/[pcre]/code/trunk/pcregrep.c
ViewVC logotype

Diff of /code/trunk/pcregrep.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 579 by ph10, Wed Nov 24 17:39:25 2010 UTC revision 588 by ph10, Sat Jan 15 11:22:47 2011 UTC
# Line 6  Line 6 
6  its pattern matching. On a Unix or Win32 system it can recurse into  its pattern matching. On a Unix or Win32 system it can recurse into
7  directories.  directories.
8    
9             Copyright (c) 1997-2010 University of Cambridge             Copyright (c) 1997-2011 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 186  static BOOL utf8 = FALSE; Line 186  static BOOL utf8 = FALSE;
186    
187  /* Structure for options and list of them */  /* Structure for options and list of them */
188    
189  enum { OP_NODATA, OP_STRING, OP_OP_STRING, OP_NUMBER, OP_OP_NUMBER,  enum { OP_NODATA, OP_STRING, OP_OP_STRING, OP_NUMBER, OP_LONGNUMBER,
190         OP_PATLIST };         OP_OP_NUMBER, OP_PATLIST };
191    
192  typedef struct option_item {  typedef struct option_item {
193    int type;    int type;
# Line 216  used to identify them. */ Line 216  used to identify them. */
216  #define N_M_LIMIT_REC  (-14)  #define N_M_LIMIT_REC  (-14)
217    
218  static option_item optionlist[] = {  static option_item optionlist[] = {
219    { OP_NODATA,    N_NULL,   NULL,              "",              "  terminate options" },    { OP_NODATA,     N_NULL,   NULL,              "",              "  terminate options" },
220    { OP_NODATA,    N_HELP,   NULL,              "help",          "display this help and exit" },    { OP_NODATA,     N_HELP,   NULL,              "help",          "display this help and exit" },
221    { OP_NUMBER,    'A',      &after_context,    "after-context=number", "set number of following context lines" },    { OP_NUMBER,     'A',      &after_context,    "after-context=number", "set number of following context lines" },
222    { OP_NUMBER,    'B',      &before_context,   "before-context=number", "set number of prior context lines" },    { OP_NUMBER,     'B',      &before_context,   "before-context=number", "set number of prior context lines" },
223    { OP_OP_STRING, N_COLOUR, &colour_option,    "color=option",  "matched text color option" },    { OP_OP_STRING,  N_COLOUR, &colour_option,    "color=option",  "matched text color option" },
224    { OP_OP_STRING, N_COLOUR, &colour_option,    "colour=option", "matched text colour option" },    { OP_OP_STRING,  N_COLOUR, &colour_option,    "colour=option", "matched text colour option" },
225    { OP_NUMBER,    'C',      &both_context,     "context=number", "set number of context lines, before & after" },    { OP_NUMBER,     'C',      &both_context,     "context=number", "set number of context lines, before & after" },
226    { OP_NODATA,    'c',      NULL,              "count",         "print only a count of matching lines per FILE" },    { OP_NODATA,     'c',      NULL,              "count",         "print only a count of matching lines per FILE" },
227    { OP_STRING,    'D',      &DEE_option,       "devices=action","how to handle devices, FIFOs, and sockets" },    { OP_STRING,     'D',      &DEE_option,       "devices=action","how to handle devices, FIFOs, and sockets" },
228    { OP_STRING,    'd',      &dee_option,       "directories=action", "how to handle directories" },    { OP_STRING,     'd',      &dee_option,       "directories=action", "how to handle directories" },
229    { OP_PATLIST,   'e',      NULL,              "regex(p)=pattern", "specify pattern (may be used more than once)" },    { OP_PATLIST,    'e',      NULL,              "regex(p)=pattern", "specify pattern (may be used more than once)" },
