Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commit1fe15d2

Browse files
committed
Fix conversion of SIMILAR TO regexes for character classes
The code that translates SIMILAR TO pattern matching expressions toPOSIX-style regular expressions did not consider that square bracketscan be nested. For example, in an expression like [[:alpha:]%_], thelogic replaced the placeholders '_' and '%' but it should not.This commit fixes the conversion logic by tracking the nesting level ofsquare brackets marking character class areas, while considering thatin expressions like []] or [^]] the first closing square bracket is aregular character. Multiple tests are added to show how the conversionsshould or should not apply applied while in a character class area, withspecific cases added for all the characters converted outside characterclasses like an opening parenthesis '(', dollar sign '$', etc.Author: Laurenz Albe <laurenz.albe@cybertec.at>Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>Reviewed-by: Michael Paquier <michael@paquier.xyz>Discussion:https://postgr.es/m/16ab039d1af455652bdf4173402ddda145f2c73b.camel@cybertec.atBackpatch-through: 13
1 parent1412781 commit1fe15d2

File tree

3 files changed

+114
-6
lines changed

3 files changed

+114
-6
lines changed

‎src/backend/utils/adt/regexp.c

Lines changed: 32 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -673,8 +673,11 @@ similar_escape_internal(text *pat_text, text *esc_text)
673673
intplen,
674674
elen;
675675
boolafterescape= false;
676-
boolincharclass= false;
677676
intnquotes=0;
677+
intcharclass_depth=0;/* Nesting level of character classes,
678+
* encompassed by square brackets */
679+
intcharclass_start=0;/* State of the character class start,
680+
* for carets */
678681

679682
p=VARDATA_ANY(pat_text);
680683
plen=VARSIZE_ANY_EXHDR(pat_text);
@@ -804,7 +807,7 @@ similar_escape_internal(text *pat_text, text *esc_text)
804807
/* fast path */
805808
if (afterescape)
806809
{
807-
if (pchar=='"'&&!incharclass)/* escape-double-quote? */
810+
if (pchar=='"'&&charclass_depth<1)/* escape-double-quote? */
808811
{
809812
/* emit appropriate part separator, per notes above */
810813
if (nquotes==0)
@@ -853,18 +856,41 @@ similar_escape_internal(text *pat_text, text *esc_text)
853856
/* SQL escape character; do not send to output */
854857
afterescape= true;
855858
}
856-
elseif (incharclass)
859+
elseif (charclass_depth>0)
857860
{
858861
if (pchar=='\\')
859862
*r++='\\';
860863
*r++=pchar;
861-
if (pchar==']')
862-
incharclass= false;
864+
865+
/*
866+
* Ignore a closing bracket at the start of a character class.
867+
* Such a bracket is taken literally rather than closing the
868+
* class. "charclass_start" is 1 right at the beginning of a
869+
* class and 2 after an initial caret.
870+
*/
871+
if (pchar==']'&&charclass_start>2)
872+
charclass_depth--;
873+
elseif (pchar=='[')
874+
charclass_depth++;
875+
876+
/*
877+
* If there is a caret right after the opening bracket, it negates
878+
* the character class, but a following closing bracket should
879+
* still be treated as a normal character. That holds only for
880+
* the first caret, so only the values 1 and 2 mean that closing
881+
* brackets should be taken literally.
882+
*/
883+
if (pchar=='^')
884+
charclass_start++;
885+
else
886+
charclass_start=3;/* definitely past the start */
863887
}
864888
elseif (pchar=='[')
865889
{
890+
/* start of a character class */
866891
*r++=pchar;
867-
incharclass= true;
892+
charclass_depth++;
893+
charclass_start=1;
868894
}
869895
elseif (pchar=='%')
870896
{

