08-18 22:32:47 +08:00
!
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Add one more attribute, to push into "dense" storage
! Create attribute
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL h5acreate_f(dataset, attrname, H5T_NATIVE_INTEGER, sid, attr, error, aapl_id=H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write data into the attribute
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
data_dims(1) = 1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, u, data_dims, error)
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify all the attributes written
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! ret = test_attr_dense_verify(dataset, (u + 1));
! CHECK(ret, FAIL, "test_attr_dense_verify");
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! CLOSE Dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Unlink dataset with attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5ldelete_f(fid, DSET1_NAME, error, H5P_DEFAULT_F)
CALL check("H5Ldelete_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check size of file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! filesize = h5_get_file_size(FILENAME);
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
! verify(filesize, empty_filesize, "h5_get_file_size")
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
END SUBROUTINE test_attr_dense_open
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_dense_verify(): Test basic H5A (attribute) code.
!** Verify attributes on object
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
SUBROUTINE test_attr_dense_verify(loc_id, max_attr, total_error)
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: loc_id
INTEGER, INTENT(IN) :: max_attr
INTEGER, INTENT(INOUT) :: total_error
INTEGER(SIZE_T), PARAMETER :: ATTR_NAME_LEN = 8 ! FIX, why if 7 does not work?
INTEGER :: u
CHARACTER(LEN=2) :: chr2
CHARACTER(LEN=ATTR_NAME_LEN) :: attrname
CHARACTER(LEN=ATTR_NAME_LEN) :: check_name
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER(HID_T) :: attr !String Attribute identifier
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: error
INTEGER :: value
data_dims = 0
[svn-r14954] re-instated the test_attr_dense_open routine for debugging
2008-05-08 23:59:38 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve the current # of reported errors
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! old_nerrs = GetTestNumErrs();
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open all the attributes by name and verify the data
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_attr -1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL h5aopen_f(loc_id, attrname, attr, error)
CALL check("h5aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Read data from the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! value = 103
data_dims(1) = 1
CALL h5aread_f(attr, H5T_NATIVE_INTEGER, value, data_dims, error)
CALL CHECK("H5Aread_F", error, total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aread_F", value, u, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open all the attributes by index and verify the data
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u=0, max_attr-1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_by_idx_f(loc_id, ".", H5_INDEX_CRT_ORDER_F, H5_ITER_INC_F, INT(u,HSIZE_T), &
attr, error, aapl_id=H5P_DEFAULT_F)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify Name
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
CALL H5Aget_name_f(attr, ATTR_NAME_LEN, check_name, error)
CALL check('H5Aget_name',error,total_error)
IF(check_name.NE.attrname) THEN
WRITE(*,*) 'ERROR: attribute name different: attr_name = ',check_name, ', should be ', attrname
total_error = total_error + 1
ENDIF
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Read data from the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
data_dims(1) = 1
CALL h5aread_f(attr, H5T_NATIVE_INTEGER, value, data_dims, error)
CALL CHECK("H5Aread_f", error, total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aread_f", value, u, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
END SUBROUTINE test_attr_dense_verify
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_corder_create_empty(): Test basic H5A (attribute) code.
!** Tests basic code to create objects with attribute creation order info
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
SUBROUTINE test_attr_corder_create_basic( fcpl, fapl, total_error )
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
[svn-r15417] Description: Changed the INTENT for total_error to be consistantly INTENT(INOUT) as noted in bug 1251
2008-07-29 03:19:11 +08:00
INTEGER, INTENT(INOUT) :: total_error
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: dcpl
INTEGER(HID_T) :: sid
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
INTEGER(HID_T) :: dataset
INTEGER :: error
INTEGER :: crt_order_flags
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
INTEGER :: minusone = -1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Output message about test being performed
