mirrors/postgres
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Rename built-in Snowball stemmer dictionaries to be english_stem,

Browse Source 
russian_stem, etc.  Per discussion.
This commit is contained in:
Tom Lane 2007-08-25 01:06:25 +00:00
parent 7351b5fa17
commit 93eab9312f
5 changed files with 20 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
src/backend/snowball/Makefile
View File

@ -2,7 +2,7 @@
#
# Makefile for src/backend/snowball
#
# $PostgreSQL: pgsql/src/backend/snowball/Makefile,v 1.1 2007/08/21 01:11:15 tgl Exp $
# $PostgreSQL: pgsql/src/backend/snowball/Makefile,v 1.2 2007/08/25 01:06:24 tgl Exp $
#
#-------------------------------------------------------------------------
@ -47,6 +47,8 @@ OBJS= dict_snowball.o api.o utilities.o \
stem_UTF_8_turkish.o
# second column is name of latin dictionary, if different
# Note order dependency: use of some other language as latin dictionary
# must come after creation of that language
LANGUAGES= \
danish danish \
dutch dutch \
@ -93,18 +95,19 @@ ifeq ($(enable_shared), yes)
while [ "$$#" -gt 0 ] ; \
do \
lang=$$1; shift; \
nonlatdictname=$$lang; \
latdictname=$$1; shift; \
if [ -s $(srcdir)/stopwords/$${lang}.stop ] ; then \
stop=", StopWords=$${lang}" ; \
else \
stop=""; \
fi; \
nonlatdictname=$$lang; \
latdictname=$$1; shift; \
cat $(srcdir)/snowball.sql.in | \
sed -e "s#_DICTNAME_#$$lang#g" | \
sed -e "s#_LANGNAME_#$$lang#g" | \
sed -e "s#_DICTNAME_#$${lang}_stem#g" | \
sed -e "s#_CFGNAME_#$$lang#g" | \
sed -e "s#_LATDICTNAME_#$$latdictname#g" | \
sed -e "s#_NONLATDICTNAME_#$$nonlatdictname#g" | \
sed -e "s#_LATDICTNAME_#$${latdictname}_stem#g" | \
sed -e "s#_NONLATDICTNAME_#$${nonlatdictname}_stem#g" | \
sed -e "s#_STOPWORDS_#$$stop#g" ; \
done >> $@
else

11
src/backend/snowball/snowball.sql.in
View File

@ -1,16 +1,15 @@
-- $PostgreSQL: pgsql/src/backend/snowball/snowball.sql.in,v 1.2 2007/08/22 01:39:44 tgl Exp $$
-- $PostgreSQL: pgsql/src/backend/snowball/snowball.sql.in,v 1.3 2007/08/25 01:06:25 tgl Exp $$
-- text search configuration for _CFGNAME_ language
-- text search configuration for _LANGNAME_ language
CREATE TEXT SEARCH DICTIONARY _DICTNAME_
(TEMPLATE = snowball,
Language = _DICTNAME_ _STOPWORDS_);
(TEMPLATE = snowball, Language = _LANGNAME_ _STOPWORDS_);
COMMENT ON TEXT SEARCH DICTIONARY _DICTNAME_ IS 'Snowball stemmer for _DICTNAME_ language';
COMMENT ON TEXT SEARCH DICTIONARY _DICTNAME_ IS 'Snowball stemmer for _LANGNAME_ language';
CREATE TEXT SEARCH CONFIGURATION _CFGNAME_
(PARSER = default);
COMMENT ON TEXT SEARCH CONFIGURATION _CFGNAME_ IS 'Configuration for _CFGNAME_ language';
COMMENT ON TEXT SEARCH CONFIGURATION _CFGNAME_ IS 'Configuration for _LANGNAME_ language';
ALTER TEXT SEARCH CONFIGURATION _CFGNAME_ ADD MAPPING
FOR email, url, host, sfloat, version, uri, file, float, int, uint

4
src/include/catalog/catversion.h
View File

@ -37,7 +37,7 @@
* Portions Copyright (c) 1996-2007, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $PostgreSQL: pgsql/src/include/catalog/catversion.h,v 1.417 2007/08/22 01:39:45 tgl Exp $
* $PostgreSQL: pgsql/src/include/catalog/catversion.h,v 1.418 2007/08/25 01:06:25 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -53,6 +53,6 @@
*/
/* yyyymmddN */
#define CATALOG_VERSION_NO 200708211
#define CATALOG_VERSION_NO 200708241
#endif

4
src/test/regress/expected/tsearch.out
View File

@ -195,13 +195,13 @@ SELECT * FROM ts_stat('SELECT a FROM test_tsvector', 'AB') ORDER BY ndoc DESC, n
(1 row)
--dictionaries and to_tsvector
SELECT ts_lexize('english', 'skies');
SELECT ts_lexize('english_stem', 'skies');
ts_lexize
-----------
{sky}
(1 row)
SELECT ts_lexize('english', 'identity');
SELECT ts_lexize('english_stem', 'identity');
ts_lexize
-----------
{ident}

4
src/test/regress/sql/tsearch.sql
View File

@ -81,8 +81,8 @@ SELECT * FROM ts_stat('SELECT a FROM test_tsvector', 'AB') ORDER BY ndoc DESC, n
--dictionaries and to_tsvector
SELECT ts_lexize('english', 'skies');
SELECT ts_lexize('english', 'identity');
SELECT ts_lexize('english_stem', 'skies');
SELECT ts_lexize('english_stem', 'identity');
SELECT * FROM ts_token_type('default');