netcdf-c/libdispatch/ncs3sdk_h5.c

1032 lines
30 KiB
C
Raw Normal View History

Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
/*
* Copyright 2018, University Corporation for Atmospheric Research
* See netcdf/COPYRIGHT file for copying and redistribution conditions.
*/
#define NOOP
#undef DEBUG
#include "config.h"
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#include "netcdf.h"
#include "nclog.h"
#include "ncrc.h"
#include "ncxml.h"
#include "ncs3sdk.h"
#include "nch5s3comms.h"
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
#define NCTRACING
#ifdef NCTRACING
#define NCTRACE(level,fmt,...) nctrace((level),__func__,fmt,##__VA_ARGS__)
#define NCTRACEMORE(level,fmt,...) nctracemore((level),fmt,##__VA_ARGS__)
#define NCUNTRACE(e) ncuntrace(__func__,NCTHROW(e),NULL)
#define NCUNTRACEX(e,fmt,...) ncuntrace(__func__,NCTHROW(e),fmt,##__VA_ARGS__)
#define NCNILTRACE(e) ((void)NCUNTRACE(e))
#else
#define NCTRACE(level,fmt,...)
#define NCTRACEMORE(level,fmt,...)
#define NCUNTRACE(e) (e)
#define NCNILTRACE(e)
#define NCUNTRACEX(e,fmt,...) (e)
#endif
#ifdef __CYGWIN__
extern char* strdup(const char*);
#endif
#ifndef SHA256_DIGEST_LENGTH
#define SHA256_DIGEST_LENGTH 32
#endif
/* Mnemonic */
#define RECLAIM 1
#define size64 unsigned long long
typedef struct NCS3CLIENT {
char* rooturl; /* The URL (minus any fragment) for the dataset root path (excludes bucket on down) */
s3r_t* h5s3client; /* From h5s3comms */
} NCS3CLIENT;
struct Object {
NClist* checksumalgorithms; /* NClist<char*> */
char* etag;
char* key;
char* lastmodified;
struct Owner {
char* displayname;
char* id;
} owner;
char* size;
char* storageclass;
};
typedef char* CommonPrefix;
/* Hold essential info from a ListObjectsV2 response */
struct LISTOBJECTSV2 {
char* istruncated;
NClist* contents; /* NClist<struct Object*> */
char* name;
char* prefix;
char* delimiter;
char* maxkeys;
NClist* commonprefixes; /* NClist<CommonPrefix> */
char* encodingtype;
char* keycount;
char* continuationtoken;
char* nextcontinuationtoken;
char* startafter;
};
/* Forward */
static void s3client_destroy(NCS3CLIENT* s3client);
static char* makes3rooturl(NCS3INFO* info);
static int makes3fullpath(const char* pathkey, const char* bucket, const char* prefix, const char* key, NCbytes* url);
static int parse_listbucketresult(char* xml, unsigned long long xmllen, struct LISTOBJECTSV2**);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
static int parse_object(ncxml_t root, NClist* objects);
static int parse_owner(ncxml_t root, struct Owner* ownerp);
static int parse_prefix(ncxml_t root, NClist* prefixes);
static int parse_checksumalgorithm(ncxml_t root, NClist* algorithms);
static struct LISTOBJECTSV2* alloclistobjectsv2(void);
static struct Object* allocobject(void);
static void reclaim_listobjectsv2(struct LISTOBJECTSV2* lo);
static void reclaim_object(struct Object* o);
static char* trim(char* s, int reclaim);
static int makes3prefix(const char* prefix, char** prefixdirp);
static int s3objectsinfo(NClist* contents, NClist* keys, NClist* lens);
static int s3commonprefixes(NClist* list, NClist* keys);
static int mergekeysets(NClist*,NClist*,NClist*);
static int rawtokeys(s3r_buf_t* response, NClist* keys, NClist* lengths, struct LISTOBJECTSV2** listv2p);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
static int queryadd(NClist* query, const char* key, const char* value);
static int queryend(NClist* query, char** querystring);
static int queryinsert(NClist* list, char* ekey, char* evalue);
2023-04-28 06:40:09 +08:00
#define NT(x) ((x)==NULL?"null":x)
2023-04-29 04:30:48 +08:00
2023-04-29 05:28:32 +08:00
#if 0
2023-04-28 06:40:09 +08:00
static void
2023-04-28 07:02:36 +08:00
dumps3info(NCS3INFO* s3info, const char* tag)
2023-04-28 06:40:09 +08:00
{
2023-04-28 07:02:36 +08:00
if(tag == NULL) tag = "dumps3info";
fprintf(stderr,">>> %s: s3info=%p\n",tag,s3info);
2023-04-28 06:40:09 +08:00
if(s3info != NULL) {
2023-04-28 07:02:36 +08:00
fprintf(stderr,">>> %s: s3info->host=%s\n",tag,NT(s3info->host));
fprintf(stderr,">>> %s: s3info->region=%s\n",tag,NT(s3info->region));
fprintf(stderr,">>> %s: s3info->bucket=%s\n",tag,NT(s3info->bucket));
fprintf(stderr,">>> %s: s3info->rootkey=%s\n",tag,NT(s3info->rootkey));
fprintf(stderr,">>> %s: s3info->profile=%s\n",tag,NT(s3info->profile));
2023-04-28 06:40:09 +08:00
}
}
2023-04-29 04:30:48 +08:00
static void
dumps3client(void* s3client0, const char* tag)
{
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
if(tag == NULL) tag = "dumps3client";
fprintf(stderr,">>> %s: s3client=%p\n",tag,s3client);
if(s3client != NULL) {
fprintf(stderr,">>> %s: s3client->rooturl=%s\n",tag,NT(s3client->rooturl));
fprintf(stderr,">>> %s: s3client->h5s3client=%p\n",tag,s3client->rooturl);
}
}
2023-04-29 05:11:42 +08:00
#endif
2023-04-29 04:30:48 +08:00
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
/**************************************************/
EXTERNL void*
NC_s3sdkcreateclient(NCS3INFO* info)
{
int stat = NC_NOERR;
const char* accessid = NULL;
const char* accesskey = NULL;
char* urlroot = NULL;
NCS3CLIENT* s3client = NULL;
NCTRACE(11,"info=%s",NC_s3dumps3info(info));
