walkero
/
pcre2
1
0
Fork 0
Code Issues Pull Requests Projects Releases 1 Wiki Activity

Add -LP and -LS (list properties, list scripts) features to pcre2test.

This commit is contained in:
Philip Hazel 2022-01-12 15:01:14 +00:00
parent 68fbc1982e
commit bf35c0518c
3 changed files with 214 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
ChangeLog
View File

@ -39,7 +39,8 @@ pcre2_substitute(), and the replacement argument of the latter, if the pointer
is NULL and the length is zero, treat as an empty string. Apparently a number
of applications treat NULL/0 in this way.
14. Added support for Bidi_Class and Bidi_Control Unicode properties.
14. Added support for Bidi_Class and a number of binary Unicode properties,
including Bidi_Control.
15. Fix some minor issues raised by clang sanitize.
@ -76,6 +77,8 @@ misaligned the frame that follows, resulting in an alignment fault when storing
a pointer to Fecode at the start of match. Patch to fix this issue by Jessica
Clarke PR#72.
20. Added -LP and -LS listing options to pcre2test.
Version 10.39 29-October-2021
-----------------------------

18
doc/pcre2test.1
View File

@ -1,4 +1,4 @@
.TH PCRE2TEST 1 "28 November 2021" "PCRE 10.40"
.TH PCRE2TEST 1 "12 January 2022" "PCRE 10.40"
.SH NAME
pcre2test - a program for testing Perl-compatible regular expressions.
.SH SYNOPSIS
@ -211,7 +211,17 @@ available, and the use of JIT for matching is verified.
\fB-LM\fP
List modifiers: write a list of available pattern and subject modifiers to the
standard output, then exit with zero exit code. All other options are ignored.
If both -C and -LM are present, whichever is first is recognized.
If both -C and any -Lx options are present, whichever is first is recognized.
.TP 10
\fB-LP\fP
List properties: write a list of recognized Unicode properties to the standard
output, then exit with zero exit code. All other options are ignored. If both
-C and any -Lx options are present, whichever is first is recognized.
.TP 10
\fB-LS\fP
List scripts: write a list of recogized Unicode script names to the standard
output, then exit with zero exit code. All other options are ignored. If both
-C and any -Lx options are present, whichever is first is recognized.
.TP 10
\fB-pattern\fP \fImodifier-list\fP
Behave as if each pattern line contains the given modifiers.
@ -2109,6 +2119,6 @@ Cambridge, England.
.rs
.sp
.nf
Last updated: 28 November 2021
Copyright (c) 1997-2021 University of Cambridge.
Last updated: 12 January 2022
Copyright (c) 1997-2022 University of Cambridge.
.fi

214
src/pcre2test.c
View File

@ -5490,7 +5490,7 @@ if ((pat_patctl.control & CTL_POSIX) != 0)
if ((pat_patctl.options & ~POSIX_SUPPORTED_COMPILE_OPTIONS) != 0)
{
show_compile_options(
pat_patctl.options & (uint32_t)(~POSIX_SUPPORTED_COMPILE_OPTIONS),
pat_patctl.options & (uint32_t)(~POSIX_SUPPORTED_COMPILE_OPTIONS),
msg, "");
msg = "";
}
@ -5499,7 +5499,7 @@ if ((pat_patctl.control & CTL_POSIX) != 0)
(uint32_t)(~POSIX_SUPPORTED_COMPILE_EXTRA_OPTIONS)) != 0)
{
show_compile_extra_options(
FLD(pat_context, extra_options) &
FLD(pat_context, extra_options) &
(uint32_t)(~POSIX_SUPPORTED_COMPILE_EXTRA_OPTIONS), msg, "");
msg = "";
}
@ -5509,7 +5509,7 @@ if ((pat_patctl.control & CTL_POSIX) != 0)
{
show_controls(
pat_patctl.control & (uint32_t)(~POSIX_SUPPORTED_COMPILE_CONTROLS),
pat_patctl.control2 & (uint32_t)(~POSIX_SUPPORTED_COMPILE_CONTROLS2),
pat_patctl.control2 & (uint32_t)(~POSIX_SUPPORTED_COMPILE_CONTROLS2),
msg);
msg = "";
}
@ -7310,7 +7310,7 @@ if (dat_datctl.replacement[0] != 0)
uint8_t *pr;
uint8_t rbuffer[REPLACE_BUFFSIZE];
uint8_t nbuffer[REPLACE_BUFFSIZE];
uint8_t *rbptr;
uint8_t *rbptr;
uint32_t xoptions;
uint32_t emoption; /* External match option */
PCRE2_SIZE j, rlen, nsize, erroroffset;
@ -7460,10 +7460,10 @@ if (dat_datctl.replacement[0] != 0)
{
PCRE2_SET_SUBSTITUTE_CALLOUT(dat_context, NULL, NULL); /* No callout */
}
/* There is a special option to set the replacement to NULL in order to test
that case. */
rbptr = ((dat_datctl.control2 & CTL2_NULL_REPLACEMENT) == 0)? rbuffer : NULL;
PCRE2_SUBSTITUTE(rc, compiled_code, pp, arg_ulen, dat_datctl.offset,
@ -7655,15 +7655,15 @@ for (gmatched = 0;; gmatched++)
}
/* The result of the match is now in capcount. First handle a successful
match. If pp was forced to be NULL (to test NULL handling) it will have been
treated as an empty string if the length was zero. So re-create that for
match. If pp was forced to be NULL (to test NULL handling) it will have been
treated as an empty string if the length was zero. So re-create that for
outputting. */
if (capcount >= 0)
{
int i;
if (pp == NULL) pp = (uint8_t *)"";
if (pp == NULL) pp = (uint8_t *)"";
if (capcount > (int)oveccount) /* Check for lunatic return value */
{
@ -8251,6 +8251,8 @@ printf(" -jit set default pattern modifier 'jit'\n");
printf(" -jitfast set default pattern modifier 'jitfast'\n");
printf(" -jitverify set default pattern modifier 'jitverify'\n");
printf(" -LM list pattern and subject modifiers, then exit\n");
printf(" -LP list non-script properties, then exit\n");
printf(" -LS list supported scripts, then exit\n");
printf(" -q quiet: do not output PCRE2 version number at start\n");
printf(" -pattern <s> set default pattern modifier fields\n");
printf(" -subject <s> set default subject modifier fields\n");
@ -8431,6 +8433,166 @@ return 0;
}
/*************************************************
* Format one property/script list item *
*************************************************/
#ifdef SUPPORT_UNICODE
static void
format_list_item(int16_t *ff, char *buff, BOOL isscript)
{
int count;
int maxi = 0;
const char *maxs = "";
size_t max = 0;
for (count = 0; ff[count] >= 0; count++) {}
/* Find the name to put first. For scripts, any 3-character name is chosen.
For non-scripts, or if there is no 3-character name, take the longest. */
for (int i = 0; ff[i] >= 0; i++)
{
const char *s = PRIV(utt_names) + ff[i];
size_t len = strlen(s);
if (isscript && len == 3)
{
maxi = i;
max = len;
maxs = s;