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
! WRITE(*,*) " - Testing Basic Code for Attributes with Creation Order Info"
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataset creation property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pcreate_f(H5P_DATASET_CREATE_F,dcpl,error)
CALL check("h5Pcreate_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Get creation order indexing on object
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_creation_order_f(dcpl, crt_order_flags, error)
CALL check("H5Pget_attr_creation_order_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Pget_attr_creation_order_f",crt_order_flags , 0, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Setting invalid combination of a attribute order creation order indexing on should fail
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pset_attr_creation_order_f(dcpl, H5P_CRT_ORDER_INDEXED_F, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Pset_attr_creation_order_f",error , minusone, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_creation_order_f(dcpl, crt_order_flags, error)
CALL check("H5Pget_attr_creation_order_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Pget_attr_creation_order_f",crt_order_flags , 0, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Set attribute creation order tracking & indexing for object
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pset_attr_creation_order_f(dcpl, IOR(H5P_CRT_ORDER_TRACKED_F, H5P_CRT_ORDER_INDEXED_F), error)
CALL check("H5Pset_attr_creation_order_f",error,total_error)
CALL H5Pget_attr_creation_order_f(dcpl, crt_order_flags, error)
CALL check("H5Pget_attr_creation_order_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Pget_attr_creation_order_f",crt_order_flags , &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IOR(H5P_CRT_ORDER_TRACKED_F, H5P_CRT_ORDER_INDEXED_F), total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create a dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid, DSET1_NAME, H5T_NATIVE_CHARACTER, sid, dataset, error, &
lcpl_id=H5P_DEFAULT_F, dapl_id=H5P_DEFAULT_F, dcpl_id=dcpl)
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspace
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl, error)
CALL check("h5pclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Re-open file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fopen_f(FileName, H5F_ACC_RDWR_F, fid, error, fapl)
CALL check("h5open_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open dataset created
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dopen_f(fid, DSET1_NAME, dataset, error, H5P_DEFAULT_F )
CALL check("h5dopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Retrieve dataset creation property list for group
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Dget_create_plist_f(dataset, dcpl, error)
CALL check("H5Dget_create_plist_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Query the attribute creation properties
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Pget_attr_creation_order_f(dcpl, crt_order_flags, error)
CALL check("H5Pget_attr_creation_order_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Pget_attr_creation_order_f",crt_order_flags , &
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IOR(H5P_CRT_ORDER_TRACKED_F, H5P_CRT_ORDER_INDEXED_F), total_error )
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close property list
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5pclose_f(dcpl, error)
CALL check("h5pclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
END SUBROUTINE test_attr_corder_create_basic
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_basic_write(): Test basic H5A (attribute) code.
!** Tests integer attributes on both datasets and groups
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
SUBROUTINE test_attr_basic_write(fapl, total_error)
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fapl
INTEGER, INTENT(INOUT) :: total_error
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid1
INTEGER(HID_T) :: sid1, sid2
CHARACTER(LEN=8) :: DSET1_NAME = "Dataset1"
INTEGER(HID_T) :: dataset
INTEGER :: i
INTEGER :: error
INTEGER(HID_T) :: attr,attr2 !String Attribute identifier
INTEGER(HID_T) :: group
CHARACTER(LEN=25) :: check_name
CHARACTER(LEN=18) :: chr_exact_size
CHARACTER(LEN=5), PARAMETER :: ATTR1_NAME="Attr1"
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
INTEGER, PARAMETER :: ATTR1_RANK = 1
INTEGER, PARAMETER :: ATTR1_DIM1 = 3
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=7), PARAMETER :: ATTR1A_NAME ="Attr1_a"
CHARACTER(LEN=18), PARAMETER :: ATTR_TMP_NAME = "Attr1_a-1234567890"
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
INTEGER, DIMENSION(ATTR1_DIM1) :: attr_data1
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER, DIMENSION(ATTR1_DIM1) :: attr_data1a
INTEGER, DIMENSION(ATTR1_DIM1) :: read_data1
INTEGER(HSIZE_T) :: attr_size ! attributes storage requirements .MSB.