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
s3client = (NCS3CLIENT*)calloc(1,sizeof(NCS3CLIENT));
if(s3client == NULL) goto done;
if(info->profile != NULL) {
if((stat = NC_s3profilelookup(info->profile, "aws_access_key_id", &accessid))) goto done;
if((stat = NC_s3profilelookup(info->profile, "aws_secret_access_key", &accesskey))) goto done;
}
if((s3client->rooturl = makes3rooturl(info))==NULL) {stat = NC_ENOMEM; goto done;}
s3client->h5s3client = NCH5_s3comms_s3r_open(s3client->rooturl,info->svc,info->region,accessid,accesskey);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if(s3client->h5s3client == NULL) {stat = NC_ES3; goto done;}
done:
nullfree(urlroot);
if(stat && s3client) {
2023-04-29 04:52:20 +08:00
NC_s3sdkclose(s3client,info,0,NULL);
2023-04-29 05:26:21 +08:00
s3client = NULL;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
}
NCNILTRACE(NC_NOERR);
return (void*)s3client;
}
EXTERNL int
NC_s3sdkbucketexists(void* s3client0, const char* bucket, int* existsp, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCbytes* url = ncbytesnew();
long httpcode = 0;
NCTRACE(11,"bucket=%s",bucket);
if(errmsgp) *errmsgp = NULL;
if((stat = makes3fullpath(s3client->rooturl,bucket,NULL,NULL,url))) goto done;
if((stat = NCH5_s3comms_s3r_head(s3client->h5s3client, ncbytescontents(url), NULL, NULL, &httpcode, NULL))) goto done;
if(existsp) {*existsp = (stat == 0 && httpcode == 200);}
done:
ncbytesfree(url);
return NCUNTRACEX(stat,"exists=%d",PTRVAL(int,existsp,-1));
}
EXTERNL int
NC_s3sdkbucketcreate(void* s3client0, const char* region, const char* bucket, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NC_UNUSED(s3client);
NCTRACE(11,"region=%s bucket=%s",region,bucket);
if(errmsgp) *errmsgp = NULL;
fprintf(stderr,"create bucket: %s\n",bucket); fflush(stderr);
return NCUNTRACE(stat);
}
EXTERNL int
NC_s3sdkbucketdelete(void* s3client0, NCS3INFO* info, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NC_UNUSED(s3client);
NCTRACE(11,"info=%s%s",NC_s3dumps3info(info));
if(errmsgp) *errmsgp = NULL;
fprintf(stderr,"delete bucket: %s\n",info->bucket); fflush(stderr);
return NCUNTRACE(stat);
}
/**************************************************/
/* Object API */
/*
@return NC_NOERR if key points to a content-bearing object.
@return NC_EEMPTY if object at key has no content.
@return NC_EXXX return true error
*/
EXTERNL int
NC_s3sdkinfo(void* s3client0, const char* bucket, const char* pathkey, size64_t* lenp, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCbytes* url = ncbytesnew();
long long len = -1;
NCTRACE(11,"bucket=%s pathkey=%s",bucket,pathkey);
if((stat = makes3fullpath(s3client->rooturl,bucket,pathkey,NULL,url))) goto done;
if((stat = NCH5_s3comms_s3r_getsize(s3client->h5s3client, ncbytescontents(url), &len))) goto done;
if(lenp) {*lenp = len;}
done:
ncbytesfree(url);
return NCUNTRACEX(stat,"len=%d",PTRVAL(int,lenp,-1));
}
/*
@return NC_NOERR if success
@return NC_EXXX if fail
*/
EXTERNL int
NC_s3sdkread(void* s3client0, const char* bucket, const char* pathkey, size64_t start, size64_t count, void* content, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCbytes* url = ncbytesnew();
struct s3r_buf_t data = {0,NULL};
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
NCTRACE(11,"bucket=%s pathkey=%s start=%llu count=%llu content=%p",bucket,pathkey,start,count,content);
if((stat = makes3fullpath(s3client->rooturl,bucket,pathkey,NULL,url))) goto done;
/* Read the data */
data.count = count;
data.content = content;
if((stat = NCH5_s3comms_s3r_read(s3client->h5s3client,ncbytescontents(url),(size_t)start,(size_t)count,&data))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
done:
ncbytesfree(url);
return NCUNTRACE(stat);
}
/*
For S3, I can see no way to do a byterange write;
so we are effectively writing the whole object
*/
EXTERNL int
NC_s3sdkwriteobject(void* s3client0, const char* bucket, const char* pathkey, size64_t count, const void* content, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCbytes* url = ncbytesnew();
s3r_buf_t data;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
NCTRACE(11,"bucket=%s pathkey=%s count=%llu content=%p",bucket,pathkey,count,content);
if((stat = makes3fullpath(s3client->rooturl,bucket,pathkey,NULL,url))) goto done;
/* Write the data */
data.count = count;
data.content = (void*)content;
if((stat = NCH5_s3comms_s3r_write(s3client->h5s3client,ncbytescontents(url),&data))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
done:
ncbytesfree(url);
return NCUNTRACE(stat);
}
EXTERNL int
NC_s3sdkclose(void* s3client0, NCS3INFO* info, int deleteit, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCTRACE(11,"info=%s deleteit=%d",NC_s3dumps3info(info),deleteit);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if(deleteit) {
/* Delete the root key; ok it if does not exist */
switch (stat = NC_s3sdkdeletekey(s3client0,info->bucket,info->rootkey,errmsgp)) {
case NC_NOERR: break;
case NC_EEMPTY: case NC_ENOTFOUND: stat = NC_NOERR; break;
default: break;
}
}
s3client_destroy(s3client);
return NCUNTRACE(stat);
}
/*
Common code for getkeys and searchkeys.
