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

Commit78212f2

Browse files
committed
Convert tsqueryin and tsvectorin to report errors softly.
This is slightly tedious because the adjustments cascade througha couple of levels of subroutines, but it's not very hard.I chose to avoid changing function signatures more than absolutelynecessary, by passing the escontext pointer in existing structswhere possible.tsquery's nuisance NOTICEs about empty queries are suppressed insoft-error mode, since they're not errors and we surely don't wantthem to be shown to the user anyway. Maybe that whole behaviorshould be reconsidered.Discussion:https://postgr.es/m/3824377.1672076822@sss.pgh.pa.us
1 parenteb8312a commit78212f2

File tree

8 files changed

+196
-52
lines changed

8 files changed

+196
-52
lines changed

‎src/backend/tsearch/to_tsany.c‎

Lines changed: 8 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -594,7 +594,8 @@ to_tsquery_byid(PG_FUNCTION_ARGS)
594594
query=parse_tsquery(text_to_cstring(in),
595595
pushval_morph,
596596
PointerGetDatum(&data),
597-
0);
597+
0,
598+
NULL);
598599

599600
PG_RETURN_TSQUERY(query);
600601
}
@@ -630,7 +631,8 @@ plainto_tsquery_byid(PG_FUNCTION_ARGS)
630631
query=parse_tsquery(text_to_cstring(in),
631632
pushval_morph,
632633
PointerGetDatum(&data),
633-
P_TSQ_PLAIN);
634+
P_TSQ_PLAIN,
635+
NULL);
634636

635637
PG_RETURN_POINTER(query);
636638
}
@@ -667,7 +669,8 @@ phraseto_tsquery_byid(PG_FUNCTION_ARGS)
667669
query=parse_tsquery(text_to_cstring(in),
668670
pushval_morph,
669671
PointerGetDatum(&data),
670-
P_TSQ_PLAIN);
672+
P_TSQ_PLAIN,
673+
NULL);
671674

672675
PG_RETURN_TSQUERY(query);
673676
}
@@ -704,7 +707,8 @@ websearch_to_tsquery_byid(PG_FUNCTION_ARGS)
704707
query=parse_tsquery(text_to_cstring(in),
705708
pushval_morph,
706709
PointerGetDatum(&data),
707-
P_TSQ_WEB);
710+
P_TSQ_WEB,
711+
NULL);
708712

709713
PG_RETURN_TSQUERY(query);
710714
}

‎src/backend/utils/adt/tsquery.c‎

Lines changed: 76 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -16,6 +16,7 @@
1616

1717
#include"libpq/pqformat.h"
1818
#include"miscadmin.h"
19+
#include"nodes/miscnodes.h"
1920
#include"tsearch/ts_locale.h"
2021
#include"tsearch/ts_type.h"
2122
#include"tsearch/ts_utils.h"
@@ -58,10 +59,16 @@ typedef enum
5859
/*
5960
* get token from query string
6061
*
61-
* *operator is filled in with OP_* when return values is PT_OPR,
62-
* but *weight could contain a distance value in case of phrase operator.
63-
* *strval, *lenval and *weight are filled in when return value is PT_VAL
62+
* All arguments except "state" are output arguments.
6463
*
64+
* If return value is PT_OPR, then *operator is filled with an OP_* code
65+
* and *weight will contain a distance value in case of phrase operator.
66+
*
67+
* If return value is PT_VAL, then *lenval, *strval, *weight, and *prefix
68+
* are filled.
69+
*
70+
* If PT_ERR is returned then a soft error has occurred. If state->escontext
71+
* isn't already filled then this should be reported as a generic parse error.
6572
*/
6673
typedefts_tokentype (*ts_tokenizer) (TSQueryParserStatestate,int8*operator,
6774
int*lenval,char**strval,
@@ -93,6 +100,9 @@ struct TSQueryParserStateData
93100

94101
/* state for value's parser */
95102
TSVectorParseStatevalstate;
103+
104+
/* context object for soft errors - must match valstate's escontext */
105+
Node*escontext;
96106
};
97107