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
INTEGER(HSIZE_T), DIMENSION(1) :: dimsa = (/3/) ! Dataset dimensions
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: rank1 = 2 ! Dataspace1 rank
INTEGER(HSIZE_T), DIMENSION(2) :: dims1 = (/4,6/) ! Dataset dimensions
INTEGER(HSIZE_T), DIMENSION(2) :: maxdims1 = (/4,6/) ! maximum dimensions
INTEGER(SIZE_T) :: size
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
!! Initialize attribute data
attr_data1(1) = 258
attr_data1(2) = 9987
attr_data1(3) = -99890
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_data1a(1) = 258
attr_data1a(2) = 1087
attr_data1a(3) = -99890
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid1, error, H5P_DEFAULT_F, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_simple_f(rank1, dims1, sid1, error, maxdims1)
CALL check("h5screate_simple_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create a dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dcreate_f(fid1, DSET1_NAME, H5T_NATIVE_CHARACTER, sid1, dataset, error, H5P_DEFAULT_F, H5P_DEFAULT_F, H5P_DEFAULT_F )
CALL check("h5dcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for attribute
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
CALL h5screate_simple_f(ATTR1_RANK, dimsa, sid2, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5screate_simple_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Try to create an attribute on the file (should create an attribute on root group)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(fid1, ATTR1_NAME, H5T_NATIVE_INTEGER, sid2, attr, error, aapl_id=H5P_DEFAULT_F, acpl_id=H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open the root group
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Gopen_f(fid1, "/", group, error, H5P_DEFAULT_F)
CALL check("H5Gopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open attribute again
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aopen_f(group, ATTR1_NAME, attr, error)
CALL check("h5aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close root group
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Gclose_f(group, error)
CALL check("h5gclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create an attribute for the dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, ATTR1_NAME, H5T_NATIVE_INTEGER, sid2, attr, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write attribute information
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
CALL h5awrite_f(attr, H5T_NATIVE_INTEGER, attr_data1, dimsa, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create an another attribute for the dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(dataset, ATTR1A_NAME, H5T_NATIVE_INTEGER, sid2, attr2, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Write attribute information
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
CALL h5awrite_f(attr2, H5T_NATIVE_INTEGER, attr_data1a, dimsa, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5awrite_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check storage size for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_storage_size_f(attr, attr_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
!EP CALL verify("h5aget_storage_size_f", INT(attr_size), 2*HSIZE_T, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Read attribute information immediately, without closing attribute
[svn-r14941] Maintenance: Cleaned up more test code, fixed more bugs in the test and source codes. This check-in should address Fortran failures on liberty and smirom. Platforms tested: kagiso with Intel, smirom with g95 -fPIC, liberty with gfortran42
2008-05-07 02:01:59 +08:00
CALL h5aread_f(attr, H5T_NATIVE_INTEGER, read_data1, dimsa, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("h5aread_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify values read in
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO i = 1, ATTR1_DIM1
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify('h5aread_f',attr_data1(i),read_data1(i), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! CLOSE attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr2, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! change attribute name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Arename_f(dataset, ATTR1_NAME, ATTR_TMP_NAME, error)
CALL check("H5Arename_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open attribute again
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aopen_f(dataset, ATTR_TMP_NAME, attr, error)
CALL check("h5aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Verify new attribute name
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! Set a deliberately small size
check_name = ' ' ! need to initialize or does not pass test
size = 1
CALL H5Aget_name_f(attr, size, check_name, error)
CALL check('H5Aget_name',error,total_error)
! Now enter with the corrected size
IF(error.NE.size)THEN
size = error
CALL H5Aget_name_f(attr, size, check_name, error)
CALL check('H5Aget_name',error,total_error)
ENDIF
IF(TRIM(ADJUSTL(check_name)).NE.TRIM(ADJUSTL(ATTR_TMP_NAME))) THEN
PRINT*,'.'//TRIM(check_name)//'.',LEN_TRIM(check_name)
PRINT*,'.'//TRIM(ATTR_TMP_NAME)//'.',LEN_TRIM(ATTR_TMP_NAME)
WRITE(*,*) 'ERROR: attribute name different: attr_name ='//TRIM(check_name)//'.'
WRITE(*,*) ' should be ='//TRIM(ATTR_TMP_NAME)//'.'
total_error = total_error + 1
stop
ENDIF
! Try with a string buffer that is exactly the correct size
size = 18
CALL H5Aget_name_f(attr, size, chr_exact_size, error)
CALL check('H5Aget_name_f',error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify('H5Aget_name_f',chr_exact_size,ATTR_TMP_NAME, total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr, error)
CALL check("h5aclose_f",error,total_error)
CALL h5sclose_f(sid1, error)
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL check("h5sclose_f",error,total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid2, error)
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL check("h5sclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close Dataset
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5dclose_f(dataset, error)
CALL check("h5dclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid1, error)
CALL check("h5fclose_f",error,total_error)
END SUBROUTINE test_attr_basic_write
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
!**
!** test_attr_many(): Test basic H5A (attribute) code.
!** Tests storing lots of attributes
!**
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!***************************************************************
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
SUBROUTINE test_attr_many(new_format, fcpl, fapl, total_error)
IMPLICIT NONE
LOGICAL, INTENT(IN) :: new_format
INTEGER(HID_T), INTENT(IN) :: fcpl
INTEGER(HID_T), INTENT(IN) :: fapl
[svn-r15417] Description: Changed the INTENT for total_error to be consistantly INTENT(INOUT) as noted in bug 1251
2008-07-29 03:19:11 +08:00
INTEGER, INTENT(INOUT) :: total_error
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=8) :: FileName = "tattr.h5"
INTEGER(HID_T) :: fid
INTEGER(HID_T) :: sid
INTEGER(HID_T) :: gid
INTEGER(HID_T) :: aid
INTEGER :: error
INTEGER(HSIZE_T), DIMENSION(7) :: data_dims
CHARACTER(LEN=5) :: chr5
CHARACTER(LEN=11) :: attrname
CHARACTER(LEN=8), PARAMETER :: GROUP1_NAME="/Group1"
INTEGER :: u
INTEGER :: nattr
LOGICAL :: exists
INTEGER, DIMENSION(1) :: attr_data1
data_dims = 0
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fcreate_f(FileName, H5F_ACC_TRUNC_F, fid, error, fcpl, fapl)
CALL check("h5fcreate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create dataspace for attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5screate_f(H5S_SCALAR_F, sid, error)
CALL check("h5screate_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create group for attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
CALL H5Gcreate_f(fid, GROUP1_NAME, gid, error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL check("H5Gcreate_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Create many attributes
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
IF(new_format)THEN
nattr = 250
ELSE
nattr = 2
ENDIF
DO u = 0, nattr - 1
WRITE(chr5,'(I5.5)') u
attrname = 'attr '//chr5
CALL H5Aexists_f( gid, attrname, exists, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists",exists,.FALSE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL H5Aexists_by_name_f(fid, GROUP1_NAME, attrname, exists, error, lapl_id = H5P_DEFAULT_F)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists_by_name_f",exists,.FALSE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5acreate_f(gid, attrname, H5T_NATIVE_INTEGER, sid, aid, error, H5P_DEFAULT_F, H5P_DEFAULT_F)
CALL check("h5acreate_f",error,total_error)
CALL H5Aexists_f(gid, attrname, exists, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists",exists,.TRUE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL H5Aexists_by_name_f(fid, GROUP1_NAME, attrname, exists, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists_by_name_f",exists,.TRUE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
attr_data1(1) = u
data_dims(1) = 1
CALL h5awrite_f(aid, H5T_NATIVE_INTEGER, attr_data1, data_dims, error)
CALL check("h5awrite_f",error,total_error)
CALL h5aclose_f(aid, error)
CALL check("h5aclose_f",error,total_error)
CALL H5Aexists_f(gid, attrname, exists, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists",exists,.TRUE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r14930] Maintenance: Fixed more bugs/typos and enabled tests that were failing previously on linew. Currently only one test (dense attributes) is failing. It looks like C library problem and we have a similar bug report in Bugzilla: when dense storage is used, attributes are not written to the file; somehow similar C test doesn't expose the problem while Fortran test does. Platforms tested: linew, kagiso with g95 and PGI
2008-05-05 09:00:37 +08:00
CALL H5Aexists_by_name_f(fid, GROUP1_NAME, attrname, exists, error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("H5Aexists_by_name_f",exists,.TRUE.,total_error )
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
ENDDO
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close group
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Gclose_f(gid, error)
CALL check("h5gclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close file
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5fclose_f(fid, error)
CALL check("h5fclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close dataspaces
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5sclose_f(sid, error)
CALL check("h5sclose_f",error,total_error)
END SUBROUTINE test_attr_many
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!-------------------------------------------------------------------------
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! * Function: attr_open_check
! *
! * Purpose: Check opening attribute on an object
! *
! * Return: Success: 0
! * Failure: -1
! *
[svn-r15025] Purpose: Cleaned up unused code (i.e. commented out) and removed extraneous output to standard out. Tested: N/A - No critical executable source lines were modified, only comments and write statements.
2008-05-17 04:46:10 +08:00
! * Programmer: Fortran version (M.S. Breitenfeld)
! * March 21, 2008
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
! *
! *-------------------------------------------------------------------------
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
!
