postgresql/contrib/ltree

contrib/ltree module

ltree - is a PostgreSQL contrib module which contains implementation of data
types, indexed access methods and queries for data organized as a tree-like
structures.
This module will works for PostgreSQL version 7.3.
(version for 7.2 version is available from http://www.sai.msu.su/~megera/postgres/gist/ltree/ltree-7.2.tar.gz)
-------------------------------------------------------------------------------
All work was done by Teodor Sigaev (teodor@stack.net) and Oleg Bartunov
(oleg@sai.msu.su). See http://www.sai.msu.su/~megera/postgres/gist for
additional information. Authors would like to thank Eugeny Rodichev for helpful
discussions. Comments and bug reports are welcome.
-------------------------------------------------------------------------------

LEGAL NOTICES: This module is released under BSD license (as PostgreSQL
itself). This work was done in framework of Russian Scientific Network and
partially supported by Russian Foundation for Basic Research and Stack Group.
-------------------------------------------------------------------------------

MOTIVATION

This is a placeholder for introduction to the problem. Hope, people reading
this document doesn't need it too much :-)

DEFINITIONS

A label of a node is a sequence of one or more words separated by blank
character '_' and containing letters and digits ( for example, [a-zA-Z0-9] for
C locale). The length of a label is limited by 256 bytes.

Example: 'Countries', 'Personal_Services'

A label path of a node is a sequence of one or more dot-separated labels
l1.l2...ln, represents path from root to the node. The length of a label path
is limited by 65Kb, but size <= 2Kb is preferrable. We consider it's not a
strict limitation ( maximal size of label path for DMOZ catalogue - http://
www.dmoz.org, is about 240 bytes !)

Example: 'Top.Countries.Europe.Russia'

We introduce several datatypes:

ltree
    - is a datatype for label path.
   
ltree[]
    - is a datatype for arrays of ltree.
   
lquery
    - is a path expression that has regular expression in the label path and
    used for ltree matching. Star symbol (*) is used to specify any number of
    labels (levels) and could be used at the beginning and the end of lquery,
    for example, '*.Europe.*'.
   
    The following quantifiers are recognized for '*' (like in Perl):
   
               {n}    Match exactly n levels
               {n,}   Match at least n levels
               {n,m}  Match at least n but not more than m levels
               {,m}   Match at maximum m levels (eq. to {0,m})
   
    It is possible to use several modifiers at the end of a label:
   
      
               @     Do case-insensitive label matching
               *     Do prefix matching for a label
               %     Don't account word separator '_' in label matching, that is 
                     'Russian%' would match 'Russian_nations', but not 'Russian'
   
    lquery could contains logical '!' (NOT) at the beginning of the label and '
    |' (OR) to specify possible alternatives for label matching.
   
    Example of lquery:
   
    
                Top.*{0,2}.sport*@.!football|tennis.Russ*|Spain
                a)  b)     c)      d)               e)
    
    A label path should
      + a) begins from a node with label 'Top'
      + b) and following zero or 2 labels until
      + c) a node with label beginning from case-insensitive prefix 'sport'
      + d) following node with label not matched 'football' or 'tennis' and
      + e) ends on node with label beginning from 'Russ' or strictly matched
        'Spain'.
   
ltxtquery
    - is a datatype for label searching (like type 'query' for full text
    searching, see contrib/tsearch). It's possible to use modifiers @,%,* at
    the end of word. The meaning of modifiers are the same as for lquery.
   
    Example: 'Europe & Russia*@ & !Transportation'
   
    Search paths contain words 'Europe' and 'Russia*' (case-insensitive) and
    not 'Transportation'. Notice, the order of words as they appear in label
    path is not important !

OPERATIONS

The following operations are defined for type ltree:

<,>,<=,>=,=, <>
    - have their usual meanings. Comparison is doing in the order of direct
    tree traversing, children of a node are sorted lexicographic.
ltree @> ltree
    - returns TRUE if left argument is an ancestor of right argument (or
    equal).
ltree <@ ltree
    - returns TRUE if left argument is a descendant of right argument (or
    equal).
ltree ~ lquery, lquery ~ ltree
    - return TRUE if node represented by ltree satisfies lquery.
ltree @ ltxtquery, ltxtquery @ ltree
    - return TRUE if node represented by ltree satisfies ltxtquery.
ltree || ltree, ltree || text, text || ltree
    - return concatenated ltree.

