Add -o option

* src/grep.c: Add --only-matching (-o) switch (see NEWS)
* doc/grep.texi, doc/grep.1, NEWS: Document changes

17 files changed, 66 insertions, 14 deletions

diff --git a/ChangeLog b/ChangeLog
index 5ec7d498..07b2ecd5 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2001-11-19  Bernhard Rosenkraenzer <bero@redhat.com>
+	* src/grep.c: Add --only-matching (-o) switch (see NEWS)
+	* doc/grep.texi, doc/grep.1, NEWS: Document changes
+
 2001-09-25  Bernhard Rosenkraenzer <bero@redhat.com>
 	* configure.in: Detect pcre correctly even when it's in
 	  non-standard locations, using pcre-config
diff --git a/NEWS b/NEWS
index 7f17da9c..703c94f4 100644
--- a/NEWS
+++ b/NEWS
@@ -1,4 +1,8 @@
 Version 2.5
+  - The new option --only-matching (-o) will print only the part of matching
+    lines that matches the pattern. This is useful, for example, to extract
+    IP addresses from log files.
+
   - i18n bug fixed ([A-Z0-9] wouldn't match A in locales other than C on
     systems using recent glibc builds
 
diff --git a/doc/grep.1 b/doc/grep.1
index 29b2cf9c..3edc0ab4 100644
--- a/doc/grep.1
+++ b/doc/grep.1
@@ -12,7 +12,7 @@
 .de Id
 .ds Dt \\$4
 ..
-.Id $Id: grep.1,v 1.21 2001/09/25 10:00:16 bero Exp $
+.Id $Id: grep.1,v 1.22 2001/11/19 13:23:29 bero Exp $
 .TH GREP 1 \*(Dt "GNU Project"
 .SH NAME
 grep, egrep, fgrep \- print lines matching a pattern
@@ -292,6 +292,10 @@ is operating, or if an I/O error occurs.
 Prefix each line of output with the line number
 within its input file.
 .TP
+.BR \-o ", " \-\^\-only-matching
+Show only the part of a matching line that matches
+.I PATTERN.
+.TP
 .BR \-\^\-line-buffering
 Use line buffering, it can be a performance penality.
 .TP
diff --git a/doc/grep.texi b/doc/grep.texi
index 70021cc3..4b39a94a 100644
--- a/doc/grep.texi
+++ b/doc/grep.texi
@@ -185,6 +185,13 @@ The scanning of every file will stop on the first match.
 @cindex line numbering
 Prefix each line of output with the line number within its input file.
 
+@item -o
+@itemx --only-matching
+@opindex -o
+@opindex --only-matching
+@cindex only matching
+Print only the part of matching lines that actually matches @var{pattern}.
+
 @item -q
 @itemx --quiet
 @itemx --silent
diff --git a/po/grep.pot b/po/grep.pot
index aec90e45..7298c5b3 100644
--- a/po/grep.pot
+++ b/po/grep.pot
@@ -6,7 +6,7 @@
 msgid ""
 msgstr ""
 "Project-Id-Version: PACKAGE VERSION\n"
-"POT-Creation-Date: 2001-09-26 18:24+0200\n"
+"POT-Creation-Date: 2001-09-26 18:44+0200\n"
 "PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
 "Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
 "Language-Team: LANGUAGE <LL@li.org>\n"
diff --git a/src/grep.c b/src/grep.c
index 558f5f1e..ac1eb797 100644
--- a/src/grep.c
+++ b/src/grep.c
@@ -69,6 +69,9 @@ static int mmap_option;
 /* If nonzero, use grep_color marker.  */
 static int color_option;
 
+/* If nonzero, show only the part of a line matching the expression. */
+static int only_matching;
+
 /* The color string used.  The user can overwrite it using the environment
    variable GREP_COLOR.  The default is to print red.  */
 static const char *grep_color = "01;31";
@@ -77,7 +80,7 @@ static struct exclude *excluded_patterns;
 static struct exclude *included_patterns;
 /* Short options.  */
 static char const short_options[] =
-"0123456789A:B:C:D:EFGHIPUVX:abcd:e:f:hiKLlm:nqRrsuvwxyZz";
+"0123456789A:B:C:D:EFGHIPUVX:abcd:e:f:hiKLlm:noqRrsuvwxyZz";
 