98108
/*
@@ -194,7 +204,7 @@ parse_phrase_operator(TSQueryParserState pstate, int16 *distance)
194204
if (ptr==endptr)
195205
return false;
196206
elseif (errno==ERANGE||l<0||l>MAXENTRYPOS)
197-
ereport(ERROR,
207+
ereturn(pstate->escontext, false,
198208
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
199209
errmsg("distance in phrase operator must be an integer value between zero and %d inclusive",
200210
MAXENTRYPOS)));
@@ -301,10 +311,8 @@ gettoken_query_standard(TSQueryParserState state, int8 *operator,
301311
}
302312
elseif (t_iseq(state->buf,':'))
303313
{
304-
ereport(ERROR,
305-
(errcode(ERRCODE_SYNTAX_ERROR),
306-
errmsg("syntax error in tsquery: \"%s\"",
307-
state->buffer)));
314+
/* generic syntax error message is fine */
315+
returnPT_ERR;
308316
}
309317
elseif (!t_isspace(state->buf))
310318
{
@@ -320,12 +328,17 @@ gettoken_query_standard(TSQueryParserState state, int8 *operator,
320328
state->state=WAITOPERATOR;
321329
returnPT_VAL;
322330
}
331+
elseif (SOFT_ERROR_OCCURRED(state->escontext))
332+
{
333+
/* gettoken_tsvector reported a soft error */
334+
returnPT_ERR;
335+
}
323336
elseif (state->state==WAITFIRSTOPERAND)
324337
{
325338
returnPT_END;
326339
}
327340
else
328-
ereport(ERROR,
341+
ereturn(state->escontext,PT_ERR,
329342
(errcode(ERRCODE_SYNTAX_ERROR),
330343
errmsg("no operand in tsquery: \"%s\"",
331344
state->buffer)));
@@ -354,6 +367,11 @@ gettoken_query_standard(TSQueryParserState state, int8 *operator,
354367
*operator=OP_PHRASE;
355368
returnPT_OPR;
356369
}
370+
elseif (SOFT_ERROR_OCCURRED(state->escontext))
371+
{
372+
/* parse_phrase_operator reported a soft error */
373+
returnPT_ERR;
374+
}
357375
elseif (t_iseq(state->buf,')'))
358376
{
359377
state->buf++;
@@ -438,6 +456,11 @@ gettoken_query_websearch(TSQueryParserState state, int8 *operator,
438456
state->state=WAITOPERATOR;
439457
returnPT_VAL;
440458
}
459+
elseif (SOFT_ERROR_OCCURRED(state->escontext))
460+
{
461+
/* gettoken_tsvector reported a soft error */
462+
returnPT_ERR;
463+
}
441464
elseif (state->state==WAITFIRSTOPERAND)
442465
{
443466
returnPT_END;
@@ -529,12 +552,12 @@ pushValue_internal(TSQueryParserState state, pg_crc32 valcrc, int distance, int
529552
QueryOperand*tmp;
530553

531554
if (distance >=MAXSTRPOS)
532-
ereport(ERROR,
555+
ereturn(state->escontext,,
533556
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
534557
errmsg("value is too big in tsquery: \"%s\"",
535558
state->buffer)));
536559
if (lenval >=MAXSTRLEN)
537-
ereport(ERROR,
560+
ereturn(state->escontext,,
538561
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
539562
errmsg("operand is too long in tsquery: \"%s\"",
540563
state->buffer)));
@@ -562,7 +585,7 @@ pushValue(TSQueryParserState state, char *strval, int lenval, int16 weight, bool
562585
pg_crc32valcrc;
563586

564587
if (lenval >=MAXSTRLEN)
565-
ereport(ERROR,
588+
ereturn(state->escontext,,
566589
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
567590
errmsg("word is too long in tsquery: \"%s\"",
568591
state->buffer)));
@@ -686,11 +709,17 @@ makepol(TSQueryParserState state,
686709
return;
687710
casePT_ERR:
688711
default:
689-
ereport(ERROR,
690-
(errcode(ERRCODE_SYNTAX_ERROR),
691-
errmsg("syntax error in tsquery: \"%s\"",
692-
state->buffer)));
712+
/* don't overwrite a soft error saved by gettoken function */
713+
if (!SOFT_ERROR_OCCURRED(state->escontext))
714+
errsave(state->escontext,
715+
(errcode(ERRCODE_SYNTAX_ERROR),
716+
errmsg("syntax error in tsquery: \"%s\"",
717+
state->buffer)));
718+
return;
693719
}
720+
/* detect soft error in pushval or recursion */
721+
if (SOFT_ERROR_OCCURRED(state->escontext))
722+
return;
694723
}
695724

696725
cleanOpStack(state,opstack,&lenstack,OP_OR/* lowest */ );
@@ -769,6 +798,8 @@ findoprnd(QueryItem *ptr, int size, bool *needcleanup)
769798

770799

771800
/*
801+
* Parse the tsquery stored in "buf".
802+
*
772803
* Each value (operand) in the query is passed to pushval. pushval can
773804
* transform the simple value to an arbitrarily complex expression using
774805
* pushValue and pushOperator. It must push a single value with pushValue,
@@ -778,19 +809,27 @@ findoprnd(QueryItem *ptr, int size, bool *needcleanup)
778809
*
779810
* opaque is passed on to pushval as is, pushval can use it to store its
780811
* private state.
812+
*
813+
* The pushval function can record soft errors via escontext.
814+
* Callers must check SOFT_ERROR_OCCURRED to detect that.
815+
*
816+
* A bitmask of flags (see ts_utils.h) and an error context object
817+
* can be provided as well. If a soft error occurs, NULL is returned.
781818
*/
782819
TSQuery
783820
parse_tsquery(char*buf,
784821
PushFunctionpushval,
785822
Datumopaque,
786-
intflags)
823+
intflags,
824+
Node*escontext)
787825
{
788826
structTSQueryParserStateDatastate;
789827
inti;
790828
TSQueryquery;
791829
intcommonlen;
792830
QueryItem*ptr;
793831
ListCell*cell;
832+
boolnoisy;
794833
boolneedcleanup;
795834
inttsv_flags=P_TSV_OPR_IS_DELIM |P_TSV_IS_TSQUERY;
796835

@@ -808,15 +847,19 @@ parse_tsquery(char *buf,
808847
else
809848
state.gettoken=gettoken_query_standard;
810849

850+
/* emit nuisance NOTICEs only if not doing soft errors */
851+
noisy= !(escontext&&IsA(escontext,ErrorSaveContext));
852+
811853
/* init state */
812854
state.buffer=buf;
813855
state.buf=buf;
814856
state.count=0;
815857
state.state=WAITFIRSTOPERAND;
816858
state.polstr=NIL;
859+
state.escontext=escontext;
817860

818861
/* init value parser's state */
819-
state.valstate=init_tsvector_parser(state.buffer,tsv_flags);
862+
state.valstate=init_tsvector_parser(state.buffer,tsv_flags,escontext);
820863

821864
/* init list of operand */
822865
state.sumlen=0;
@@ -829,19 +872,23 @@ parse_tsquery(char *buf,
829872

830873
close_tsvector_parser(state.valstate);
831874

875+
if (SOFT_ERROR_OCCURRED(escontext))
876+
returnNULL;
877+
832878
if (state.polstr==NIL)
833879
{
834-
ereport(NOTICE,
835-
(errmsg("text-search query doesn't contain lexemes: \"%s\"",
836-
state.buffer)));
880+
if (noisy)
881+
ereport(NOTICE,
882+
(errmsg("text-search query doesn't contain lexemes: \"%s\"",
883+
state.buffer)));
837884
query= (TSQuery)palloc(HDRSIZETQ);
838885
SET_VARSIZE(query,HDRSIZETQ);
839886
query->size=0;
840887
returnquery;
841888
}
842889

843890
if (TSQUERY_TOO_BIG(list_length(state.polstr),state.sumlen))
844-
ereport(ERROR,
891+
ereturn(escontext,NULL,
845892
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
846893
errmsg("tsquery is too large")));
847894
commonlen=COMPUTESIZE(list_length(state.polstr),state.sumlen);
@@ -889,7 +936,7 @@ parse_tsquery(char *buf,
889936
* If there are QI_VALSTOP nodes, delete them and simplify the tree.
890937
*/
891938
if (needcleanup)
892-
query=cleanup_tsquery_stopwords(query);
939+
query=cleanup_tsquery_stopwords(query,noisy);
893940

894941
returnquery;
895942
}
@@ -908,8 +955,13 @@ Datum
908955
tsqueryin(PG_FUNCTION_ARGS)
909956
{
910957
char*in=PG_GETARG_CSTRING(0);
958+
Node*escontext=fcinfo->context;
911959

912-
PG_RETURN_TSQUERY(parse_tsquery(in,pushval_asis,PointerGetDatum(NULL),0));
960+
PG_RETURN_TSQUERY(parse_tsquery(in,
961+
pushval_asis,
962+
PointerGetDatum(NULL),
963+
0,
964+
escontext));
913965
}
914966