230    { OP_NODATA,    'F',      NULL,              "fixed-strings", "patterns are sets of newline-separated strings" },    { OP_NODATA,     'F',      NULL,              "fixed-strings", "patterns are sets of newline-separated strings" },
231    { OP_STRING,    'f',      &pattern_filename, "file=path",     "read patterns from file" },    { OP_STRING,     'f',      &pattern_filename, "file=path",     "read patterns from file" },
232    { OP_NODATA,    N_FOFFSETS, NULL,            "file-offsets",  "output file offsets, not text" },    { OP_NODATA,     N_FOFFSETS, NULL,            "file-offsets",  "output file offsets, not text" },
233    { OP_NODATA,    'H',      NULL,              "with-filename", "force the prefixing filename on output" },    { OP_NODATA,     'H',      NULL,              "with-filename", "force the prefixing filename on output" },
234    { OP_NODATA,    'h',      NULL,              "no-filename",   "suppress the prefixing filename on output" },    { OP_NODATA,     'h',      NULL,              "no-filename",   "suppress the prefixing filename on output" },
235    { OP_NODATA,    'i',      NULL,              "ignore-case",   "ignore case distinctions" },    { OP_NODATA,     'i',      NULL,              "ignore-case",   "ignore case distinctions" },
236    { OP_NODATA,    'l',      NULL,              "files-with-matches", "print only FILE names containing matches" },    { OP_NODATA,     'l',      NULL,              "files-with-matches", "print only FILE names containing matches" },
237    { OP_NODATA,    'L',      NULL,              "files-without-match","print only FILE names not containing matches" },    { OP_NODATA,     'L',      NULL,              "files-without-match","print only FILE names not containing matches" },
238    { OP_STRING,    N_LABEL,  &stdin_name,       "label=name",    "set name for standard input" },    { OP_STRING,     N_LABEL,  &stdin_name,       "label=name",    "set name for standard input" },
239    { OP_NODATA,    N_LBUFFER, NULL,             "line-buffered", "use line buffering" },    { OP_NODATA,     N_LBUFFER, NULL,             "line-buffered", "use line buffering" },
240    { OP_NODATA,    N_LOFFSETS, NULL,            "line-offsets",  "output line numbers and offsets, not text" },    { OP_NODATA,     N_LOFFSETS, NULL,            "line-offsets",  "output line numbers and offsets, not text" },
241    { OP_STRING,    N_LOCALE, &locale,           "locale=locale", "use the named locale" },    { OP_STRING,     N_LOCALE, &locale,           "locale=locale", "use the named locale" },
242    { OP_NUMBER,    N_M_LIMIT,&match_limit,      "match-limit=number", "set PCRE match limit option" },    { OP_LONGNUMBER, N_M_LIMIT, &match_limit,     "match-limit=number", "set PCRE match limit option" },
243    { OP_NUMBER,    N_M_LIMIT_REC,&match_limit_recursion, "recursion-limit=number", "set PCRE match recursion limit option" },    { OP_LONGNUMBER, N_M_LIMIT_REC, &match_limit_recursion, "recursion-limit=number", "set PCRE match recursion limit option" },
244    { OP_NODATA,    'M',      NULL,              "multiline",     "run in multiline mode" },    { OP_NODATA,     'M',      NULL,              "multiline",     "run in multiline mode" },
245    { OP_STRING,    'N',      &newline,          "newline=type",  "set newline type (CR, LF, CRLF, ANYCRLF or ANY)" },    { OP_STRING,     'N',      &newline,          "newline=type",  "set newline type (CR, LF, CRLF, ANYCRLF or ANY)" },
