Fix ts_headline() edge cases for empty query and empty search text.

tsquery's GETQUERY() macro is only safe to apply to a tsquery
that is known non-empty; otherwise it gives a pointer to garbage.
Before commit 5a617d75d, ts_headline() avoided this pitfall, but
only in a very indirect, nonobvious way.  (hlCover could not reach
its TS_execute call, because if the query contains no lexemes
then hlFirstIndex would surely return -1.)  After that commit,
it fell into the trap, resulting in weird errors such as
"unrecognized operator" and/or valgrind complaints.  In HEAD,
fix this by not calling TS_execute_locations() at all for an
empty query.  In the back branches, add a defensive check to
hlCover() --- that's not fixing any live bug, but I judge the
code a bit too fragile as-is.

Also, both mark_hl_fragments() and mark_hl_words() were careless
about the possibility of empty search text: in the cases where
no match has been found, they'd end up telling mark_fragment() to
mark from word indexes 0 to 0 inclusive, even when there is no
word 0.  This is harmless since we over-allocated the prs->words
array, but it does annoy valgrind.  Fix so that the end index is -1
and thus mark_fragment() will do nothing in such cases.

Bottom line is that this fixes a live bug in HEAD, but in the
back branches it's only getting rid of a valgrind nitpick.
Back-patch anyway.

Per report from Alexander Lakhin.

Discussion: https://postgr.es/m/c27f642d-020b-01ff-ae61-086af287c4fd@gmail.com

diff --git a/src/backend/tsearch/wparser_def.c b/src/backend/tsearch/wparser_def.c
index 840a44ec007b105d6e4f78023959f62740d988aa..23e4e9d98a9efcae9ec9ade10208e53993910c6a 100644 (file)
--- a/src/backend/tsearch/wparser_def.c
+++ b/src/backend/tsearch/wparser_def.c
@@ -2417,7 +2417,8 @@ mark_hl_fragments(HeadlineParsedText *prs, TSQuery query, List *locations,
    /* show the first min_words words if we have not marked anything */
    if (num_f <= 0)
    {
-       startpos = endpos = curlen = 0;
+       startpos = curlen = 0;
+       endpos = -1;
        for (i = 0; i < prs->curwords && curlen < min_words; i++)
        {
            if (!NONWORDTOKEN(prs->words[i].type))
@@ -2571,7 +2572,7 @@ mark_hl_words(HeadlineParsedText *prs, TSQuery query, List *locations,
        if (bestlen < 0)
        {
            curlen = 0;
-           pose = 0;
+           pose = -1;
            for (i = 0; i < prs->curwords && curlen < min_words; i++)
            {
                if (!NONWORDTOKEN(prs->words[i].type))
@@ -2601,7 +2602,6 @@ prsd_headline(PG_FUNCTION_ARGS)
    HeadlineParsedText *prs = (HeadlineParsedText *) PG_GETARG_POINTER(0);
    List       *prsoptions = (List *) PG_GETARG_POINTER(1);
    TSQuery     query = PG_GETARG_TSQUERY(2);
-   hlCheck     ch;
    List       *locations;
 
    /* default option values: */
@@ -2671,10 +2671,17 @@ prsd_headline(PG_FUNCTION_ARGS)
    }
 
    /* Locate words and phrases matching the query */
-   ch.words = prs->words;
-   ch.len = prs->curwords;
-   locations = TS_execute_locations(GETQUERY(query), &ch, TS_EXEC_EMPTY,
-                                    checkcondition_HL);
+   if (query->size > 0)
+   {
+       hlCheck     ch;
+
+       ch.words = prs->words;
+       ch.len = prs->curwords;
+       locations = TS_execute_locations(GETQUERY(query), &ch, TS_EXEC_EMPTY,
+                                        checkcondition_HL);
+   }
+   else
+       locations = NIL;        /* empty query matches nothing */
 
    /* Apply appropriate headline selector */
    if (max_fragments == 0)

diff --git a/src/test/regress/expected/tsearch.out b/src/test/regress/expected/tsearch.out
index 0e682457430a69c6b2717d2cec5a0198c598b4f6..6b3ad38f39689d5fec91368ce290a9f121f2adfb 100644 (file)
--- a/src/test/regress/expected/tsearch.out
+++ b/src/test/regress/expected/tsearch.out
@@ -2127,6 +2127,27 @@ to_tsquery('english','Lorem') && phraseto_tsquery('english','ullamcorper urna'),
  <b>Lorem</b> ipsum <b>urna</b>.  Nullam nullam <b>ullamcorper</b> <b>urna</b>
 (1 row)
 
+-- Edge cases with empty query
+SELECT ts_headline('english',
+'', ''::tsquery);
+NOTICE:  text-search query doesn't contain lexemes: ""
+LINE 2: '', ''::tsquery);
+            ^
+ ts_headline 
+-------------
+ 
+(1 row)
+
+SELECT ts_headline('english',
+'foo bar', ''::tsquery);
+NOTICE:  text-search query doesn't contain lexemes: ""
+LINE 2: 'foo bar', ''::tsquery);
+                   ^
+ ts_headline 
+-------------
+ foo bar
+(1 row)
+
 --Rewrite sub system
 CREATE TABLE test_tsquery (txtkeyword TEXT, txtsample TEXT);
 \set ECHO none

diff --git a/src/test/regress/sql/tsearch.sql b/src/test/regress/sql/tsearch.sql
index b56477a813988e7a7409515047bc0d6fbd104b69..f6f5fb5c2704971850ff2a18a4fbbb74f331475e 100644 (file)
--- a/src/test/regress/sql/tsearch.sql
+++ b/src/test/regress/sql/tsearch.sql
@@ -640,6 +640,12 @@ SELECT ts_headline('english',
 to_tsquery('english','Lorem') && phraseto_tsquery('english','ullamcorper urna'),
 'MaxFragments=100, MaxWords=100, MinWords=1');
 
+-- Edge cases with empty query
+SELECT ts_headline('english',
+'', ''::tsquery);
+SELECT ts_headline('english',
+'foo bar', ''::tsquery);
+
 --Rewrite sub system
 
 CREATE TABLE test_tsquery (txtkeyword TEXT, txtsample TEXT);