‎src/test/regress/expected/strings.out

Lines changed: 62 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -564,6 +564,68 @@ SELECT 'abcdefg' SIMILAR TO '_bcd%' ESCAPE NULL AS null;
564564
SELECT 'abcdefg' SIMILAR TO '_bcd#%' ESCAPE '##' AS error;
565565
ERROR: invalid escape string
566566
HINT: Escape string must be empty or one character.
567+
-- Characters that should be left alone in character classes when a
568+
-- SIMILAR TO regexp pattern is converted to POSIX style.
569+
-- Underscore "_"
570+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '_[_[:alpha:]_]_';
571+
QUERY PLAN
572+
------------------------------------------------
573+
Seq Scan on text_tbl
574+
Filter: (f1 ~ '^(?:.[_[:alpha:]_].)$'::text)
575+
(2 rows)
576+
577+
-- Percentage "%"
578+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '%[%[:alnum:]%]%';
579+
QUERY PLAN
580+
--------------------------------------------------
581+
Seq Scan on text_tbl
582+
Filter: (f1 ~ '^(?:.*[%[:alnum:]%].*)$'::text)
583+
(2 rows)
584+
585+
-- Dot "."
586+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '.[.[:alnum:].].';
587+
QUERY PLAN
588+
--------------------------------------------------
589+
Seq Scan on text_tbl
590+
Filter: (f1 ~ '^(?:\.[.[:alnum:].]\.)$'::text)
591+
(2 rows)
592+
593+
-- Dollar "$"
594+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '$[$[:alnum:]$]$';
595+
QUERY PLAN
596+
--------------------------------------------------
597+
Seq Scan on text_tbl
598+
Filter: (f1 ~ '^(?:\$[$[:alnum:]$]\$)$'::text)
599+
(2 rows)
600+
601+
-- Opening parenthesis "("
602+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '([([:alnum:](](';
603+
ERROR: invalid regular expression: parentheses () not balanced
604+
-- Caret "^"
605+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '^[^[:alnum:]^[^^][[^^]][\^][[\^]]\^]^';
606+
QUERY PLAN
607+
------------------------------------------------------------------------
608+
Seq Scan on text_tbl
609+
Filter: (f1 ~ '^(?:\^[^[:alnum:]^[^^][[^^]][\^][[\^]]\^]\^)$'::text)
610+
(2 rows)
611+
612+
-- Closing square bracket "]" at the beginning of character class
613+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '[]%][^]%][^%]%';
614+
QUERY PLAN
615+
------------------------------------------------
616+
Seq Scan on text_tbl
617+
Filter: (f1 ~ '^(?:[]%][^]%][^%].*)$'::text)
618+
(2 rows)
619+
620+
-- Closing square bracket effective after two carets at the beginning
621+
-- of character class.
622+
EXPLAIN (COSTS OFF) SELECT * FROM TEXT_TBL WHERE f1 SIMILAR TO '[^^]^';
623+
QUERY PLAN
624+
---------------------------------------
625+
Seq Scan on text_tbl
626+
Filter: (f1 ~ '^(?:[^^]\^)$'::text)
627+
(2 rows)
628+
567629
-- Test back reference in regexp_replace
568630
SELECT regexp_replace('1112223333', E'(\\d{3})(\\d{3})(\\d{4})', E'(\\1) \\2-\\3');
569631
regexp_replace

‎src/test/regress/sql/strings.sql

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -185,6 +185,26 @@ SELECT 'abcd\efg' SIMILAR TO '_bcd\%' ESCAPE '' AS true;
185185
SELECT'abcdefg' SIMILAR TO'_bcd%' ESCAPENULLASnull;
186186
SELECT'abcdefg' SIMILAR TO'_bcd#%' ESCAPE'##'AS error;
187187

188+
-- Characters that should be left alone in character classes when a
189+
-- SIMILAR TO regexp pattern is converted to POSIX style.
190+
-- Underscore "_"
191+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'_[_[:alpha:]_]_';
192+
-- Percentage "%"
193+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'%[%[:alnum:]%]%';
194+
-- Dot "."
195+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'.[.[:alnum:].].';
196+
-- Dollar "$"
197+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'$[$[:alnum:]$]$';
198+
-- Opening parenthesis "("
199+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'([([:alnum:](](';
200+
-- Caret "^"
201+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'^[^[:alnum:]^[^^][[^^]][\^][[\^]]\^]^';
202+
-- Closing square bracket "]" at the beginning of character class
203+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'[]%][^]%][^%]%';
204+
-- Closing square bracket effective after two carets at the beginning
205+
-- of character class.
206+
EXPLAIN (COSTS OFF)SELECT*FROM TEXT_TBLWHERE f1 SIMILAR TO'[^^]^';
207+
188208
-- Test back reference in regexp_replace
189209
SELECT regexp_replace('1112223333', E'(\\d{3})(\\d{3})(\\d{4})', E'(\\1)\\2-\\3');
190210
SELECT regexp_replace('AAA BBB CCC', E'\\s+','','g');

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp