| 67 |
#endif |
#endif |
| 68 |
|
|
| 69 |
|
|
| 70 |
#define PCRE_SPY /* For Win32 build, import data, not export */ |
/* We have to include pcre_internal.h because we need the internal info for |
| 71 |
|
displaying the results of pcre_study() and we also need to know about the |
| 72 |
/* We include pcre_internal.h because we need the internal info for displaying |
internal macros, structures, and other internal data values; pcretest has |
| 73 |
the results of pcre_study() and we also need to know about the internal |
"inside information" compared to a program that strictly follows the PCRE API. |
| 74 |
macros, structures, and other internal data values; pcretest has "inside |
|
| 75 |
information" compared to a program that strictly follows the PCRE API. */ |
Although pcre_internal.h does itself include pcre.h, we explicitly include it |
| 76 |
|
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set |
| 77 |
|
appropriately for an application, not for building PCRE. */ |
| 78 |
|
|
| 79 |
|
#include "pcre.h" |
| 80 |
#include "pcre_internal.h" |
#include "pcre_internal.h" |
| 81 |
|
|
| 82 |
/* We need access to the data tables that PCRE uses. So as not to have to keep |
/* We need access to the data tables that PCRE uses. So as not to have to keep |
| 663 |
*************************************************/ |
*************************************************/ |
| 664 |
|
|
| 665 |
/* This is used both at compile and run-time to check for <xxx> escapes, where |
/* This is used both at compile and run-time to check for <xxx> escapes, where |
| 666 |
xxx is LF, CR, CRLF, or ANY. Print a message and return 0 if there is no match. |
xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is |
| 667 |
|
no match. |
| 668 |
|
|
| 669 |
Arguments: |
Arguments: |
| 670 |
p points after the leading '<' |
p points after the leading '<' |
| 679 |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
| 680 |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
| 681 |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
| 682 |
|
if (strncmp((char *)p, "anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
| 683 |
if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY; |
if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY; |
| 684 |
fprintf(f, "Unknown newline type at: <%s\n", p); |
fprintf(f, "Unknown newline type at: <%s\n", p); |
| 685 |
return 0; |
return 0; |
| 852 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
| 853 |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
| 854 |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
| 855 |
|
(rc == -2)? "ANYCRLF" : |
| 856 |
(rc == -1)? "ANY" : "???"); |
(rc == -1)? "ANY" : "???"); |
| 857 |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
| 858 |
printf(" Internal link size = %d\n", rc); |
printf(" Internal link size = %d\n", rc); |
| 1448 |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
| 1449 |
break; |
break; |
| 1450 |
|
|
| 1451 |
|
case PCRE_NEWLINE_ANYCRLF: |
| 1452 |
|
fprintf(outfile, "Forced newline sequence: ANYCRLF\n"); |
| 1453 |
|
break; |
| 1454 |
|
|
| 1455 |
case PCRE_NEWLINE_ANY: |
case PCRE_NEWLINE_ANY: |
| 1456 |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
| 1457 |
break; |
break; |
| 1601 |
for (;;) |
for (;;) |
| 1602 |
{ |
{ |
| 1603 |
uschar *q; |
uschar *q; |
| 1604 |
uschar *bptr = dbuffer; |
uschar *bptr; |
| 1605 |
int *use_offsets = offsets; |
int *use_offsets = offsets; |
| 1606 |
int use_size_offsets = size_offsets; |
int use_size_offsets = size_offsets; |
| 1607 |
int callout_data = 0; |
int callout_data = 0; |
| 1657 |
p = buffer; |
p = buffer; |
| 1658 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
| 1659 |
|
|
| 1660 |
q = dbuffer; |
bptr = q = dbuffer; |
| 1661 |
while ((c = *p++) != 0) |
while ((c = *p++) != 0) |
| 1662 |
{ |
{ |
| 1663 |
int i = 0; |
int i = 0; |
| 1982 |
|
|
| 1983 |
for (;; gmatched++) /* Loop for /g or /G */ |
for (;; gmatched++) /* Loop for /g or /G */ |
| 1984 |
{ |
{ |
|
int gany_fudge; |
|
| 1985 |
if (timeitm > 0) |
if (timeitm > 0) |
| 1986 |
{ |
{ |
| 1987 |
register int i; |
register int i; |
| 2221 |
} |
} |
| 2222 |
|
|
| 2223 |
/* Failed to match. If this is a /g or /G loop and we previously set |
/* Failed to match. If this is a /g or /G loop and we previously set |
| 2224 |
g_notempty after a null match, this is not necessarily the end. |
g_notempty after a null match, this is not necessarily the end. We want |
