src/version.c, src/readname.c: New files

src/word-split.c, src/html-strip.l: Add support for --from-file
(-T) option. 
Makefile.am (libswis.a): New goal
swis.h: Add new prototypes
gnulib.modules: Add obstack


git-svn-id: file:///svnroot/swis/trunk@13 05ba3e8d-823b-0410-8fb2-de0ee4edb5ba

8 files changed, 235 insertions, 13 deletions

diff --git a/ChangeLog b/ChangeLog
index 5059bff..2da714c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,13 @@
 2007-10-04  Sergey Poznyakoff  <gray@gnu.org.ua>
 
+	* src/version.c, src/readname.c: New files
+	* src/word-split.c, src/html-strip.l: Add support for --from-file
+	(-T) option. 
+	* Makefile.am (libswis.a): New goal
+	* swis.h: Add new prototypes
+	* gnulib.modules: Add obstack
+
+
 	* src/Makefile.am (LDADD): Add LIBICONV
 	* configure.ac: Minor change
 	* gnulib.modules: Add iconv
diff --git a/gnulib.modules b/gnulib.modules
index c8f9eaf..ca43b92 100644
--- a/gnulib.modules
+++ b/gnulib.modules
@@ -14,6 +14,8 @@ progname
 getline
 strerror
 iconv
+fnmatch
+obstack
 mbchar
 mbfile
 mbiter
diff --git a/src/Makefile.am b/src/Makefile.am
index 71f7ab8..9b46d15 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -14,11 +14,16 @@
 # You should have received a copy of the GNU General Public License
 # along with SWIS.  If not, see <http://www.gnu.org/licenses/>.
 
+noinst_LIBRARIES=libswis.a
 bin_PROGRAMS=html-strip word-split
 html_strip_SOURCES=html-strip.l
 word_split_SOURCES=word-split.c
 
+libswis_a_SOURCES=\
+ readname.c\
+ version.c
+
 INCLUDES=-I${top_srcdir}/gnu -I../gnu
-LDADD=../gnu/libgnu.a $(LIBICONV)
+LDADD=./libswis.a ../gnu/libgnu.a $(LIBICONV)
 EXTRA_DIST=swis.h
 AM_LFLAGS=-d
diff --git a/src/html-strip.l b/src/html-strip.l
index f337c5d..a2dded3 100644
--- a/src/html-strip.l
+++ b/src/html-strip.l
@@ -241,11 +241,11 @@ struct option options[] = {
   { "version", no_argument, NULL, 'v' },
   { "output", required_argument, NULL, 'o' },
   { "tag", no_argument, NULL, 't' },
+  { "files-from", required_argument, NULL, 'T' },
+  { "null", no_argument, NULL, '0' },
   { NULL }
 };
 
-DECL_COPYRIGHT;
-
 void
 usage ()
 {
@@ -255,6 +255,8 @@ usage ()
   printf ("  -d, --debug                    output debugging info\n");
   printf ("  -o, --output=FILE              direct output to FILE instead of stdout\n");
   printf ("  -t, --tag                      tag each output block with the source file name");
+  printf ("  -T, --from-file=FILE           read input file names from FILE\n");
+  printf ("  -0, --null                     -T reads null-terminated names\n");
   printf ("\n");
   printf ("  -h, --help                     print this help list\n");
   printf ("  -v, --version                  print program version and exit\n");
@@ -270,10 +272,14 @@ main (int argc, char **argv)
   program_name = argv[0];
   yy_flex_debug = 0;
 
-  while ((c = getopt_long (argc, argv, "dho:tv", options, NULL)) != EOF)
+  while ((c = getopt_long (argc, argv, "0dhoT::tv", options, NULL)) != EOF)
     {
       switch (c)
 	{
+	case '0':
+	  filename_terminator = 0;
+	  break;
+	  
 	case 'd':
 	  yy_flex_debug = 1;
 	  break;
@@ -292,13 +298,16 @@ main (int argc, char **argv)
 	    error (1, errno, "cannot open output file %s", optarg);
 	  break;
 
+	case 'T':
+	  read_names_from_file (optarg);
+	  break;
+	  
 	case 't':
 	  tag_option = 1;
 	  break;
 	  
 	case 'v':
-	  version_etc (stdout, "html-strip", PACKAGE_NAME, VERSION, 
-		       PACKAGE_AUTHOR, NULL);
+	  swis_version (stdout, "html-strip");
 	  exit (0);
 	  
 	default:
@@ -309,6 +318,8 @@ main (int argc, char **argv)
   argc -= optind;
   argv += optind;
 