Return a list of names of legal objects immediately below a specified key.
In theory, the returned list should be sorted in lexical order,
but it possible that it is not.
*/
static int
getkeys(void* s3client0, const char* bucket, const char* prefixkey0, const char* delim, size_t* nkeysp, char*** keysp, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
char* prefixdir = NULL;
NClist* query = NULL;
char* querystring = NULL;
NCURI* purl = NULL;
NCbytes* listurl = ncbytesnew();
NClist* allkeys = nclistnew();
struct LISTOBJECTSV2* listv2 = NULL;
int istruncated = 0;
char* continuetoken = NULL;
s3r_buf_t response = {0,NULL};
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
NCTRACE(11,"bucket=%s prefixkey0=%s",bucket,prefixkey0);
/* cleanup the prefix */
if((stat = makes3prefix(prefixkey0,&prefixdir))) return NCUNTRACE(stat);
do {
nclistfreeall(query);
query = nclistnew();
nullfree(querystring);
querystring = NULL;
ncbytesclear(listurl);
nullfree(response.content); response.content = NULL; response.count = 0;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
/* Make sure order is sorted (after encoding) */
if((stat = queryadd(query,"list-type","2"))) goto done;
if((stat = queryadd(query,"prefix",prefixdir))) goto done;
if(delim != NULL) {
if((stat = queryadd(query,"delimiter",delim))) goto done;
}
if(istruncated && continuetoken != NULL) {
if((stat = queryadd(query,"continuation-token",continuetoken))) goto done;
}
if((stat = queryend(query,&querystring))) goto done;
/* Build the proper url; leave off prefix because it will appear in the query */
if((stat = makes3fullpath(s3client->rooturl, bucket, NULL, NULL, listurl))) goto done;
/* Append the query string */
ncbytescat(listurl,"?");
ncbytescat(listurl,querystring);
if((stat = NCH5_s3comms_s3r_getkeys(s3client->h5s3client, ncbytescontents(listurl), &response))) goto done;
if((stat = rawtokeys(&response,allkeys,NULL,&listv2))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
istruncated = (strcasecmp(listv2->istruncated,"true")==0?1:0);
nullfree(continuetoken);
continuetoken = nulldup(listv2->nextcontinuationtoken);
reclaim_listobjectsv2(listv2); listv2 = NULL;
} while(istruncated);
if(nkeysp) {*nkeysp = nclistlength(allkeys);}
if(keysp) {*keysp = nclistextract(allkeys);}
done:
nullfree(continuetoken);
reclaim_listobjectsv2(listv2);
nclistfreeall(allkeys);
nclistfreeall(query);
nullfree(querystring);
ncurifree(purl);
ncbytesfree(listurl);
nullfree(response.content);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if(prefixdir) free(prefixdir);
return NCUNTRACEX(stat,"nkeys=%u",PTRVAL(unsigned,nkeysp,0));
}
/*
Return a list of names of legal objects immediately below a specified key.
In theory, the returned list should be sorted in lexical order,
but it possible that it is not.
*/
EXTERNL int
NC_s3sdkgetkeys(void* s3client0, const char* bucket, const char* prefixkey0, size_t* nkeysp, char*** keysp, char** errmsgp)
{
NCTRACE(11,"bucket=%s prefixkey0=%s",bucket,prefixkey0);
Mitigate S3 test interference + Unlimited Dimensions in NCZarr This PR started as an attempt to add unlimited dimensions to NCZarr. It did that, but this exposed significant problems with test interference. So this PR is mostly about fixing -- well mitigating anyway -- test interference. The problem of test interference is now documented in the document docs/internal.md. The solutions implemented here are also describe in that document. The solution is somewhat fragile but multiple cleanup mechanisms are provided. Note that this feature requires that the AWS command line utility must be installed. ## Unlimited Dimensions. The existing NCZarr extensions to Zarr are modified to support unlimited dimensions. NCzarr extends the Zarr meta-data for the ".zgroup" object to include netcdf-4 model extensions. This information is stored in ".zgroup" as dictionary named "_nczarr_group". Inside "_nczarr_group", there is a key named "dims" that stores information about netcdf-4 named dimensions. The value of "dims" is a dictionary whose keys are the named dimensions. The value associated with each dimension name has one of two forms Form 1 is a special case of form 2, and is kept for backward compatibility. Whenever a new file is written, it uses format 1 if possible, otherwise format 2. * Form 1: An integer representing the size of the dimension, which is used for simple named dimensions. * Form 2: A dictionary with the following keys and values" - "size" with an integer value representing the (current) size of the dimension. - "unlimited" with a value of either "1" or "0" to indicate if this dimension is an unlimited dimension. For Unlimited dimensions, the size is initially zero, and as variables extend the length of that dimension, the size value for the dimension increases. That dimension size is shared by all arrays referencing that dimension, so if one array extends an unlimited dimension, it is implicitly extended for all other arrays that reference that dimension. This is the standard semantics for unlimited dimensions. Adding unlimited dimensions required a number of other changes to the NCZarr code-base. These included the following. * Did a partial refactor of the slice handling code in zwalk.c to clean it up. * Added a number of tests for unlimited dimensions derived from the same test in nc_test4. * Added several NCZarr specific unlimited tests; more are needed. * Add test of endianness. ## Misc. Other Changes * Modify libdispatch/ncs3sdk_aws.cpp to optionally support use of the AWS Transfer Utility mechanism. This is controlled by the ```#define TRANSFER```` command in that file. It defaults to being disabled. * Parameterize both the standard Unidata S3 bucket (S3TESTBUCKET) and the netcdf-c test data prefix (S3TESTSUBTREE). * Fixed an obscure memory leak in ncdump. * Removed some obsolete unit testing code and test cases. * Uncovered a bug in the netcdf-c handling of big-endian floats and doubles. Have not fixed yet. See tst_h5_endians.c. * Renamed some nczarr_tests testcases to avoid name conflicts with nc_test4. * Modify the semantics of zmap\#ncsmap_write to only allow total rewrite of objects. * Modify the semantics of zodom to properly handle stride > 1. * Add a truncate operation to the libnczarr zmap code.