| 2225 |
We want to advance the start offset, and continue. In the case of UTF-8 |
to advance the start offset, and continue. We won't be at the end of the |
| 2226 |
matching, the advance must be one character, not one byte. Fudge the |
string - that was checked before setting g_notempty. |
| 2227 |
offset values to achieve this. We won't be at the end of the string - |
|
| 2228 |
that was checked before setting g_notempty. */ |
Complication arises in the case when the newline option is "any" or |
| 2229 |
|
"anycrlf". If the previous match was at the end of a line terminated by |
| 2230 |
|
CRLF, an advance of one character just passes the \r, whereas we should |
| 2231 |
|
prefer the longer newline sequence, as does the code in pcre_exec(). |
| 2232 |
|
Fudge the offset value to achieve this. |
| 2233 |
|
|
| 2234 |
|
Otherwise, in the case of UTF-8 matching, the advance must be one |
| 2235 |
|
character, not one byte. */ |
| 2236 |
|
|
| 2237 |
else |
else |
| 2238 |
{ |
{ |
| 2239 |
if (g_notempty != 0) |
if (g_notempty != 0) |
| 2240 |
{ |
{ |
| 2241 |
int onechar = 1; |
int onechar = 1; |
| 2242 |
|
unsigned int obits = ((real_pcre *)re)->options; |
| 2243 |
use_offsets[0] = start_offset; |
use_offsets[0] = start_offset; |
| 2244 |
if (use_utf8) |
if ((obits & PCRE_NEWLINE_BITS) == 0) |
| 2245 |
|
{ |
| 2246 |
|
int d; |
| 2247 |
|
(void)pcre_config(PCRE_CONFIG_NEWLINE, &d); |
| 2248 |
|
obits = (d == '\r')? PCRE_NEWLINE_CR : |
| 2249 |
|
(d == '\n')? PCRE_NEWLINE_LF : |
| 2250 |
|
(d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF : |
| 2251 |
|
(d == -2)? PCRE_NEWLINE_ANYCRLF : |
| 2252 |
|
(d == -1)? PCRE_NEWLINE_ANY : 0; |
| 2253 |
|
} |
| 2254 |
|
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
| 2255 |
|
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
| 2256 |
|
&& |
| 2257 |
|
start_offset < len - 1 && |
| 2258 |
|
bptr[start_offset] == '\r' && |
| 2259 |
|
bptr[start_offset+1] == '\n') |
| 2260 |
|
onechar++; |
| 2261 |
|
else if (use_utf8) |
| 2262 |
{ |
{ |
| 2263 |
while (start_offset + onechar < len) |
while (start_offset + onechar < len) |
| 2264 |
{ |
{ |
| 2290 |
what Perl's /g options does. This turns out to be rather cunning. First |
what Perl's /g options does. This turns out to be rather cunning. First |
| 2291 |
we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the |
we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the |
| 2292 |
same point. If this fails (picked up above) we advance to the next |
same point. If this fails (picked up above) we advance to the next |
| 2293 |
character. |
character. */ |
|
|
|
|
Yet more complication arises in the case when the newline option is |
|
|
"any" and a pattern in multiline mode has to match at the start of a |
|
|
line. If a previous match was at the end of a line, and advance of one |
|
|
character just passes the \r, whereas we should prefer the longer newline |
|
|
sequence, as does the code in pcre_exec(). So we fudge it. */ |
|
| 2294 |
|
|
| 2295 |
g_notempty = 0; |
g_notempty = 0; |
|
gany_fudge = 0; |
|
| 2296 |
|
|
| 2297 |
if (use_offsets[0] == use_offsets[1]) |
if (use_offsets[0] == use_offsets[1]) |
| 2298 |
{ |
{ |
| 2299 |
if (use_offsets[0] == len) break; |
if (use_offsets[0] == len) break; |
| 2300 |
g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED; |
g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED; |
|
if ((((real_pcre *)re)->options & PCRE_STARTLINE) != 0 && |
|
|
(((real_pcre *)re)->options & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY && |
|
|
use_offsets[0] < len - 1 && |
|
|
bptr[use_offsets[0]] == '\r' && |
|
|
bptr[use_offsets[0]+1] == '\n') |
|
|
gany_fudge = 1; |
|
| 2301 |
} |
} |
| 2302 |
|
|
| 2303 |
/* For /g, update the start offset, leaving the rest alone */ |
/* For /g, update the start offset, leaving the rest alone */ |
| 2304 |
|
|
| 2305 |
if (do_g) start_offset = use_offsets[1] + gany_fudge; |
if (do_g) start_offset = use_offsets[1]; |
| 2306 |
|
|
| 2307 |
/* For /G, update the pointer and length */ |
/* For /G, update the pointer and length */ |
| 2308 |
|
|
| 2309 |
else |
else |
| 2310 |
{ |
{ |
| 2311 |
bptr += use_offsets[1] + gany_fudge; |
bptr += use_offsets[1]; |
| 2312 |
len -= use_offsets[1] + gany_fudge; |
len -= use_offsets[1]; |
| 2313 |
} |
} |
| 2314 |
} /* End of loop for /g and /G */ |
} /* End of loop for /g and /G */ |
| 2315 |
|
|