 /* Non-boolean long options that have no corresponding short equivalents.  */
 enum
@@ -124,6 +127,7 @@ static struct option const long_options[] =
   {"no-messages", no_argument, NULL, 's'},
   {"null", no_argument, NULL, 'Z'},
   {"null-data", no_argument, NULL, 'z'},
+  {"only-matching", no_argument, NULL, 'o'},
   {"perl-regexp", no_argument, NULL, 'P'},
   {"quiet", no_argument, NULL, 'q'},
   {"recursive", no_argument, NULL, 'r'},
@@ -518,6 +522,28 @@ prline (char const *beg, char const *lim, int sep)
 #endif
       print_offset_sep (pos, sep);
     }
+  if (only_matching)
+    {
+      size_t match_size;
+      size_t match_offset;
+      while ((match_offset = (*execute) (beg, lim - beg, &match_size, 1))
+	  != (size_t) -1)
+        {
+	  char const *b = beg + match_offset;
+	  if (b == lim)
+	    break;
+	  if(color_option)
+	    printf("\33[%sm", grep_color);
+	  fwrite(b, sizeof (char), match_size, stdout);
+	  if(color_option)
+	    fputs("\33[00m", stdout);
+	  fputs("\n", stdout);
+	  beg = b + match_size;
+        }
+      if(line_buffered)
+	fflush(stdout);
+      return;
+    }
   if (color_option)
     {
       size_t match_size;
@@ -1021,6 +1047,7 @@ Output control:\n\
       --line-buffered       flush output on every line\n\
   -H, --with-filename       print the filename for each match\n\
   -h, --no-filename         suppress the prefixing filename on output\n\
+  -o, --only-matching       show only the part of a line matching PATTERN\n\
   -q, --quiet, --silent     suppress all normal output\n\
       --binary-files=TYPE   assume that binary files are TYPE\n\
                             TYPE is 'binary', 'text', or 'without-match'\n\
@@ -1263,6 +1290,8 @@ main (int argc, char **argv)
   out_after = out_before = -1;
   /* Default before/after context: chaged by -C/-NUM options */
   default_context = 0;
+  /* Changed by -o option */
+  only_matching = 0;
 
   /* Internationalization. */
 #if defined(HAVE_SETLOCALE)
@@ -1443,6 +1472,10 @@ main (int argc, char **argv)
 	out_line = 1;
 	break;
 
+      case 'o':
+	only_matching = 1;
+	break;
+
       case 'q':
 	exit_on_match = 1;
 	break;
diff --git a/tests/backref.sh b/tests/backref.sh
index b87ac853..83f6dcc9 100755
--- a/tests/backref.sh
+++ b/tests/backref.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Test that backrefs are local to regex.
 #
 #
diff --git a/tests/bre.sh b/tests/bre.sh
index eecdd3ee..e7ccf0a0 100644
--- a/tests/bre.sh
+++ b/tests/bre.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Regression test for GNU grep.
 
 : ${srcdir=.}
diff --git a/tests/empty.sh b/tests/empty.sh
index cae640f9..d4225fec 100644
--- a/tests/empty.sh
+++ b/tests/empty.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # test that the empty file means no pattern
 # and an empty pattern means match all.
 
diff --git a/tests/ere.sh b/tests/ere.sh
index 5bf16952..d57e2fae 100644
--- a/tests/ere.sh
+++ b/tests/ere.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Regression test for GNU grep.
 
 : ${srcdir=.}
diff --git a/tests/file.sh b/tests/file.sh
index c3979790..3db95bed 100755
--- a/tests/file.sh
+++ b/tests/file.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Test for POSIX.2 options for grep
 #
 # grep -E -f pattern_file file
diff --git a/tests/khadafy.sh b/tests/khadafy.sh
index 141b3d81..e3d2cde4 100644
--- a/tests/khadafy.sh
+++ b/tests/khadafy.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Regression test for GNU grep.
 
 : ${srcdir=.}
diff --git a/tests/options.sh b/tests/options.sh
index 34150185..0f185297 100644
--- a/tests/options.sh
+++ b/tests/options.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Test for POSIX.2 options for grep
 #
 # grep [ -E| -F][ -c| -l| -q ][-insvx] -e pattern_list 
diff --git a/tests/spencer1.sh b/tests/spencer1.sh
index f09b3fa9..2391761a 100644
--- a/tests/spencer1.sh
+++ b/tests/spencer1.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Regression test for GNU grep.
 
 : ${srcdir=.}
diff --git a/tests/spencer2.sh b/tests/spencer2.sh
index 19caa8de..a26869c5 100644
--- a/tests/spencer2.sh
+++ b/tests/spencer2.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Regression test for GNU grep.
 
 : ${srcdir=.}
diff --git a/tests/status.sh b/tests/status.sh
index 4f08b8a3..94f59294 100644
--- a/tests/status.sh
+++ b/tests/status.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 # Test for status code for GNU grep.
 # status code
 #  0 match found
diff --git a/tests/warning.sh b/tests/warning.sh
index d2dc6d5a..bfca533a 100644
--- a/tests/warning.sh
+++ b/tests/warning.sh
@@ -1,4 +1,4 @@
-#! /bin/sh
+#!/bin/sh
 #
 # Tell them not to be alarmed.