diff options
Diffstat (limited to 'src/3rdparty/pcre2/src/pcre2_dfa_match.c')
| -rw-r--r-- | src/3rdparty/pcre2/src/pcre2_dfa_match.c | 87 |
1 files changed, 76 insertions, 11 deletions
diff --git a/src/3rdparty/pcre2/src/pcre2_dfa_match.c b/src/3rdparty/pcre2/src/pcre2_dfa_match.c index 060dc7669a..b16e594cc0 100644 --- a/src/3rdparty/pcre2/src/pcre2_dfa_match.c +++ b/src/3rdparty/pcre2/src/pcre2_dfa_match.c @@ -7,7 +7,7 @@ and semantics are as close as possible to those of the Perl 5 language. Written by Philip Hazel Original API code Copyright (c) 1997-2012 University of Cambridge - New API code Copyright (c) 2016-2021 University of Cambridge + New API code Copyright (c) 2016-2022 University of Cambridge ----------------------------------------------------------------------------- Redistribution and use in source and binary forms, with or without @@ -350,7 +350,7 @@ Returns: the return from the callout */ static int -do_callout(PCRE2_SPTR code, PCRE2_SIZE *offsets, PCRE2_SPTR current_subject, +do_callout_dfa(PCRE2_SPTR code, PCRE2_SIZE *offsets, PCRE2_SPTR current_subject, PCRE2_SPTR ptr, dfa_match_block *mb, PCRE2_SIZE extracode, PCRE2_SIZE *lengthptr) { @@ -1193,6 +1193,11 @@ for (;;) OK = prop->script == code[2]; break; + case PT_SCX: + OK = (prop->script == code[2] || + MAPBIT(PRIV(ucd_script_sets) + UCD_SCRIPTX_PROP(prop), code[2]) != 0); + break; + /* These are specials for combination cases. */ case PT_ALNUM: @@ -1240,6 +1245,15 @@ for (;;) c >= 0xe000; break; + case PT_BIDICL: + OK = UCD_BIDICLASS(c) == code[2]; + break; + + case PT_BOOL: + OK = MAPBIT(PRIV(ucd_boolprop_sets) + + UCD_BPROPS_PROP(prop), code[2]) != 0; + break; + /* Should never occur, but keep compilers from grumbling. */ default: @@ -1451,6 +1465,11 @@ for (;;) OK = prop->script == code[3]; break; + case PT_SCX: + OK = (prop->script == code[3] || + MAPBIT(PRIV(ucd_script_sets) + UCD_SCRIPTX_PROP(prop), code[3]) != 0); + break; + /* These are specials for combination cases. */ case PT_ALNUM: @@ -1498,6 +1517,15 @@ for (;;) c >= 0xe000; break; + case PT_BIDICL: + OK = UCD_BIDICLASS(c) == code[3]; + break; + + case PT_BOOL: + OK = MAPBIT(PRIV(ucd_boolprop_sets) + + UCD_BPROPS_PROP(prop), code[3]) != 0; + break; + /* Should never occur, but keep compilers from grumbling. */ default: @@ -1692,6 +1720,11 @@ for (;;) OK = prop->script == code[3]; break; + case PT_SCX: + OK = (prop->script == code[3] || + MAPBIT(PRIV(ucd_script_sets) + UCD_SCRIPTX_PROP(prop), code[3]) != 0); + break; + /* These are specials for combination cases. */ case PT_ALNUM: @@ -1739,6 +1772,15 @@ for (;;) c >= 0xe000; break; + case PT_BIDICL: + OK = UCD_BIDICLASS(c) == code[3]; + break; + + case PT_BOOL: + OK = MAPBIT(PRIV(ucd_boolprop_sets) + + UCD_BPROPS_PROP(prop), code[3]) != 0; + break; + /* Should never occur, but keep compilers from grumbling. */ default: @@ -1958,6 +2000,12 @@ for (;;) OK = prop->script == code[1 + IMM2_SIZE + 2]; break; + case PT_SCX: + OK = (prop->script == code[1 + IMM2_SIZE + 2] || + MAPBIT(PRIV(ucd_script_sets) + UCD_SCRIPTX_PROP(prop), + code[1 + IMM2_SIZE + 2]) != 0); + break; + /* These are specials for combination cases. */ case PT_ALNUM: @@ -2005,6 +2053,15 @@ for (;;) c >= 0xe000; break; + case PT_BIDICL: + OK = UCD_BIDICLASS(c) == code[1 + IMM2_SIZE + 2]; + break; + + case PT_BOOL: + OK = MAPBIT(PRIV(ucd_boolprop_sets) + + UCD_BPROPS_PROP(prop), code[1 + IMM2_SIZE + 2]) != 0; + break; + /* Should never occur, but keep compilers from grumbling. */ default: @@ -2742,7 +2799,7 @@ for (;;) || code[LINK_SIZE + 1] == OP_CALLOUT_STR) { PCRE2_SIZE callout_length; - rrc = do_callout(code, offsets, current_subject, ptr, mb, + rrc = do_callout_dfa(code, offsets, current_subject, ptr, mb, 1 + LINK_SIZE, &callout_length); if (rrc < 0) return rrc; /* Abandon */ if (rrc > 0) break; /* Fail this thread */ @@ -3139,7 +3196,7 @@ for (;;) case OP_CALLOUT_STR: { PCRE2_SIZE callout_length; - rrc = do_callout(code, offsets, current_subject, ptr, mb, 0, + rrc = do_callout_dfa(code, offsets, current_subject, ptr, mb, 0, &callout_length); if (rrc < 0) return rrc; /* Abandon */ if (rrc == 0) @@ -3285,20 +3342,22 @@ rws->next = NULL; rws->size = RWS_BASE_SIZE; rws->free = RWS_BASE_SIZE - RWS_ANCHOR_SIZE; -/* A length equal to PCRE2_ZERO_TERMINATED implies a zero-terminated -subject string. */ +/* Recognize NULL, length 0 as an empty string. */ -if (length == PCRE2_ZERO_TERMINATED) - { - length = PRIV(strlen)(subject); - was_zero_terminated = 1; - } +if (subject == NULL && length == 0) subject = (PCRE2_SPTR)""; /* Plausibility checks */ if ((options & ~PUBLIC_DFA_MATCH_OPTIONS) != 0) return PCRE2_ERROR_BADOPTION; if (re == NULL || subject == NULL || workspace == NULL || match_data == NULL) return PCRE2_ERROR_NULL; + +if (length == PCRE2_ZERO_TERMINATED) + { + length = PRIV(strlen)(subject); + was_zero_terminated = 1; + } + if (wscount < 20) return PCRE2_ERROR_DFA_WSSIZE; if (start_offset > length) return PCRE2_ERROR_BADOFFSET; @@ -3998,4 +4057,10 @@ while (rws->next != NULL) return rc; } +/* These #undefs are here to enable unity builds with CMake. */ + +#undef NLBLOCK /* Block containing newline information */ +#undef PSSTART /* Field containing processed string start */ +#undef PSEND /* Field containing processed string end */ + /* End of pcre2_dfa_match.c */ |