+  update_argcv (&argc, &argv);
+  
   if (argc)
     {
       input_file = argv;
diff --git a/src/readname.c b/src/readname.c
new file mode 100644
index 0000000..def30b6
--- /dev/null
+++ b/src/readname.c
@@ -0,0 +1,154 @@
+/* This file is part of SWIS
+   Copyright (C) 2007 Sergey Poznyakoff
+ 
+   SWIS is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3, or (at your option)
+   any later version.
+
+   SWIS is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with SWIS.  If not, see <http://www.gnu.org/licenses/>. */
+
+#include "swis.h"
+
+int filename_terminator = '\n';
+
+enum read_file_list_state  /* Result of reading file name from the list file */
+  {
+    file_list_success,     /* OK, name read successfully */
+    file_list_end,         /* End of list file */
+    file_list_zero,        /* Zero separator encountered where it should not */
+    file_list_skip         /* Empty (zero-length) entry encountered, skip it */
+  };
+
+/* Read from FP a sequence of characters up to FILENAME_TERMINATOR and put them
+   into STK.
+ */
+static enum read_file_list_state
+read_name_from_file (FILE *fp, struct obstack *stk)
+{
+  int c;
+  size_t counter = 0;
+
+  for (c = getc (fp); c != EOF && c != filename_terminator; c = getc (fp))
+    {
+      if (c == 0)
+	{
+	  /* We have read a zero separator. The file possibly is
+	     zero-separated */
+	  return file_list_zero;
+	}
+      obstack_1grow (stk, c);
+      counter++;
+    }
+
+  if (counter == 0 && c != EOF)
+    return file_list_skip;
+
+  obstack_1grow (stk, 0);
+
+  return (counter == 0 && c == EOF) ? file_list_end : file_list_success;
+}
+
+struct obstack argv_stk;
+int stk_init;
+size_t argcount;
+
+void
+read_names_from_file (const char *filename)
+{
+  FILE *fp;
+  size_t count = 0, i;
+  enum read_file_list_state read_state;
+  
+  if (!strcmp (filename, "-"))
+    {
+      //      request_stdin ("-T");
+      fp = stdin;
+    }
+  else
+    {
+      if ((fp = fopen (filename, "r")) == NULL)
+	error (1, errno, "cannot open file %s", filename);
+    }
+
+  if (!stk_init)
+    {
+      stk_init = 1;
+      obstack_init (&argv_stk);
+    }
+  
+  while ((read_state = read_name_from_file (fp, &argv_stk)) != file_list_end)
+    {
+      switch (read_state)
+	{
+	case file_list_success:
+	  argcount++;
+	  break;
+
+	case file_list_end: /* won't happen, just to pacify gcc */
+	  break;
+
+	case file_list_zero:
+	  {
+	    size_t size;
+	    char *p;
+	    
+	    error (0, 0, "%s: file name read contains nul character",
+		   filename);
+
+	    /* Prepare new stack contents */
+	    /* FIXME */
+	    size = obstack_object_size (&argv_stk);
+	    p = obstack_finish (&argv_stk);
+	    for (i = 0; size > 0; size--, p++)
+	      {
+		if (*p)
+		  obstack_1grow (&argv_stk, *p);
+		else if (i < argcount)
+		  i++;
+		else
+		  obstack_1grow (&argv_stk, '\n');
+	      }
+	    obstack_1grow (&argv_stk, 0);
+	    count = 1;
+	    /* Read rest of files using new filename terminator */
+	    filename_terminator = 0;
+	    break;
+	  }
+
+	case file_list_skip:
+	  break;
+	}
+    }
+
+  fclose (fp);
+  argcount += count;
+}
+
+void
+update_argcv (int *pargc, char ***pargv)
+{
+  int i, new_argc;
+  char **new_argv;
+  char *p, *start;
+  
+  new_argc = *pargc + argcount + 1;
+  new_argv = xmalloc (sizeof (new_argv[0]) * (new_argc + 1));
+  memcpy (new_argv, *pargv, sizeof (new_argv[0]) * *pargc);
+
+  start = obstack_finish (&argv_stk);
+  for (i = *pargc, p = start; *p; p += strlen (p) + 1, i++)
+    new_argv[i] = xstrdup (p);
+  new_argv[i] = NULL;
+
+  obstack_free (&argv_stk, NULL);
+
+  *pargv = new_argv;
+  *pargc = new_argc;
+}
diff --git a/src/swis.h b/src/swis.h
index 9adb00c..b16c0a4 100644
--- a/src/swis.h
+++ b/src/swis.h
@@ -39,7 +39,13 @@
 #endif
 #include "version-etc.h"
 