Operations for arrays of ltree (ltree[]):

ltree[] @> ltree, ltree <@ ltree[]
    - returns TRUE if array ltree[] contains an ancestor of ltree.
ltree @> ltree[], ltree[] <@ ltree
    - returns TRUE if array ltree[] contains a descendant of ltree.
ltree[] ~ lquery, lquery ~ ltree[]
    - returns TRUE if array ltree[] contains label paths matched lquery.
ltree[] @ ltxtquery, ltxtquery @ ltree[]
    - returns TRUE if array ltree[] contains label paths matched ltxtquery
    (full text search).
ltree[] ?@> ltree, ltree ?<@ ltree[], ltree[] ?~ lquery, ltree[] ?@ ltxtquery
    - returns first element of array ltree[] satisfies corresponding condition
    and NULL in vice versa.

REMARK

Operations <@, @>, @ and ~ have analogues - ^<@, ^@>, ^@, ^~, which doesn't use
indices !

INDICES

Various indices could be created to speed up execution of operations:

  * B-tree index over ltree:
    <, <=, =, =>, >
  * GiST index over ltree:
    <, <=, =, =>, >, @>, <@, @, ~
    Example:
    create index path_gist_idx on test using gist (path);
  * GiST index over ltree[]:
    ltree[]<@ ltree, ltree @> ltree[], @, ~.
    Example:
    create index path_gist_idx on test using gist (array_path);
    Notices: This index is lossy.

FUNCTIONS

ltree subltree
      ltree subltree(ltree, start, end) 
        returns subpath of ltree from start (inclusive) until the end.
        # select subltree('Top.Child1.Child2',1,2);
        subltree
        --------
        Child1
ltree subpath
      ltree subpath(ltree, OFFSET,LEN)
      ltree subpath(ltree, OFFSET)
        returns subpath of ltree from OFFSET (inclusive) with length LEN.
        If OFFSET is negative returns subpath starts that far from the end 
        of the path.  If LENGTH is omitted, returns everything to the end
        of the path.  If LENGTH is negative, leaves that many labels off 
        the end of the path.
        # select subpath('Top.Child1.Child2',1,2);
        subpath
        -------
        Child1.Child2
    
        # select subpath('Top.Child1.Child2',-2,1);
        subpath 
        ---------
        Child1
int4 nlevel
      
       int4 nlevel(ltree) - returns level of the node.
        # select nlevel('Top.Child1.Child2');
        nlevel 
        --------
          3
    Note, that arguments start, end, OFFSET, LEN have meaning of level of the
    node !
   
ltree lca(ltree,ltree,...) (up to 8 arguments)
    ltree lca(ltree[])
    Returns Lowest Common Ancestor (lca)
         # select lca('1.2.2.3','1.2.3.4.5.6');
         lca 
         -----
          1.2
         # select lca('{la.2.3,1.2.3.4.5.6}') is null;
         ?column? 
         ----------
            f
         

INSTALLATION

  cd contrib/ltree
  make
  make install
  make installcheck

EXAMPLE OF USAGE

 createdb ltreetest
 psql ltreetest < /usr/local/pgsql/share/contrib/ltree.sql
 psql ltreetest < ltreetest.sql

Now, we have a database ltreetest populated with a data describing hierarchy
shown below:

 
                            TOP
                         /   |  \     
                 Science Hobbies Collections  
                     /       |              \
            Astronomy   Amateurs_Astronomy Pictures
               /  \                            |
    Astrophysics  Cosmology                Astronomy
                                            /  |    \
                                     Galaxies Stars Astronauts

Inheritance:

ltreetest=# select path from test where path <@ 'Top.Science';
                path                
------------------------------------
 Top.Science
 Top.Science.Astronomy
 Top.Science.Astronomy.Astrophysics
 Top.Science.Astronomy.Cosmology
(4 rows)

Matching:

ltreetest=# select path from test where path ~ '*.Astronomy.*';
                     path                      
-----------------------------------------------
 Top.Science.Astronomy
 Top.Science.Astronomy.Astrophysics
 Top.Science.Astronomy.Cosmology
 Top.Collections.Pictures.Astronomy
 Top.Collections.Pictures.Astronomy.Stars
 Top.Collections.Pictures.Astronomy.Galaxies
 Top.Collections.Pictures.Astronomy.Astronauts
(7 rows)
ltreetest=# select path from test where path ~ '*.!pictures@.*.Astronomy.*';
                path                
------------------------------------
 Top.Science.Astronomy
 Top.Science.Astronomy.Astrophysics
 Top.Science.Astronomy.Cosmology
(3 rows)

Full text search:

ltreetest=# select path from test where path @ 'Astro*% & !pictures@';
                path                
------------------------------------
 Top.Science.Astronomy
 Top.Science.Astronomy.Astrophysics
 Top.Science.Astronomy.Cosmology
 Top.Hobbies.Amateurs_Astronomy
(4 rows)

ltreetest=# select path from test where path @ 'Astro* & !pictures@';
                path                
------------------------------------
 Top.Science.Astronomy
 Top.Science.Astronomy.Astrophysics
 Top.Science.Astronomy.Cosmology
(3 rows)

Using Functions:

ltreetest=# select subpath(path,0,2)||'Space'||subpath(path,2) from test where path <@ 'Top.Science.Astronomy';
                 ?column?                 
------------------------------------------
 Top.Science.Space.Astronomy
 Top.Science.Space.Astronomy.Astrophysics
 Top.Science.Space.Astronomy.Cosmology
(3 rows)
We could create SQL-function:
CREATE FUNCTION ins_label(ltree, int4, text) RETURNS ltree 
AS 'select subpath($1,0,$2) || $3 || subpath($1,$2);'
LANGUAGE SQL WITH (ISCACHABLE);

and previous select could be rewritten as:

ltreetest=# select ins_label(path,2,'Space') from test where path <@ 'Top.Science.Astronomy';
                ins_label                 
------------------------------------------
 Top.Science.Space.Astronomy
 Top.Science.Space.Astronomy.Astrophysics
 Top.Science.Space.Astronomy.Cosmology
(3 rows)

Or with another arguments:

CREATE FUNCTION ins_label(ltree, ltree, text) RETURNS ltree
AS 'select subpath($1,0,nlevel($2)) || $3 || subpath($1,nlevel($2));'
LANGUAGE SQL WITH (ISCACHABLE);

ltreetest=# select ins_label(path,'Top.Science'::ltree,'Space') from test where path <@ 'Top.Science.Astronomy';
                ins_label                 
------------------------------------------
 Top.Science.Space.Astronomy
 Top.Science.Space.Astronomy.Astrophysics
 Top.Science.Space.Astronomy.Cosmology
(3 rows)

ADDITIONAL DATA

To get more feeling from our ltree module you could download
dmozltree-eng.sql.gz (about 3Mb tar.gz archive containing 300,274 nodes),
available from http://www.sai.msu.su/~megera/postgres/gist/ltree/
dmozltree-eng.sql.gz, which is DMOZ catalogue, prepared for use with ltree.
Setup your test database (dmoz), load ltree module and issue command:

zcat dmozltree-eng.sql.gz| psql dmoz

Data will be loaded into database dmoz and all indices will be created.

BENCHMARKS

All runs were performed on my IBM ThinkPad T21 (256 MB RAM, 750Mhz) using DMOZ
data, containing 300,274 nodes (see above for download link). We used some
basic queries typical for walking through catalog.

