bool
match(const char *str, const char *pattern)
{
int r;
r = MatchText(str, strlen(str), pattern, strlen(pattern), NULL);
return (r == LIKE_TRUE);
}
List *
capture(const char *str, const char *pattern, int nparams)
{
int r;
List *params = NIL;
r = MatchText(str, strlen(str), pattern, strlen(pattern), ¶ms);
if (r == LIKE_TRUE && list_length(params) == nparams)
return params;
list_free_deep(params);
return NIL;
}
bool FeedFilter::Term::MatchValue(const char* szStrValue, long long iIntValue)
{
double fFloatValue = (double)iIntValue;
char szIntBuf[100];
if (m_eCommand < fcEqual && !szStrValue)
{
snprintf(szIntBuf, 100, "%lld", iIntValue);
szIntBuf[100-1] = '\0';
szStrValue = szIntBuf;
}
else if (m_eCommand >= fcEqual && szStrValue)
{
fFloatValue = atof(szStrValue);
iIntValue = (long long)fFloatValue;
}
switch (m_eCommand)
{
case fcText:
return MatchText(szStrValue);
case fcRegex:
return MatchRegex(szStrValue);
case fcEqual:
return m_bFloat ? fFloatValue == m_fFloatParam : iIntValue == m_iIntParam;
case fcLess:
return m_bFloat ? fFloatValue < m_fFloatParam : iIntValue < m_iIntParam;
case fcLessEqual:
return m_bFloat ? fFloatValue <= m_fFloatParam : iIntValue <= m_iIntParam;
case fcGreater:
return m_bFloat ? fFloatValue > m_fFloatParam : iIntValue > m_iIntParam;
case fcGreaterEqual:
return m_bFloat ? fFloatValue >= m_fFloatParam : iIntValue >= m_iIntParam;
default:
return false;
}
}
static int
MatchText(char *t, int tlen, char *p, int plen,
pg_locale_t locale, bool locale_is_c)
{
/* Fast path for match-everything pattern */
if (plen == 1 && *p == '%')
return LIKE_TRUE;
/*
* In this loop, we advance by char when matching wildcards (and thus on
* recursive entry to this function we are properly char-synced). On other
* occasions it is safe to advance by byte, as the text and pattern will
* be in lockstep. This allows us to perform all comparisons between the
* text and pattern on a byte by byte basis, even for multi-byte
* encodings.
*/
while (tlen > 0 && plen > 0)
{
if (*p == '\\')
{
/* Next pattern byte must match literally, whatever it is */
NextByte(p, plen);
/* ... and there had better be one, per SQL standard */
if (plen <= 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_ESCAPE_SEQUENCE),
errmsg("LIKE pattern must not end with escape character")));
if (GETCHAR(*p) != GETCHAR(*t))
return LIKE_FALSE;
}
else if (*p == '%')
{
char firstpat;
/*
* % processing is essentially a search for a text position at
* which the remainder of the text matches the remainder of the
* pattern, using a recursive call to check each potential match.
*
* If there are wildcards immediately following the %, we can skip
* over them first, using the idea that any sequence of N _'s and
* one or more %'s is equivalent to N _'s and one % (ie, it will
* match any sequence of at least N text characters). In this way
* we will always run the recursive search loop using a pattern
* fragment that begins with a literal character-to-match, thereby
* not recursing more than we have to.
*/
NextByte(p, plen);
while (plen > 0)
{
if (*p == '%')
NextByte(p, plen);
else if (*p == '_')
{
/* If not enough text left to match the pattern, ABORT */
if (tlen <= 0)
return LIKE_ABORT;
NextChar(t, tlen);
NextByte(p, plen);
}
else
break; /* Reached a non-wildcard pattern char */
}
/*
* If we're at end of pattern, match: we have a trailing % which
* matches any remaining text string.
*/
if (plen <= 0)
return LIKE_TRUE;
/*
* Otherwise, scan for a text position at which we can match the
* rest of the pattern. The first remaining pattern char is known
* to be a regular or escaped literal character, so we can compare
* the first pattern byte to each text byte to avoid recursing
* more than we have to. This fact also guarantees that we don't
* have to consider a match to the zero-length substring at the
* end of the text.
*/
if (*p == '\\')
{
if (plen < 2)
ereport(ERROR,
(errcode(ERRCODE_INVALID_ESCAPE_SEQUENCE),
errmsg("LIKE pattern must not end with escape character")));
firstpat = GETCHAR(p[1]);
}
else
firstpat = GETCHAR(*p);
while (tlen > 0)
{
if (GETCHAR(*t) == firstpat)
{
int matched = MatchText(t, tlen, p, plen,
locale, locale_is_c);
if (matched != LIKE_FALSE)
return matched; /* TRUE or ABORT */
}
NextChar(t, tlen);
}
/*
* End of text with no match, so no point in trying later places
* to start matching this pattern.