246    { OP_NODATA,    'n',      NULL,              "line-number",   "print line number with output lines" },    { OP_NODATA,     'n',      NULL,              "line-number",   "print line number with output lines" },
247    { OP_OP_NUMBER, 'o',      &only_matching,    "only-matching=n", "show only the part of the line that matched" },    { OP_OP_NUMBER,  'o',      &only_matching,    "only-matching=n", "show only the part of the line that matched" },
248    { OP_NODATA,    'q',      NULL,              "quiet",         "suppress output, just set return code" },    { OP_NODATA,     'q',      NULL,              "quiet",         "suppress output, just set return code" },
249    { OP_NODATA,    'r',      NULL,              "recursive",     "recursively scan sub-directories" },    { OP_NODATA,     'r',      NULL,              "recursive",     "recursively scan sub-directories" },
250    { OP_STRING,    N_EXCLUDE,&exclude_pattern,  "exclude=pattern","exclude matching files when recursing" },    { OP_STRING,     N_EXCLUDE,&exclude_pattern,  "exclude=pattern","exclude matching files when recursing" },
251    { OP_STRING,    N_INCLUDE,&include_pattern,  "include=pattern","include matching files when recursing" },    { OP_STRING,     N_INCLUDE,&include_pattern,  "include=pattern","include matching files when recursing" },
252    { OP_STRING,    N_EXCLUDE_DIR,&exclude_dir_pattern, "exclude-dir=pattern","exclude matching directories when recursing" },    { OP_STRING,     N_EXCLUDE_DIR,&exclude_dir_pattern, "exclude-dir=pattern","exclude matching directories when recursing" },
253    { OP_STRING,    N_INCLUDE_DIR,&include_dir_pattern, "include-dir=pattern","include matching directories when recursing" },    { OP_STRING,     N_INCLUDE_DIR,&include_dir_pattern, "include-dir=pattern","include matching directories when recursing" },
254    
255    /* These two were accidentally implemented with underscores instead of    /* These two were accidentally implemented with underscores instead of
256    hyphens in the option names. As this was not discovered for several releases,    hyphens in the option names. As this was not discovered for several releases,
# Line 296  const char utf8_table4[] = { Line 296  const char utf8_table4[] = {
296    
297    
298  /*************************************************  /*************************************************
299    *         Exit from the program                  *
300    *************************************************/
301    
302    /* If there has been a resource error, give a suitable message.
303    
304    Argument:  the return code
305    Returns:   does not return
306    */
307    
308    static void
309    pcregrep_exit(int rc)
310    {
311    if (resource_error)
312      {
313      fprintf(stderr, "pcregrep: Error %d or %d means that a resource limit "
314        "was exceeded.\n", PCRE_ERROR_MATCHLIMIT, PCRE_ERROR_RECURSIONLIMIT);
315      fprintf(stderr, "pcregrep: Check your regex for nested unlimited loops.\n");
316      }
317    
318    exit(rc);
319    }
320    
321    
322    /*************************************************
323  *            OS-specific functions               *  *            OS-specific functions               *
324  *************************************************/  *************************************************/
325    
# Line 565  return sys_errlist[n]; Line 589  return sys_errlist[n];
589    
590    
591  /*************************************************  /*************************************************
 *         Exit from the program                  *  
 *************************************************/  
   