2023-09-27 06:56:48 +08:00
return NCUNTRACE(getkeys(s3client0, bucket, prefixkey0, "/", nkeysp, keysp, errmsgp));
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
}
/*
Return a list of full keys of legal objects immediately below a specified key.
Not necessarily sorted.
Essentially same as getkeys, but with no delimiter.
*/
EXTERNL int
NC_s3sdksearch(void* s3client0, const char* bucket, const char* prefixkey0, size_t* nkeysp, char*** keysp, char** errmsgp)
{
NCTRACE(11,"bucket=%s prefixkey0=%s",bucket,prefixkey0);
Mitigate S3 test interference + Unlimited Dimensions in NCZarr This PR started as an attempt to add unlimited dimensions to NCZarr. It did that, but this exposed significant problems with test interference. So this PR is mostly about fixing -- well mitigating anyway -- test interference. The problem of test interference is now documented in the document docs/internal.md. The solutions implemented here are also describe in that document. The solution is somewhat fragile but multiple cleanup mechanisms are provided. Note that this feature requires that the AWS command line utility must be installed. ## Unlimited Dimensions. The existing NCZarr extensions to Zarr are modified to support unlimited dimensions. NCzarr extends the Zarr meta-data for the ".zgroup" object to include netcdf-4 model extensions. This information is stored in ".zgroup" as dictionary named "_nczarr_group". Inside "_nczarr_group", there is a key named "dims" that stores information about netcdf-4 named dimensions. The value of "dims" is a dictionary whose keys are the named dimensions. The value associated with each dimension name has one of two forms Form 1 is a special case of form 2, and is kept for backward compatibility. Whenever a new file is written, it uses format 1 if possible, otherwise format 2. * Form 1: An integer representing the size of the dimension, which is used for simple named dimensions. * Form 2: A dictionary with the following keys and values" - "size" with an integer value representing the (current) size of the dimension. - "unlimited" with a value of either "1" or "0" to indicate if this dimension is an unlimited dimension. For Unlimited dimensions, the size is initially zero, and as variables extend the length of that dimension, the size value for the dimension increases. That dimension size is shared by all arrays referencing that dimension, so if one array extends an unlimited dimension, it is implicitly extended for all other arrays that reference that dimension. This is the standard semantics for unlimited dimensions. Adding unlimited dimensions required a number of other changes to the NCZarr code-base. These included the following. * Did a partial refactor of the slice handling code in zwalk.c to clean it up. * Added a number of tests for unlimited dimensions derived from the same test in nc_test4. * Added several NCZarr specific unlimited tests; more are needed. * Add test of endianness. ## Misc. Other Changes * Modify libdispatch/ncs3sdk_aws.cpp to optionally support use of the AWS Transfer Utility mechanism. This is controlled by the ```#define TRANSFER```` command in that file. It defaults to being disabled. * Parameterize both the standard Unidata S3 bucket (S3TESTBUCKET) and the netcdf-c test data prefix (S3TESTSUBTREE). * Fixed an obscure memory leak in ncdump. * Removed some obsolete unit testing code and test cases. * Uncovered a bug in the netcdf-c handling of big-endian floats and doubles. Have not fixed yet. See tst_h5_endians.c. * Renamed some nczarr_tests testcases to avoid name conflicts with nc_test4. * Modify the semantics of zmap\#ncsmap_write to only allow total rewrite of objects. * Modify the semantics of zodom to properly handle stride > 1. * Add a truncate operation to the libnczarr zmap code.