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
SUBROUTINE attr_open_check(fid, dsetname, obj_id, max_attrs, total_error )
IMPLICIT NONE
INTEGER(HID_T), INTENT(IN) :: fid
CHARACTER(LEN=*), INTENT(IN) :: dsetname
INTEGER(HID_T), INTENT(IN) :: obj_id
INTEGER, INTENT(IN) :: max_attrs
INTEGER, INTENT(INOUT) :: total_error
INTEGER :: u
CHARACTER (LEN=8) :: attrname
INTEGER :: error
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
LOGICAL :: f_corder_valid ! Indicates whether the the creation order data is valid for this attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
INTEGER :: corder ! Is a positive integer containing the creation order of the attribute
INTEGER :: cset ! Indicates the character set used for the attribute’s name
INTEGER(HSIZE_T) :: data_size ! indicates the size, in the number of characters
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
INTEGER(HSIZE_T) :: storage_size ! attributes storage requirements
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CHARACTER(LEN=2) :: chr2
INTEGER(HID_T) attr_id
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open each attribute on object by index and check that it's the correct one
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
DO u = 0, max_attrs-1
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
WRITE(chr2,'(I2.2)') u
attrname = 'attr '//chr2
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aopen_f(obj_id, attrname, attr_id, error)
CALL check("h5aopen_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Get the attribute's information
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_f(attr_id, f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_f",error,total_error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check that the object's attributes are correct
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f.corder",corder,u,total_error)
CALL verify("h5aget_info_f.corder_valid",f_corder_valid,.TRUE.,total_error)
CALL verify("h5aget_info_f.cset", cset, H5T_CSET_ASCII_F, total_error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
CALL h5aget_storage_size_f(attr_id, storage_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r18197] Description: Trim trailing whitespace from source code files with this command: find . \( -name "*.[ch]" -or -name "*.cpp" -or -name "*.f90" \) -print |xargs -n 1 sed -i "" 's/[[:blank:]]*$//' Tested on: None - eyeballed only
2010-01-30 12:29:13 +08:00
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f.data_size", INT(data_size), INT(storage_size), total_error)
[svn-r14987] Purpose: Fixed the validation problem with H5Aget_info_f Desciption: Recast the interger type passed into 'validate' subroutine.
2008-05-14 00:10:19 +08:00
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr_id, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_by_name_f(obj_id, ".", attrname, attr_id, error, lapl_id=H5P_DEFAULT_F, aapl_id=H5P_DEFAULT_F)
CALL check("H5Aopen_by_name_f", error, total_error)
CALL h5aget_info_f(attr_id, f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check the attribute's information
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f",corder,u,total_error)
CALL verify("h5aget_info_f",f_corder_valid,.TRUE.,total_error)
CALL verify("h5aget_info_f", cset, H5T_CSET_ASCII_F, total_error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
CALL h5aget_storage_size_f(attr_id, storage_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f", INT(data_size), INT(storage_size), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr_id, error)
CALL check("h5aclose_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Open the attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL H5Aopen_by_name_f(fid, dsetname, attrname, attr_id, error)
CALL check("H5Aopen_by_name_f", error, total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Get the attribute's information
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aget_info_f(attr_id, f_corder_valid, corder, cset, data_size, error)
CALL check("h5aget_info_f",error,total_error)
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Check the attribute's information
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f",corder,u,total_error)
CALL verify("h5aget_info_f",f_corder_valid,.TRUE.,total_error)
CALL verify("h5aget_info_f", cset, H5T_CSET_ASCII_F, total_error)
[svn-r14979] Purpose: Added additional verification test for H5Aget_info_f
2008-05-13 04:24:53 +08:00
CALL h5aget_storage_size_f(attr_id, storage_size, error)
CALL check("h5aget_storage_size_f",error,total_error)
[svn-r27493] Trying again to merge the F2003_v1.10 branch to the trunk. Tested: h5committest --PASSED--
2015-08-11 22:35:30 +08:00
CALL verify("h5aget_info_f", INT(data_size), INT(storage_size), total_error)
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
[svn-r26479] Fix for: HDFFV-9162 h5pget_class_f should return hid_t for classtype Also removed all left over C comment artifacts /* and */ in the Fortran comments. Tested: h5committest
2015-03-20 01:11:47 +08:00
! Close attribute
[svn-r14923] Maintenance: This check-in merges changes from the fortran_1_8 branch back into the trunk (up to revision 14921) Platforms tested: kagiso with g95 and Intel compilers; more testing will be done after checking in a fresh copy from the trunk. New code itself was tested with all Fortran compilers available at THG
2008-05-04 07:39:37 +08:00
CALL h5aclose_f(attr_id, error)
CALL check("h5aclose_f",error,total_error)
ENDDO
END SUBROUTINE attr_open_check
[svn-r24967] Maintenance: Reorganized and cleaned the code to remove compiler warnings in the Fortran test code and examples. Platforms tested: Manual testing in place and using srcdir on jam, platypus, and emu with default and PGI, Intel and new GNU compilers. ifort compiler was also tested with -i8 and -r8 flags on jam. CMake tested on jam.
2014-04-06 23:56:21 +08:00
END MODULE TH5A_1_8
Reference in New Issue Copy Permalink