break;
}
else if (len > max)
{
max = len;
maxi = i;
maxs = s;
}
}
strcpy(buff, maxs);
buff += max;
if (count > 1)
{
const char *sep = " (";
for (int i = 0; i < count; i++)
{
if (i == maxi) continue;
buff += sprintf(buff, "%s%s", sep, PRIV(utt_names) + ff[i]);
sep = ", ";
}
(void)sprintf(buff, ")");
}
}
#endif /* SUPPORT_UNICODE */
/*************************************************
* Display scripts or properties *
*************************************************/
#define MAX_SYNONYMS 5
static void
display_properties(BOOL wantscripts)
{
#ifndef SUPPORT_UNICODE
printf("** This version of PCRE2 was compiled without Unicode support.\n");
#else
const char *typename;
uint16_t seentypes[1024];
uint16_t seenvalues[1024];
int seencount = 0;
int16_t found[256][MAX_SYNONYMS + 1];
int fc = 0;
int colwidth = 40;
int n;
if (wantscripts)
{
n = ucp_Script_Count;
typename = "SCRIPTS";
}
else
{
n = ucp_Bprop_Count;
typename = "PROPERTIES";
}
for (size_t i = 0; i < PRIV(utt_size); i++)
{
int k;
int m = 0;
int16_t *fv;
const ucp_type_table *t = PRIV(utt) + i;
unsigned int value = t->value;
if (wantscripts)
{
if (t->type != PT_SC && t->type != PT_SCX) continue;
}
else
{
if (t->type != PT_BOOL) continue;
}
for (k = 0; k < seencount; k++)
{
if (t->type == seentypes[k] && t->value == seenvalues[k]) break;
}
if (k < seencount) continue;
seentypes[seencount] = t->type;
seenvalues[seencount++] = t->value;
fv = found[fc++];
fv[m++] = t->name_offset;
for (size_t j = i + 1; j < PRIV(utt_size); j++)
{
const ucp_type_table *tt = PRIV(utt) + j;
if (tt->type != t->type || tt->value != value) continue;
if (m >= MAX_SYNONYMS)
printf("** Too many synonyms: %s ignored\n",
PRIV(utt_names) + tt->name_offset);
else fv[m++] = tt->name_offset;
}
fv[m] = -1;
}
printf("-------------------------- SUPPORTED %s --------------------------\n\n",
typename);
if (!wantscripts) printf(
"This release of PCRE2 supports Unicode's general category properties such\n"
"as Lu (upper case letter), bi-directional properties such as Bidi_Class,\n"
"and the following binary (yes/no) properties:\n\n");
for (int k = 0; k < (n+1)/2; k++)
{
int x;
char buff1[128];
char buff2[128];
format_list_item(found[k], buff1, wantscripts);
x = k + (n+1)/2;
if (x < n) format_list_item(found[x], buff2, wantscripts);
else buff2[0] = 0;
x = printf("%s", buff1);
while (x++ < colwidth) printf(" ");
printf("%s\n", buff2);
}
#endif /* SUPPORT_UNICODE */
}
/*************************************************
* Display one modifier *
@ -8445,8 +8607,8 @@ printf("%c%s", c, m->name);
for (size_t i = 0; i < C1MODLISTCOUNT; i++)
{
if (strcmp(m->name, c1modlist[i].fullname) == 0)
printf(" (%c)", c1modlist[i].onechar);
}
printf(" (%c)", c1modlist[i].onechar);
}
}
@ -8505,19 +8667,19 @@ for (i = 0; i < MODLISTCOUNT; i++)
break;
}
if (for_pattern == is_pattern)
if (for_pattern == is_pattern)
{
extra[n] = 0;
extra[n] = 0;
for (size_t k = 0; k < C1MODLISTCOUNT; k++)
{
if (strcmp(m->name, c1modlist[k].fullname) == 0)
{
extra[n] += 4;
extra[n] += 4;
break;
}
}
}
}
list[n++] = i;
}
}
}
/* Now print from the list in two columns. */
@ -8672,6 +8834,22 @@ while (argc > 1 && argv[op][0] == '-' && argv[op][1] != 0)
goto EXIT;
}
/* List properties and exit */
if (strcmp(arg, "-LP") == 0)
{
display_properties(FALSE);
goto EXIT;
}
/* List scripts and exit */
if (strcmp(arg, "-LS") == 0)
{
display_properties(TRUE);
goto EXIT;
}
/* Display and/or set return code for configuration options. */
if (strcmp(arg, "-C") == 0)