From 5c86bfd3e83b2bde7706746c59ac148b1553e717 Mon Sep 17 00:00:00 2001 From: David Geier Date: Thu, 23 Apr 2026 11:08:33 +0200 Subject: [PATCH v1 3/3] Use correct collation for comparison --- contrib/pg_trgm/Makefile | 6 +-- contrib/pg_trgm/meson.build | 1 + contrib/pg_trgm/pg_trgm--1.6--1.7.sql | 20 +++++++++ contrib/pg_trgm/pg_trgm.control | 2 +- contrib/pg_trgm/trgm_gin.c | 65 +++++++++++++++++++++++++++ 5 files changed, 90 insertions(+), 4 deletions(-) create mode 100644 contrib/pg_trgm/pg_trgm--1.6--1.7.sql diff --git a/contrib/pg_trgm/Makefile b/contrib/pg_trgm/Makefile index 26b3028b75e..556b76f49f2 100644 --- a/contrib/pg_trgm/Makefile +++ b/contrib/pg_trgm/Makefile @@ -9,9 +9,9 @@ OBJS = \ trgm_regexp.o EXTENSION = pg_trgm -DATA = pg_trgm--1.5--1.6.sql pg_trgm--1.4--1.5.sql pg_trgm--1.3--1.4.sql \ - pg_trgm--1.3.sql pg_trgm--1.2--1.3.sql pg_trgm--1.1--1.2.sql \ - pg_trgm--1.0--1.1.sql +DATA = pg_trgm--1.6--1.7.sql pg_trgm--1.5--1.6.sql pg_trgm--1.4--1.5.sql \ + pg_trgm--1.3--1.4.sql pg_trgm--1.3.sql pg_trgm--1.2--1.3.sql \ + pg_trgm--1.1--1.2.sql pg_trgm--1.0--1.1.sql PGFILEDESC = "pg_trgm - trigram matching" REGRESS = pg_trgm pg_utf8_trgm pg_word_trgm pg_strict_word_trgm pg_trgm_collation diff --git a/contrib/pg_trgm/meson.build b/contrib/pg_trgm/meson.build index 5eafa774435..9fdf8c2d07e 100644 --- a/contrib/pg_trgm/meson.build +++ b/contrib/pg_trgm/meson.build @@ -28,6 +28,7 @@ install_data( 'pg_trgm--1.3.sql', 'pg_trgm--1.4--1.5.sql', 'pg_trgm--1.5--1.6.sql', + 'pg_trgm--1.6--1.7.sql', 'pg_trgm.control', kwargs: contrib_data_args, ) diff --git a/contrib/pg_trgm/pg_trgm--1.6--1.7.sql b/contrib/pg_trgm/pg_trgm--1.6--1.7.sql new file mode 100644 index 00000000000..dc5552e375e --- /dev/null +++ b/contrib/pg_trgm/pg_trgm--1.6--1.7.sql @@ -0,0 +1,20 @@ +/* contrib/pg_trgm/pg_trgm--1.6--1.7.sql */ + +-- complain if script is sourced in psql, rather than via ALTER EXTENSION +\echo Use "ALTER EXTENSION pg_trgm UPDATE TO '1.7'" to load this file. \quit + +-- Create collation-aware comparison function for trigrams +CREATE FUNCTION gin_compare_value_trgm(int4, int4) +RETURNS int4 +AS 'MODULE_PATHNAME' +LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE; + +-- Replace btint4cmp with gin_compare_value_trgm in the operator family +-- This ensures trigram comparisons respect collation settings +-- First drop the old function, then add the new one +ALTER OPERATOR FAMILY gin_trgm_ops USING gin DROP + FUNCTION 1 (text, text); + +ALTER OPERATOR FAMILY gin_trgm_ops USING gin ADD + FUNCTION 1 (text, text) gin_compare_value_trgm (int4, int4); + diff --git a/contrib/pg_trgm/pg_trgm.control b/contrib/pg_trgm/pg_trgm.control index 1d6a9ddf259..6e3ee43c510 100644 --- a/contrib/pg_trgm/pg_trgm.control +++ b/contrib/pg_trgm/pg_trgm.control @@ -1,6 +1,6 @@ # pg_trgm extension comment = 'text similarity measurement and index searching based on trigrams' -default_version = '1.6' +default_version = '1.7' module_pathname = '$libdir/pg_trgm' relocatable = true trusted = true diff --git a/contrib/pg_trgm/trgm_gin.c b/contrib/pg_trgm/trgm_gin.c index 14a892c657d..dadbc789349 100644 --- a/contrib/pg_trgm/trgm_gin.c +++ b/contrib/pg_trgm/trgm_gin.c @@ -5,8 +5,11 @@ #include "access/gin.h" #include "access/stratnum.h" +#include "catalog/pg_collation.h" +#include "catalog/pg_type.h" #include "fmgr.h" #include "trgm.h" +#include "utils/pg_locale.h" #include "varatt.h" PG_FUNCTION_INFO_V1(gin_extract_trgm); @@ -14,6 +17,7 @@ PG_FUNCTION_INFO_V1(gin_extract_value_trgm); PG_FUNCTION_INFO_V1(gin_extract_query_trgm); PG_FUNCTION_INFO_V1(gin_trgm_consistent); PG_FUNCTION_INFO_V1(gin_trgm_triconsistent); +PG_FUNCTION_INFO_V1(gin_compare_value_trgm); /* * This function can only be called if a pre-9.1 version of the GIN operator @@ -169,6 +173,67 @@ gin_extract_query_trgm(PG_FUNCTION_ARGS) PG_RETURN_POINTER(entries); } +/* + * Compare two trigram values for GIN index. + * This function considers the active collation when comparing trigrams, + * unlike btint4cmp which treats them as plain integers. + */ +Datum +gin_compare_value_trgm(PG_FUNCTION_ARGS) +{ + int32 a = PG_GETARG_INT32(0); + int32 b = PG_GETARG_INT32(1); + Oid collid = PG_GET_COLLATION(); + pg_locale_t locale = 0; + + /* + * If a non-default collation is specified, we need to compare the + * trigrams character-by-character using the collation's rules. + */ + if (collid != DEFAULT_COLLATION_OID) + locale = pg_newlocale_from_collation(collid); + + if (locale && locale->collate_is_c) + locale = 0; /* C collation can use simple comparison */ + + if (locale && locale->collate) + { + /* + * For non-C collations, extract the three bytes from each trigram + * and compare them using the collation's comparison function. + */ + char str_a[3]; + char str_b[3]; + int result; + + /* Extract bytes from the packed integer representation */ + str_a[0] = (a >> 16) & 0xFF; + str_a[1] = (a >> 8) & 0xFF; + str_a[2] = a & 0xFF; + + str_b[0] = (b >> 16) & 0xFF; + str_b[1] = (b >> 8) & 0xFF; + str_b[2] = b & 0xFF; + + /* Use collation-aware comparison */ + result = pg_strncoll(str_a, 3, str_b, 3, locale); + PG_RETURN_INT32(result); + } + else + { + /* + * For C collation or default collation, simple integer comparison + * is sufficient and faster. + */ + if (a < b) + PG_RETURN_INT32(-1); + else if (a > b) + PG_RETURN_INT32(1); + else + PG_RETURN_INT32(0); + } +} + Datum gin_trgm_consistent(PG_FUNCTION_ARGS) { -- 2.51.0