 /* If there has been a resource error, give a suitable message.  
   
 Argument:  the return code  
 Returns:   does not return  
 */  
   
 static void  
 pcregrep_exit(int rc)  
 {  
 if (resource_error)  
   {  
   fprintf(stderr, "pcregrep: Error %d or %d means that a resource limit "  
     "was exceeded.\n", PCRE_ERROR_MATCHLIMIT, PCRE_ERROR_RECURSIONLIMIT);  
   fprintf(stderr, "pcregrep: Check your regex for nested unlimited loops.\n");  
   }  
   
 exit(rc);  
 }  
   
   
   
 /*************************************************  
592  *            Read one line of input              *  *            Read one line of input              *
593  *************************************************/  *************************************************/
594    
# Line 635  Arguments: Line 634  Arguments:
634    endptr    end of available data    endptr    end of available data
635    lenptr    where to put the length of the eol sequence    lenptr    where to put the length of the eol sequence
636    
637  Returns:    pointer to the last byte of the line  Returns:    pointer to the last byte of the line, including the newline byte(s)
638  */  */
639    
640  static char *  static char *
# Line 1317  while (ptr < endptr) Line 1316  while (ptr < endptr)
1316        (invert not set). Because the PCRE_FIRSTLINE option is set, the start of        (invert not set). Because the PCRE_FIRSTLINE option is set, the start of
1317        the match will always be before the first newline sequence. */        the match will always be before the first newline sequence. */
1318    
1319        if (multiline)        if (multiline & !invert)
1320          {          {
1321          int ellength;          char *endmatch = ptr + offsets[1];
1322          char *endmatch = ptr;          t = ptr;
1323          if (!invert)          while (t < endmatch)
1324            {            {
1325            endmatch += offsets[1];            t = end_of_line(t, endptr, &endlinelength);
1326            t = ptr;            if (t < endmatch) linenumber++; else break;
           while (t < endmatch)  
             {  
             t = end_of_line(t, endptr, &ellength);  
             if (t <= endmatch) linenumber++; else break;  
             }  
1327            }            }
1328          endmatch = end_of_line(endmatch, endptr, &ellength);          linelength = t - ptr - endlinelength;
         linelength = endmatch - ptr - ellength;  
1329          }          }
1330    
1331        /*** NOTE: Use only fwrite() to output the data line, so that binary        /*** NOTE: Use only fwrite() to output the data line, so that binary
# Line 1355  while (ptr < endptr) Line 1348  while (ptr < endptr)
1348  #endif  #endif
1349    
1350        /* We have to split the line(s) up if colouring, and search for further        /* We have to split the line(s) up if colouring, and search for further
1351        matches. */        matches, but not of course if the line is a non-match. */
1352    
1353        if (do_colour)        if (do_colour && !invert)
1354          {          {
1355            int plength;
1356          int last_offset = 0;          int last_offset = 0;
1357          FWRITE(ptr, 1, offsets[0], stdout);          FWRITE(ptr, 1, offsets[0], stdout);
1358          fprintf(stdout, "%c[%sm", 0x1b, colour_string);          fprintf(stdout, "%c[%sm", 0x1b, colour_string);
# Line 1369  while (ptr < endptr) Line 1363  while (ptr < endptr)
1363            last_offset += offsets[1];            last_offset += offsets[1];
1364            matchptr += offsets[1];            matchptr += offsets[1];
1365            length -= offsets[1];            length -= offsets[1];
1366            if (!match_patterns(matchptr, length, offsets, &mrc)) break;            if (last_offset >= linelength + endlinelength ||
1367                  !match_patterns(matchptr, length, offsets, &mrc)) break;
1368            FWRITE(matchptr, 1, offsets[0], stdout);            FWRITE(matchptr, 1, offsets[0], stdout);
1369            fprintf(stdout, "%c[%sm", 0x1b, colour_string);            fprintf(stdout, "%c[%sm", 0x1b, colour_string);
1370            FWRITE(matchptr + offsets[0], 1, offsets[1] - offsets[0], stdout);            FWRITE(matchptr + offsets[0], 1, offsets[1] - offsets[0], stdout);
1371            fprintf(stdout, "%c[00m", 0x1b);            fprintf(stdout, "%c[00m", 0x1b);
1372            }            }
1373          FWRITE(ptr + last_offset, 1,  
1374            (linelength + endlinelength) - last_offset, stdout);          /* In multiline mode, we may have already printed the complete line
1375            and its line-ending characters (if they matched the pattern), so there
1376            may be no more to print. */
1377    
1378            plength = (linelength + endlinelength) - last_offset;
1379            if (plength > 0)
1380              FWRITE(ptr + last_offset, 1, plength, stdout);
1381          }          }
1382    
1383        /* Not colouring; no need to search for further matches */        /* Not colouring; no need to search for further matches */
# Line 2278  for (i = 1; i < argc; i++) Line 2279  for (i = 1; i < argc; i++)
2279    
2280    /* Otherwise, deal with single string or numeric data values. */    /* Otherwise, deal with single string or numeric data values. */
2281    
2282    else if (op->type != OP_NUMBER && op->type != OP_OP_NUMBER)    else if (op->type != OP_NUMBER && op->type != OP_LONGNUMBER &&
2283               op->type != OP_OP_NUMBER)
2284      {      {
2285      *((char **)op->dataptr) = option_data;      *((char **)op->dataptr) = option_data;
2286      }      }
# Line 2308  for (i = 1; i < argc; i++) Line 2310  for (i = 1; i < argc; i++)
2310            option_data, op->one_char);            option_data, op->one_char);
2311        pcregrep_exit(usage(2));        pcregrep_exit(usage(2));
2312        }        }
2313      *((int *)op->dataptr) = n;      if (op->type == OP_LONGNUMBER)
2314            *((unsigned long int *)op->dataptr) = n;
2315        else
2316            *((int *)op->dataptr) = n;
2317      }      }
2318    }    }
2319    
Legend:
Removed from v.579  
changed lines
  Added in v.588
webmaster@exim.org
 ViewVC Help
Powered by ViewVC 1.1.12