915967
/*

‎src/backend/utils/adt/tsquery_cleanup.c‎

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -383,7 +383,7 @@ calcstrlen(NODE *node)
383383
* Remove QI_VALSTOP (stopword) nodes from TSQuery.
384384
*/
385385
TSQuery
386-
cleanup_tsquery_stopwords(TSQueryin)
386+
cleanup_tsquery_stopwords(TSQueryin,boolnoisy)
387387
{
388388
int32len,
389389
lenstr,
@@ -403,8 +403,9 @@ cleanup_tsquery_stopwords(TSQuery in)
403403
root=clean_stopword_intree(maketree(GETQUERY(in)),&ladd,&radd);
404404
if (root==NULL)
405405
{
406-
ereport(NOTICE,
407-
(errmsg("text-search query contains only stop words or doesn't contain lexemes, ignored")));
406+
if (noisy)
407+
ereport(NOTICE,
408+
(errmsg("text-search query contains only stop words or doesn't contain lexemes, ignored")));
408409
out=palloc(HDRSIZETQ);
409410
out->size=0;
410411
SET_VARSIZE(out,HDRSIZETQ);

‎src/backend/utils/adt/tsvector.c‎

Lines changed: 11 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -15,6 +15,7 @@
1515
#include"postgres.h"
1616

1717
#include"libpq/pqformat.h"
18+
#include"nodes/miscnodes.h"
1819
#include"tsearch/ts_locale.h"
1920
#include"tsearch/ts_utils.h"
2021
#include"utils/builtins.h"
@@ -178,6 +179,7 @@ Datum
178179
tsvectorin(PG_FUNCTION_ARGS)
179180
{
180181
char*buf=PG_GETARG_CSTRING(0);
182+
Node*escontext=fcinfo->context;
181183
TSVectorParseStatestate;
182184
WordEntryIN*arr;
183185
inttotallen;
@@ -201,7 +203,7 @@ tsvectorin(PG_FUNCTION_ARGS)
201203
char*cur;
202204
intbuflen=256;/* allocated size of tmpbuf */
203205

204-
state=init_tsvector_parser(buf,0);
206+
state=init_tsvector_parser(buf,0,escontext);
205207

206208
arrlen=64;
207209
arr= (WordEntryIN*)palloc(sizeof(WordEntryIN)*arrlen);
@@ -210,14 +212,14 @@ tsvectorin(PG_FUNCTION_ARGS)
210212
while (gettoken_tsvector(state,&token,&toklen,&pos,&poslen,NULL))
211213
{
212214
if (toklen >=MAXSTRLEN)
213-
ereport(ERROR,
215+
ereturn(escontext, (Datum)0,
214216
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
215217
errmsg("word is too long (%ld bytes, max %ld bytes)",
216218
(long)toklen,
217219
(long) (MAXSTRLEN-1))));
218220

219221
if (cur-tmpbuf>MAXSTRPOS)
220-
ereport(ERROR,
222+
ereturn(escontext, (Datum)0,
221223
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
222224
errmsg("string is too long for tsvector (%ld bytes, max %ld bytes)",
223225
(long) (cur-tmpbuf), (long)MAXSTRPOS)));
@@ -261,13 +263,17 @@ tsvectorin(PG_FUNCTION_ARGS)
261263

262264
close_tsvector_parser(state);
263265

266+
/* Did gettoken_tsvector fail? */
267+
if (SOFT_ERROR_OCCURRED(escontext))
268+
PG_RETURN_NULL();
269+
264270
if (len>0)
265271
len=uniqueentry(arr,len,tmpbuf,&buflen);
266272
else
267273
buflen=0;
268274

269275
if (buflen>MAXSTRPOS)
270-
ereport(ERROR,
276+
ereturn(escontext, (Datum)0,
271277
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
272278
errmsg("string is too long for tsvector (%d bytes, max %d bytes)",buflen,MAXSTRPOS)));
273279

@@ -285,6 +291,7 @@ tsvectorin(PG_FUNCTION_ARGS)
285291
stroff+=arr[i].entry.len;
286292
if (arr[i].entry.haspos)
287293
{
294+
/* This should be unreachable because of MAXNUMPOS restrictions */
288295
if (arr[i].poslen>0xFFFF)
289296
elog(ERROR,"positions array too long");
290297

0 commit comments

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