2023-09-27 06:56:48 +08:00
return NCUNTRACE(getkeys(s3client0, bucket, prefixkey0, NULL, nkeysp, keysp, errmsgp));
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
}
EXTERNL int
NC_s3sdkdeletekey(void* s3client0, const char* bucket, const char* pathkey, char** errmsgp)
{
int stat = NC_NOERR;
NCS3CLIENT* s3client = (NCS3CLIENT*)s3client0;
NCbytes* url = ncbytesnew();
long httpcode = 0;
NCTRACE(11,"s3client0=%p bucket=%s pathkey=%s",s3client0,bucket,pathkey);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if((stat = makes3fullpath(s3client->rooturl,bucket,pathkey,NULL,url))) goto done;
if((stat = NCH5_s3comms_s3r_deletekey(s3client->h5s3client, ncbytescontents(url), &httpcode))) goto done;
done:
ncbytesfree(url);
return NCUNTRACE(stat);
}
/**************************************************/
/* Utilities */
/*
Convert raw getkeys response to vector of keys
*/
static int
rawtokeys(s3r_buf_t* response, NClist* allkeys, NClist* lengths, struct LISTOBJECTSV2** listv2p)
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
{
int stat = NC_NOERR;
struct LISTOBJECTSV2* listv2 = NULL;
NClist* realkeys = nclistnew();
NClist* commonkeys = nclistnew();
if((stat = parse_listbucketresult(response->content,response->count,&listv2))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if(nclistlength(listv2->contents) > 0) {
if((stat = s3objectsinfo(listv2->contents,realkeys,lengths))) goto done;
}
/* Add common prefixes */
if(nclistlength(listv2->commonprefixes) > 0) {
if((stat = s3commonprefixes(listv2->commonprefixes,commonkeys))) goto done;
}
if((stat=mergekeysets(realkeys, commonkeys, allkeys))) goto done;
if(listv2p) {*listv2p = listv2; listv2 = NULL;}
done:
nclistfreeall(realkeys);
nclistfreeall(commonkeys);
reclaim_listobjectsv2(listv2);
return stat;
}
/* Create a path-format root url */
static char*
makes3rooturl(NCS3INFO* info)
{
NCbytes* buf = ncbytesnew();
char* result = NULL;
ncbytescat(buf,"https://");
ncbytescat(buf,info->host);
result = ncbytesextract(buf);
ncbytesfree(buf);
return result;
}
static int
makes3fullpath(const char* rooturl, const char* bucket, const char* prefix, const char* key, NCbytes* url)
{
int stat = NC_NOERR;
2023-04-28 05:12:26 +08:00
assert(url != NULL);
assert(rooturl != NULL);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
ncbytescat(url,rooturl);
if(bucket) {
if(ncbyteslength(url) > 0 && ncbytesget(url,ncbyteslength(url)-1) != '/') ncbytescat(url,"/");
if(*bucket == '/') bucket++;
ncbytescat(url,bucket);
}
if(prefix) {
if(ncbyteslength(url) > 0 && ncbytesget(url,ncbyteslength(url)-1) != '/') ncbytescat(url,"/");
if(*prefix == '/') prefix++;
ncbytescat(url,prefix);
}
if(key) {
if(ncbyteslength(url) > 0 && ncbytesget(url,ncbyteslength(url)-1) != '/') ncbytescat(url,"/");
if(*key == '/') key++;
ncbytescat(url,key);
}
/* Remove any trailing '/' */
if(ncbyteslength(url) > 0 && ncbytesget(url,ncbyteslength(url)-1) == '/') ncbytessetlength(url,ncbyteslength(url)-1);
return stat;
}
/* (1) Ensure trailing '/' and (2) Strip off leading '/' */
/* Note that order of '/' handling is important; we want
"/" to be converted to "", but "/a/b/c" to be converted to "a/b/c/"
*/
static int
makes3prefix(const char* prefix, char** prefixdirp)
{
size_t plen;
char* prefixdir;
if(prefix == NULL) prefix = "";
if(*prefix == '/') prefix++; /* Remove any leading '/' */
plen = strlen(prefix);
prefixdir = (char*)malloc(plen+1+1); /* possible '/' and nul */
if(prefixdir == NULL) return NC_ENOMEM;
memcpy(prefixdir,prefix,plen+1); /* include trailing nul */
if(plen > 0)
if(prefixdir[plen-1] != '/') strlcat(prefixdir,"/",plen+1+1); /* Ensure trailing '/' */
/* else leave prefix as "" */
*prefixdirp = prefixdir; prefixdir = NULL;
return NC_NOERR;
}
Mitigate S3 test interference + Unlimited Dimensions in NCZarr This PR started as an attempt to add unlimited dimensions to NCZarr. It did that, but this exposed significant problems with test interference. So this PR is mostly about fixing -- well mitigating anyway -- test interference. The problem of test interference is now documented in the document docs/internal.md. The solutions implemented here are also describe in that document. The solution is somewhat fragile but multiple cleanup mechanisms are provided. Note that this feature requires that the AWS command line utility must be installed. ## Unlimited Dimensions. The existing NCZarr extensions to Zarr are modified to support unlimited dimensions. NCzarr extends the Zarr meta-data for the ".zgroup" object to include netcdf-4 model extensions. This information is stored in ".zgroup" as dictionary named "_nczarr_group". Inside "_nczarr_group", there is a key named "dims" that stores information about netcdf-4 named dimensions. The value of "dims" is a dictionary whose keys are the named dimensions. The value associated with each dimension name has one of two forms Form 1 is a special case of form 2, and is kept for backward compatibility. Whenever a new file is written, it uses format 1 if possible, otherwise format 2. * Form 1: An integer representing the size of the dimension, which is used for simple named dimensions. * Form 2: A dictionary with the following keys and values" - "size" with an integer value representing the (current) size of the dimension. - "unlimited" with a value of either "1" or "0" to indicate if this dimension is an unlimited dimension. For Unlimited dimensions, the size is initially zero, and as variables extend the length of that dimension, the size value for the dimension increases. That dimension size is shared by all arrays referencing that dimension, so if one array extends an unlimited dimension, it is implicitly extended for all other arrays that reference that dimension. This is the standard semantics for unlimited dimensions. Adding unlimited dimensions required a number of other changes to the NCZarr code-base. These included the following. * Did a partial refactor of the slice handling code in zwalk.c to clean it up. * Added a number of tests for unlimited dimensions derived from the same test in nc_test4. * Added several NCZarr specific unlimited tests; more are needed. * Add test of endianness. ## Misc. Other Changes * Modify libdispatch/ncs3sdk_aws.cpp to optionally support use of the AWS Transfer Utility mechanism. This is controlled by the ```#define TRANSFER```` command in that file. It defaults to being disabled. * Parameterize both the standard Unidata S3 bucket (S3TESTBUCKET) and the netcdf-c test data prefix (S3TESTSUBTREE). * Fixed an obscure memory leak in ncdump. * Removed some obsolete unit testing code and test cases. * Uncovered a bug in the netcdf-c handling of big-endian floats and doubles. Have not fixed yet. See tst_h5_endians.c. * Renamed some nczarr_tests testcases to avoid name conflicts with nc_test4. * Modify the semantics of zmap\#ncsmap_write to only allow total rewrite of objects. * Modify the semantics of zodom to properly handle stride > 1. * Add a truncate operation to the libnczarr zmap code.
2023-09-27 06:56:48 +08:00
/* Move keys1 concat keys2 into merge; note that merge list may not be empty. */
/* Will leave keys1 and keys2 empty */
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
static int
mergekeysets(NClist* keys1, NClist* keys2, NClist* merge)
{
int stat = NC_NOERR;
int i;
size_t nkeys1 = nclistlength(keys1);
size_t nkeys2 = nclistlength(keys2);
for(i=0;i<nkeys1;i++) nclistpush(merge,nclistremove(keys1,0));
for(i=0;i<nkeys2;i++) nclistpush(merge,nclistremove(keys2,0));
nclistnull(merge);
return NCTHROW(stat);
}
static void
s3client_destroy(NCS3CLIENT* s3client)
{
if(s3client) {
nullfree(s3client->rooturl);
(void)NCH5_s3comms_s3r_close(s3client->h5s3client);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
free(s3client);
}
}
/**************************************************/
/* XML Response Parser(s) */
/**
Action: List objects (V2)
Response XML:
=========================
HTTP/1.1 200
<?xml version="1.0" encoding="UTF-8"?>
<ListBucketResult>
<Name>string</Name>
<Prefix>string</Prefix>
<KeyCount>integer</KeyCount>
<MaxKeys>integer</MaxKeys>
<Delimiter>string</Delimiter>
<IsTruncated>boolean</IsTruncated>
<Contents>
<Key>string</Key>
<LastModified>timestamp</LastModified>
<ETag>string</ETag>
<Size>integer</Size>
<StorageClass>string</StorageClass>
<ChecksumAlgorithm>string</ChecksumAlgorithm>
<Owner>
<DisplayName>string</DisplayName>
<ID>string</ID>
</Owner>
#ifdef GOOGLES3
<Generation>string</Generation>
<MetaGeneration>string</MetaGeneration>
#endif
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
...
</Contents>
...
<CommonPrefixes>
<Prefix>string</Prefix>
</CommonPrefixes>
...
<EncodingType>string</EncodingType>
<ContinuationToken>string</ContinuationToken>
<NextContinuationToken>string</NextContinuationToken>
<StartAfter>string</StartAfter>
</ListBucketResult>
=========================
*/
static int
parse_listbucketresult(char* xml, unsigned long long xmllen, struct LISTOBJECTSV2** resultp)
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
{
int stat = NC_NOERR;
ncxml_doc_t doc = NULL;
ncxml_t x;
struct LISTOBJECTSV2* result = NULL;
#ifdef DEBUG
printf("-------------------------%s\n-------------------------\n",xml);
#endif
doc = ncxml_parse(xml,xmllen);
if(doc == NULL) {stat = NC_ES3; goto done;}
ncxml_t dom = ncxml_root(doc);
/* Verify top level element */
if(strcmp(ncxml_name(dom),"ListBucketResult")!=0) {
nclog(NCLOGERR,"Expected: <ListBucketResult> actual: <%s>",ncxml_name(dom));
stat = NC_ES3;
goto done;
}
if((result = alloclistobjectsv2())==NULL) {stat = NC_ENOMEM; goto done;}
/* Iterate next-level elements */
for(x=ncxml_child_first(dom);x != NULL;x=ncxml_child_next(x)) {
const char* elem = ncxml_name(x);
if(strcmp(elem,"IsTruncated")==0) {
result->istruncated = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Contents")==0) {
if((stat = parse_object(x,result->contents))) goto done;
} else if(strcmp(elem,"Name")==0) {
result->name = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Prefix")==0) {
result->prefix = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Delimiter")==0) {
result->delimiter = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"MaxKeys")==0) {
result->maxkeys = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"CommonPrefixes")==0) {
if((stat = parse_prefix(x,result->commonprefixes))) goto done;
} else if(strcmp(elem,"EncodingType")==0) {
result->encodingtype = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"KeyCount")==0) {
result->keycount = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"ContinuationToken")==0) {
result->continuationtoken = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"NextContinuationToken")==0) {
result->nextcontinuationtoken = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"StartAfter")==0) {
result->startafter = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"StartAfter")==0) {
result->startafter = trim(ncxml_text(x),RECLAIM);
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
} else {
nclog(NCLOGERR,"Unexpected Element: <%s>",elem);
stat = NC_ES3;
goto done;
}
}
if(resultp) {*resultp = result; result = NULL;}
done:
if(result) reclaim_listobjectsv2(result);
if(doc) ncxml_free(doc);
return NCTHROW(stat);
}
static int
parse_object(ncxml_t root, NClist* objects)
{
int stat = NC_NOERR;
ncxml_t x;
struct Object* object = NULL;
/* Verify top level element */
if(strcmp(ncxml_name(root),"Contents")!=0) {
nclog(NCLOGERR,"Expected: <Contents> actual: <%s>",ncxml_name(root));
stat = NC_ES3;
goto done;
}
if((object = allocobject())==NULL) {stat = NC_ENOMEM; goto done;}
for(x=ncxml_child_first(root);x != NULL;x=ncxml_child_next(x)) {
const char* elem = ncxml_name(x);
if(strcmp(elem,"ChecksumAlgorithm")==0) {
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
if((stat = parse_checksumalgorithm(x,object->checksumalgorithms))) goto done;
} else if(strcmp(elem,"ETag")==0) {
object->etag = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Key")==0) {
object->key = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"LastModified")==0) {
object->lastmodified = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Owner")==0) {
if((stat = parse_owner(x,&object->owner))) goto done;
} else if(strcmp(elem,"Size")==0) {
object->size = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"StorageClass")==0) {
object->storageclass = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"Generation")==0) {
/* Ignore */
} else if(strcmp(elem,"MetaGeneration")==0) {
/* Ignore */
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
} else {
nclog(NCLOGERR,"Unexpected Element: <%s>",elem);
stat = NC_ES3;
goto done;
}
}
nclistpush(objects,object);
object = NULL;
done:
if(object) reclaim_object(object);
return NCTHROW(stat);
}
static int
parse_owner(ncxml_t root, struct Owner* owner)
{
int stat = NC_NOERR;
ncxml_t x;
/* Verify top level element */
if(strcmp(ncxml_name(root),"Owner")!=0) {
nclog(NCLOGERR,"Expected: <Owner> actual: <%s>",ncxml_name(root));
stat = NC_ES3;
goto done;
}
for(x=ncxml_child_first(root);x != NULL;x=ncxml_child_next(x)) {
const char* elem = ncxml_name(x);
if(strcmp(elem,"DisplayName")==0) {
owner->displayname = trim(ncxml_text(x),RECLAIM);
} else if(strcmp(elem,"ID")==0) {
owner->id = trim(ncxml_text(x),RECLAIM);
} else {
nclog(NCLOGERR,"Unexpected Element: <%s>",elem);
stat = NC_ES3;
goto done;
}
}
done:
return NCTHROW(stat);
}
static int
parse_prefix(ncxml_t root, NClist* prefixes)
{
int stat = NC_NOERR;
ncxml_t x;
char* prefix = NULL;
/* Verify top level element */
if(strcmp(ncxml_name(root),"CommonPrefixes")!=0) {
nclog(NCLOGERR,"Expected: <CommonPrefixes> actual: <%s>",ncxml_name(root));
stat = NC_ES3;
goto done;
}
for(x=ncxml_child_first(root);x != NULL;x=ncxml_child_next(x)) {
const char* elem = ncxml_name(x);
if(strcmp(elem,"Prefix")==0) {
prefix = trim(ncxml_text(x),RECLAIM);
nclistpush(prefixes,prefix);
prefix = NULL;
} else {
nclog(NCLOGERR,"Unexpected Element: <%s>",elem);
stat = NC_ES3;
goto done;
}
}
done:
nullfree(prefix);
return NCTHROW(stat);
}
static int
parse_checksumalgorithm(ncxml_t root, NClist* algorithms)
{
int stat = NC_NOERR;
char* alg = NULL;
/* Verify top level element */
if(strcmp(ncxml_name(root),"ChecksumAlgorithm")!=0) {
nclog(NCLOGERR,"Expected: <ChecksumAlgorithm> actual: <%s>",ncxml_name(root));
stat = NC_ES3;
goto done;
}
alg = trim(ncxml_text(root),RECLAIM);
nclistpush(algorithms,alg);
alg = NULL;
done:
nullfree(alg);
return NCTHROW(stat);
}
static struct LISTOBJECTSV2*
alloclistobjectsv2(void)
{
struct LISTOBJECTSV2* lov2 = NULL;
if((lov2 = calloc(1,sizeof(struct LISTOBJECTSV2))) == NULL)
return lov2;
lov2->contents = nclistnew();
lov2->commonprefixes = nclistnew();
return lov2;
}
static struct Object*
allocobject(void)
{
struct Object* obj = NULL;
if((obj = calloc(1,sizeof(struct Object))) == NULL)
return obj;
obj->checksumalgorithms = nclistnew();
return obj;
}
static void
reclaim_listobjectsv2(struct LISTOBJECTSV2* lo)
{
int i;
if(lo == NULL) return;
nullfree(lo->istruncated);
for(i=0;i<nclistlength(lo->contents);i++)
reclaim_object((struct Object*)nclistget(lo->contents,i));
nclistfree(lo->contents);
nullfree(lo->name);
nullfree(lo->prefix);
nullfree(lo->delimiter);
nullfree(lo->maxkeys);
nclistfreeall(lo->commonprefixes);
nullfree(lo->encodingtype);
nullfree(lo->keycount);
nullfree(lo->continuationtoken);
nullfree(lo->nextcontinuationtoken);
nullfree(lo->startafter);
free(lo);
}
static void
reclaim_object(struct Object* o)
{
if(o == NULL) return;
nclistfreeall(o->checksumalgorithms);
nullfree(o->etag);
nullfree(o->key);
nullfree(o->lastmodified);
nullfree(o->owner.displayname);
nullfree(o->owner.id);
nullfree(o->size);
nullfree(o->storageclass);
free(o);
}
static char*
trim(char* s, int reclaim)
{
ptrdiff_t first=0,last=0;
const char* p;
char* t = NULL;
size_t len;
for(p=s;*p;p++) {
if(*p > ' ') {first = (p - s); break;}
}
for(p=s+(strlen(s)-1);p >= s;p--) {
if(*p > ' ') {last = (p - s); break;}
}
len = (last - first) + 1;
if((t = (char*)malloc(len+1))==NULL) return t;
memcpy(t,s+first,len);
t[len] = '\0';
if(reclaim) nullfree(s);
return t;
}
/*
Get Info about a single object from a vector
*/
static int
s3objectinfo1(const struct Object* s3_object, char** fullkeyp, uintptr_t* lenp)
{
int stat = NC_NOERR;
const char* key = NULL;
char* tmp = NULL;
unsigned long long len;
assert(fullkeyp);
key = s3_object->key;
len = strlen(key);
if((tmp = (char*)malloc(len+1+1))==NULL) {stat = NC_ENOMEM; goto done;}
tmp[0] = '\0';
if(key[0] != '/') strlcat(tmp,"/",len+1+1);
strlcat(tmp,key,len+1+1);
sscanf(s3_object->size,"%llu",&len);
if(fullkeyp) {*fullkeyp = tmp; tmp = NULL;}
if(lenp) *lenp = (size64_t)len;
done:
if(tmp) free(tmp);
return NCTHROW(stat);
}
/*
Get Info about a vector of objects; Keys are fixed up to start with a '/'
*/
static int
s3objectsinfo(NClist* contents, NClist* keys, NClist* lengths)
{
int stat = NC_NOERR;
size_t i;
char* key = NULL;
uintptr_t length = 0;
for(i=0;i<nclistlength(contents);i++) {
struct Object* s3_object = (struct Object*)nclistget(contents,i);
if((stat = s3objectinfo1(s3_object,&key,&length))) goto done;
nclistpush(keys,key); key = NULL;
nclistpush(lengths,(void*)length);
}
nclistnull(keys);
nclistnull(lengths);
done:
nullfree(key); /* avoid mem leak */
return NCTHROW(stat);
}
/* Ensure each key starts with '/' */
static int
s3commonprefixes(NClist* list, NClist* keys)
{
int stat = NC_NOERR;
int i;
for (i=0;i<nclistlength(list);i++) {
char* p;
size_t len;
const char* prefix = (char*)nclistget(list,i);
len = strlen(prefix);
if((p = (char*) malloc(len+1+1))==NULL) /* for nul + leading '/' */
{stat = NC_ENOMEM; goto done;}
*p = '\0';
if(*prefix != '/') strlcat(p,"/",len+1+1);
strlcat(p,prefix,len+1+1);
nclistpush(keys,p); p = NULL;
}
done:
nclistnull(keys);
return NCTHROW(stat);
}
/* Add new key,value pair to a query; encode before insertion; keep in sorted order */
static int
queryadd(NClist* query, const char* key, const char* value)
{
int stat = NC_NOERR;
char* ekey = NULL;
char* evalue = NULL;
if(key == NULL) {stat = NC_EINVAL; goto done;}
if((stat = NCH5_s3comms_uriencode(&ekey, key, strlen(key), 1/*true*/, NULL))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
evalue = NULL;
if(value != NULL) {
if((stat = NCH5_s3comms_uriencode(&evalue, value, strlen(value), 1/*true*/, NULL))) goto done;
Improve S3 Documentation and Support ## Improvements to S3 Documentation * Create a new document *quickstart_paths.md* that give a summary of the legal path formats used by netcdf-c. This includes both file paths and URL paths. * Modify *nczarr.md* to remove most of the S3 related text. * Move the S3 text from *nczarr.md* to a new document *cloud.md*. * Add some S3-related text to the *byterange.md* document. Hopefully, this will make it easier for users to find the information they want. ## Rebuild NCZarr Testing In order to avoid problems with running make check in parallel, two changes were made: 1. The *nczarr_test* test system was rebuilt. Now, for each test. any generated files are kept in a test-specific directory, isolated from all other test executions. 2. Similarly, since the S3 test bucket is shared, any generated S3 objects are isolated using a test-specific key path. ## Other S3 Related Changes * Add code to ensure that files created on S3 are reclaimed at end of testing. * Used the bash "trap" command to ensure S3 cleanup even if the test fails. * Cleanup the S3 related configure.ac flag set since S3 is used in several places. So now one should use the option *--enable-s3* instead of *--enable-nczarr-s3*, although the latter is still kept as a deprecated alias for the former. * Get some of the github actions yml to work with S3; required fixing various test scripts adding a secret to access the Unidata S3 bucket. * Cleanup S3 portion of libnetcdf.settings.in and netcdf_meta.h.in and test_common.in. * Merge partial S3 support into dhttp.c. * Create an experimental s3 access library especially for use with Windows. It is enabled by using the options *--enable-s3-internal* (automake) or *-DENABLE_S3_INTERNAL=ON* (CMake). Also add a unit-test for it. * Move some definitions from ncrc.h to ncs3sdk.h ## Other Changes * Provide a default implementation of strlcpy and move this and similar defaults into *dmissing.c*.
2023-04-26 07:15:06 +08:00
}
/* Insert encoded key+value keeping sorted order */
if((stat = queryinsert(query, ekey, evalue))) goto done;
ekey = NULL;
evalue = NULL;
done:
nullfree(ekey);
nullfree(evalue);
return NCTHROW(stat);
}
static int
queryend(NClist* query, char** querystring)
{
int stat = NC_NOERR;
size_t i;
NCbytes* buf = ncbytesnew(); /* accumulate final query string */
/* build the final query string */
for(i=0;i<nclistlength(query);i+=2) {
const char* key = (const char*)nclistget(query,i);
const char* value = (const char*)nclistget(query,i+1);
if(ncbyteslength(buf) > 0) ncbytescat(buf,"&");
ncbytescat(buf,key);
ncbytescat(buf,"=");
if(value != NULL)
ncbytescat(buf,value);
}
if(querystring) {*querystring = ncbytesextract(buf);}
ncbytesfree(buf);
return NCTHROW(stat);
}
/* Insert encoded key+value keeping sorted order */
static int
queryinsert(NClist* list, char* ekey, char* evalue)
{
int pos,i,stat = NC_NOERR;
for(pos=-1,i=0;i<nclistlength(list);i+=2) {
const char* key = (const char*)nclistget(list,i);
int cmp = strcmp(key,ekey);
if(cmp == 0) {stat = NC_EINVAL; goto done;} /* duplicate keys */
if(cmp > 0) {pos = i; break;} /* key > ekey => insert ekey before key */
}
if(pos < 0) pos = nclistlength(list); /* insert at end; also works if |list|==0 */
nclistinsert(list,pos,evalue);
nclistinsert(list,pos,ekey);
done:
return NCTHROW(stat);
}
2023-04-28 06:40:09 +08:00