*/
return LIKE_ABORT;
}
else if (*p == '_')
{
/* _ matches any single character, and we know there is one */
NextChar(t, tlen);
NextByte(p, plen);
continue;
}
else if (GETCHAR(*p) != GETCHAR(*t))
{
/* non-wildcard pattern char fails to match text char */
return LIKE_FALSE;
}
/*
* Pattern and text match, so advance.
*
* It is safe to use NextByte instead of NextChar here, even for
* multi-byte character sets, because we are not following immediately
* after a wildcard character. If we are in the middle of a multibyte
* character, we must already have matched at least one byte of the
* character from both text and pattern; so we cannot get out-of-sync
* on character boundaries. And we know that no backend-legal
* encoding allows ASCII characters such as '%' to appear as non-first
* bytes of characters, so we won't mistakenly detect a new___ wildcard.
*/
NextByte(t, tlen);
NextByte(p, plen);
}
if (tlen > 0)
return LIKE_FALSE; /* end of pattern, but not of text */
/*
* End of text, but perhaps not of pattern. Match iff the remaining
* pattern can match a zero-length string, ie, it's zero or more %'s.
*/
while (plen > 0 && *p == '%')
NextByte(p, plen);
if (plen <= 0)
return LIKE_TRUE;
/*
* End of text with no match, so no point in trying later places to start
* matching this pattern.
*/
return LIKE_ABORT;
} /* MatchText() */
static int
MatchText(char *t, int tlen, char *p, int plen)
{
/* Fast path for match-everything pattern */
if ((plen == 1) && (*p == '%'))
return LIKE_TRUE;
/*
* In this loop, we advance by char when matching wildcards (and thus on
* recursive entry to this function we are properly char-synced). On other
* occasions it is safe to advance by byte, as the text and pattern will
* be in lockstep. This allows us to perform all comparisons between the
* text and pattern on a byte by byte basis, even for multi-byte
* encodings.
*/
while ((tlen > 0) && (plen > 0))
{
if (*p == '\\')
{
/* Next byte must match literally, whatever it is */
NextByte(p, plen);
if ((plen <= 0) || TCHAR(*p) != TCHAR(*t))
return LIKE_FALSE;
}
else if (*p == '%')
{
/*
* % processing is essentially a search for a match for what
* follows the %, plus a recursive match of the remainder. We
* succeed if and only if both conditions are met.
*/
/* %% is the same as % according to the SQL standard */
/* Advance past all %'s */
while ((plen > 0) && (*p == '%'))
NextByte(p, plen);
/* Trailing percent matches everything. */
if (plen <= 0)
return LIKE_TRUE;
/*
* Otherwise, scan for a text position at which we can match the
* rest of the pattern.
*/
if (*p == '_')
{
/* %_ is the same as _% - avoid matching _ repeatedly */
NextChar(t, tlen);
NextByte(p, plen);
if (tlen <= 0)
{
return (plen <= 0) ? LIKE_TRUE : LIKE_ABORT;
}
else if (plen <= 0)
{
return LIKE_FALSE;
}
while (tlen > 0)
{
int matched = MatchText(t, tlen, p, plen);
if (matched != LIKE_FALSE)
return matched; /* TRUE or ABORT */
NextChar(t, tlen);
}
}
else
{
char firstpat = TCHAR(*p);
if (*p == '\\')
{
if (plen < 2)
return LIKE_FALSE;
firstpat = TCHAR(p[1]);
}
while (tlen > 0)
{
/*
* Optimization to prevent most recursion: don't recurse
* unless first pattern byte matches first text byte.
*/
if (TCHAR(*t) == firstpat)
{
int matched = MatchText(t, tlen, p, plen);
if (matched != LIKE_FALSE)
return matched; /* TRUE or ABORT */
}
NextChar(t, tlen);
}
}
/*
* End of text with no match, so no point in trying later places
* to start matching this pattern.
*/
return LIKE_ABORT;
}
else if (*p == '_')
{
NextChar(t, tlen);
NextByte(p, plen);
continue;
}
else if (TCHAR(*t) != TCHAR(*p))
{
/*
* Not the single-character wildcard and no explicit match? Then
* time to quit...
*/
return LIKE_FALSE;
}
/*
* It is safe to use NextByte instead of NextChar here, even for
* multi-byte character sets, because we are not following immediately
* after a wildcard character. If we are in the middle of a multibyte
* character, we must already have matched at least one byte of the
* character from both text and pattern; so we cannot get out-of-sync
* on character boundaries. And we know that no backend-legal
* encoding allows ASCII characters such as '%' to appear as non-first
* bytes of characters, so we won't mistakenly detect a new wildcard.