QUERIES

  * Q0: Count all rows (sort of base time for comparison)
    select count(*) from dmoz;
     count  
    --------
     300274
    (1 row)
  * Q1: Get direct children (without inheritance)
    select path from dmoz where path ~ 'Top.Adult.Arts.Animation.*{1}';
                   path                
    -----------------------------------
     Top.Adult.Arts.Animation.Cartoons
     Top.Adult.Arts.Animation.Anime
    (2 rows)
  * Q2: The same as Q1 but with counting of successors
    select path as parentpath , (select count(*)-1 from dmoz where path <@
    p.path) as count from dmoz p where path ~ 'Top.Adult.Arts.Animation.*{1}';
                parentpath             | count 
    -----------------------------------+-------
     Top.Adult.Arts.Animation.Cartoons |     2
     Top.Adult.Arts.Animation.Anime    |    61
    (2 rows)
  * Q3: Get all parents
    select path from dmoz where path @> 'Top.Adult.Arts.Animation' order by
    path asc;
               path           
    --------------------------
     Top
     Top.Adult
     Top.Adult.Arts
     Top.Adult.Arts.Animation
    (4 rows)
  * Q4: Get all parents with counting of children
    select path, (select count(*)-1 from dmoz where path <@ p.path) as count
    from dmoz p where path @> 'Top.Adult.Arts.Animation' order by path asc;
               path           | count  
    --------------------------+--------
     Top                      | 300273
     Top.Adult                |   4913
     Top.Adult.Arts           |    339
     Top.Adult.Arts.Animation |     65
    (4 rows)
  * Q5: Get all children with levels
    select path, nlevel(path) - nlevel('Top.Adult.Arts.Animation') as level
    from dmoz where path ~ 'Top.Adult.Arts.Animation.*{1,2}' order by path asc;
                          path                      | level 
    ------------------------------------------------+-------
     Top.Adult.Arts.Animation.Anime                 |     1
     Top.Adult.Arts.Animation.Anime.Fan_Works       |     2
     Top.Adult.Arts.Animation.Anime.Games           |     2
     Top.Adult.Arts.Animation.Anime.Genres          |     2
     Top.Adult.Arts.Animation.Anime.Image_Galleries |     2
     Top.Adult.Arts.Animation.Anime.Multimedia      |     2
     Top.Adult.Arts.Animation.Anime.Resources       |     2
     Top.Adult.Arts.Animation.Anime.Titles          |     2
     Top.Adult.Arts.Animation.Cartoons              |     1
     Top.Adult.Arts.Animation.Cartoons.AVS          |     2
     Top.Adult.Arts.Animation.Cartoons.Members      |     2
    (11 rows)

Timings

+---------------------------------------------+
|Query|Rows|Time (ms) index|Time (ms) no index|
|-----+----+---------------+------------------|
|   Q0|   1|             NA|           1453.44|
|-----+----+---------------+------------------|
|   Q1|   2|           0.49|           1001.54|
|-----+----+---------------+------------------|
|   Q2|   2|           1.48|           3009.39|
|-----+----+---------------+------------------|
|   Q3|   4|           0.55|            906.98|
|-----+----+---------------+------------------|
|   Q4|   4|       24385.07|           4951.91|
|-----+----+---------------+------------------|
|   Q5|  11|           0.85|           1003.23|
+---------------------------------------------+
Timings without indices were obtained using operations which doesn't use
indices (see above)

Remarks

We didn't run full-scale tests, also we didn't present (yet) data for
operations with arrays of ltree (ltree[]) and full text searching. We'll
appreciate your input. So far, below some (rather obvious) results:

  * Indices does help execution of queries
  * Q4 performs bad because one needs to read almost all data from the HDD

CHANGES

Aug 9, 2002
   Fixed very stupid but important bug :-)
July 31, 2002
   Now works on 64-bit platforms.
   Added function lca - lowest common ancestor
   Version for 7.2 is distributed as separate package - 
   http://www.sai.msu.su/~megera/postgres/gist/ltree/ltree-7.2.tar.gz
July 13, 2002
   Initial release.

TODO

  * Testing on 64-bit platforms. There are several known problems with byte
    alignment; -- RESOLVED
  * Better documentation;
  * We plan (probably) to improve regular expressions processing using
    non-deterministic automata;
  * Some sort of XML support;
  * Better full text searching;

SOME BACKGROUNDS

The approach we use for ltree is much like one we used in our other GiST based
contrib modules (intarray, tsearch, tree, btree_gist, rtree_gist). Theoretical
background is available in papers referenced from our GiST development page
(http://www.sai.msu.su/~megera/postgres/gist).

A hierarchical data structure (tree) is a set of nodes. Each node has a
signature (LPS) of a fixed size, which is a hashed label path of that node.
Traversing a tree we could *certainly* prune branches if

LQS (bitwise AND) LPS != LQS

where LQS is a signature of lquery or ltxtquery, obtained in the same way as
LPS.

ltree[]:
For array of ltree LPS is a bitwise OR-ed signatures of *ALL* children
reachable from that node. Signatures are stored in RD-tree, implemented using
GiST, which provides indexed access.

ltree:
For ltree we store LPS in a B-tree, implemented using GiST. Each node entry is
represented by (left_bound, signature, right_bound), so that we could speedup
operations <, <=, =, =>, > using left_bound, right_bound and prune branches of
a tree using signature.
-------------------------------------------------------------------------------
We ask people who find the module useful to send us a postcards to:
Moscow, 119899, Universitetski pr.13, Moscow State University, Sternberg
Astronomical Institute, Russia
For: Bartunov O.S.
and
Moscow, Bratislavskaya str.23, appt. 18, Russia
For: Sigaev F.G.