+#define obstack_chunk_alloc malloc
+#define obstack_chunk_free free
+#include "obstack.h"
+
 #define PACKAGE_AUTHOR "Sergey Poznyakoff"
-#define COPYRIGHT_STRING "Copyright %s %d " PACKAGE_AUTHOR
 
-#define DECL_COPYRIGHT const char version_etc_copyright[] = COPYRIGHT_STRING
+void swis_version (FILE *fp, const char *progname);
+void read_names_from_file (const char *filename);
+void update_argcv (int *pargc, char ***pargv);
+extern int filename_terminator;
diff --git a/src/version.c b/src/version.c
new file mode 100644
index 0000000..9510a33
--- /dev/null
+++ b/src/version.c
@@ -0,0 +1,25 @@
+/* This file is part of SWIS
+   Copyright (C) 2007 Sergey Poznyakoff
+ 
+   SWIS is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3, or (at your option)
+   any later version.
+
+   SWIS is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with SWIS.  If not, see <http://www.gnu.org/licenses/>. */
+
+#include "swis.h"
+
+const char version_etc_copyright[] = "Copyright %s %d " PACKAGE_AUTHOR;
+
+void
+swis_version (FILE *fp, const char *progname)
+{
+  version_etc (fp, progname, PACKAGE_NAME, VERSION, PACKAGE_AUTHOR, NULL);
+}
diff --git a/src/word-split.c b/src/word-split.c
index 59b11d6..b283223 100644
--- a/src/word-split.c
+++ b/src/word-split.c
@@ -26,11 +26,11 @@ struct option options[] = {
   { "help", no_argument, NULL, 'h' },
   { "version", no_argument, NULL, 'v' },
   { "output", required_argument, NULL, 'o' },
+  { "files-from", required_argument, NULL, 'T' },
+  { "null", no_argument, NULL, '0' },
   { NULL }
 };
 
-DECL_COPYRIGHT;
-
 void
 usage ()
 {
@@ -40,6 +40,8 @@ usage ()
   /*  printf ("  -d, --debug                    output debugging info\n"); */
   printf ("  -o, --output=FILE              direct output to FILE instead of stdout\n");
   printf ("  -t, --tag                      preserve file name tags\n");
+  printf ("  -T, --from-file=FILE           read input file names from FILE\n");
+  printf ("  -0, --null                     -T reads null-terminated names\n");
   printf ("\n");
   printf ("  -h, --help                     print this help list\n");
   printf ("  -v, --version                  print program version and exit\n");
@@ -153,10 +155,14 @@ main (int argc, char **argv)
   
   program_name = argv[0];
 
-  while ((c = getopt_long (argc, argv, "dho:tv", options, NULL)) != EOF)
+  while ((c = getopt_long (argc, argv, "0dho:T:tv", options, NULL)) != EOF)
     {
       switch (c)
 	{
+	case '0':
+	  filename_terminator = 0;
+	  break;
+	  
 	case 'd':
 	  error (0, 0, "warning: the --debug option is not yet supported");
 	  /* FIXME */
@@ -176,13 +182,16 @@ main (int argc, char **argv)
 	    error (1, errno, "cannot open output file %s", optarg);
 	  break;
 
+	case 'T':
+	  read_names_from_file (optarg);
+	  break;
+	  
 	case 't':
 	  tag_option = 1;
 	  break;
 	  
 	case 'v':
-	  version_etc (stdout, "word-split", PACKAGE_NAME, VERSION, 
-		       PACKAGE_AUTHOR, NULL);
+	  swis_version (stdout, "word-split");
 	  exit (0);
 	  
 	default:
@@ -193,6 +202,8 @@ main (int argc, char **argv)
   argc -= optind;
   argv += optind;
 
+  update_argcv (&argc, &argv);
+
   if (argc)
     {
       input_file = argv;