*/
NextByte(t, tlen);
NextByte(p, plen);
}
if (tlen > 0)
return LIKE_FALSE; /* end of pattern, but not of text */
/* End of input string. Do we have matching pattern remaining? */
while ((plen > 0) && (*p == '%')) /* allow multiple %'s at end of
* pattern */
NextByte(p, plen);
if (plen <= 0)
return LIKE_TRUE;
/*
* End of text with no match, so no point in trying later places to start
* matching this pattern.
*/
return LIKE_ABORT;
} /* MatchText() */
static int
MatchText(unsigned char *t, int tlen, unsigned char *p, int plen)
{
/* Fast path for match-everything pattern */
if ((plen == 1) && (*p == '%'))
return LIKE_TRUE;
while ((tlen > 0) && (plen > 0))
{
if (*p == '\\')
{
/* Next pattern char must match literally, whatever it is */
NextChar(p, plen);
if ((plen <= 0) || !CHAREQ(t, p))
return LIKE_FALSE;
}
else if (*p == '%')
{
/* %% is the same as % according to the SQL standard */
/* Advance past all %'s */
while ((plen > 0) && (*p == '%'))
NextChar(p, plen);
/* Trailing percent matches everything. */
if (plen <= 0)
return LIKE_TRUE;
/*
* Otherwise, scan for a text position at which we can match
* the rest of the pattern.
*/
while (tlen > 0)
{
/*
* Optimization to prevent most recursion: don't recurse
* unless first pattern char might match this text char.
*/
if (CHAREQ(t, p) || (*p == '\\') || (*p == '_'))
{
int matched = MatchText(t, tlen, p, plen);
if (matched != LIKE_FALSE)
return matched; /* TRUE or ABORT */
}
NextChar(t, tlen);
}
/*
* End of text with no match, so no point in trying later
* places to start matching this pattern.
*/
return LIKE_ABORT;
}
else if ((*p != '_') && !CHAREQ(t, p))
{
/*
* Not the single-character wildcard and no explicit match?
* Then time to quit...
*/
return LIKE_FALSE;
}
NextChar(t, tlen);
NextChar(p, plen);
}
if (tlen > 0)
return LIKE_FALSE; /* end of pattern, but not of text */
/* End of input string. Do we have matching pattern remaining? */
while ((plen > 0) && (*p == '%')) /* allow multiple %'s at end of
* pattern */
NextChar(p, plen);
if (plen <= 0)
return LIKE_TRUE;
/*
* End of text with no match, so no point in trying later places to
* start matching this pattern.
*/
return LIKE_ABORT;
} /* MatchText() */
static int
MatchText(const char *t, size_t tlen, const char *p, size_t plen, List **params)
{
while (tlen > 0 && plen > 0)
{
if (plen < 2 || *p != '%')
{
/* non-wildcard pattern char fails to match text char */
if (*p != *t)
return LIKE_FALSE;
}
else if (p[1] == '%')
{
/* %% is % */
NextByte(p, plen);
if (*p != *t)
return LIKE_FALSE;
}
else
{
const char *begin = p;
const char *w = t;
char firstpat;
/* Skip until the type specifer */
p = strpbrk(begin + 1, "diouxXeEfFgGaAcspm");
if (p == NULL)
return LIKE_FALSE; /* bad format */
p++;
plen -= p - begin;
if (plen <= 0)
{
if (params)
*params = lcons(strdup_with_len(t, tlen), *params);
return LIKE_TRUE; /* matches everything. */
}
/*
* Otherwise, scan for a text position at which we can match the
* rest of the pattern.
*/
firstpat = *p;
while (tlen > 0)
{
/*
* Optimization to prevent most recursion: don't recurse
* unless first pattern byte matches first text byte.
*/
if (*t == firstpat)
{
int matched = MatchText(t, tlen, p, plen, params);
if (matched == LIKE_TRUE && params)
*params = lcons(strdup_with_len(w, t - w), *params);
if (matched != LIKE_FALSE)
return matched; /* TRUE or ABORT */
}
NextChar(t, tlen);
}
/*
* End of text with no match, so no point in trying later places
* to start matching this pattern.
*/
return LIKE_ABORT;
}
NextByte(t, tlen);
NextByte(p, plen);
}
if (tlen > 0)
return LIKE_FALSE; /* end of pattern, but not of text */
/* End of text string. Do we have matching pattern remaining? */
while (plen > 0 && *p == '%')
{
const char *begin = p;
p = strpbrk(begin + 1, "diouxXeEfFgGaAcspm");
if (p == NULL)
return LIKE_FALSE; /* bad format */
p++;
plen -= p - begin;
}
if (plen <= 0)
return LIKE_TRUE;
/*
* End of text with no match, so no point in trying later places to start
* matching this pattern.
*/
return LIKE_ABORT;
}
