PERFORCE change 143422 for review

[Gabor Kovesdan]

http://perforce.freebsd.org/chv.cgi?CH=143422

Change 143422 by gabor at gabor_server on 2008/06/13 17:51:54

	- Check in some work-in-progress code for -J

Affected files ...

.. //depot/projects/soc2008/gabor_textproc/grep/Makefile#3 edit
.. //depot/projects/soc2008/gabor_textproc/grep/binary.c#5 edit
.. //depot/projects/soc2008/gabor_textproc/grep/file.c#4 edit
.. //depot/projects/soc2008/gabor_textproc/grep/grep.c#16 edit
.. //depot/projects/soc2008/gabor_textproc/grep/grep.h#12 edit

Differences ...

==== //depot/projects/soc2008/gabor_textproc/grep/Makefile#3 (text+ko) ====

@@ -15,9 +15,9 @@
 	grep.1 zegrep.1 \
 	grep.1 zfgrep.1
 
-CFLAGS+= -std=c99 -Wall -pedantic
+CFLAGS+= -std=c99 -Wall -pedantic -ggdb
 
-LDADD=	-lz
-DPADD=	${LIBZ}
+LDADD=	-lz -lbz2
+DPADD=	${LIBZ} ${LIBBZ2}
 
 .include <bsd.prog.mk>

==== //depot/projects/soc2008/gabor_textproc/grep/binary.c#5 (text+ko) ====

@@ -34,6 +34,7 @@
 #endif
 #endif /* not lint */
 
+#include <bzlib.h>
 #include <ctype.h>
 #include <err.h>
 #include <stdio.h>

==== //depot/projects/soc2008/gabor_textproc/grep/file.c#4 (text+ko) ====

@@ -36,6 +36,7 @@
 
 #include <sys/param.h>
 
+#include <bzlib.h>
 #include <err.h>
 #include <stdio.h>
 #include <stdlib.h>
@@ -51,6 +52,7 @@
 #define FILE_STDIO	0
 #define FILE_MMAP	1
 #define FILE_GZIP	2
+#define FILE_BZIP	3
 
 struct file {
 	int		 type;
@@ -58,9 +60,27 @@
 	FILE		*f;
 	struct mmfile	*mmf;
 	gzFile		*gzf;
+	BZFILE		*bzf;
 };
 
 static char *
+bzfgetln(BZFILE *f, size_t *len)
+{
+	int	bzerror;
+
+	if (lnbuflen != *len)
+	{
+		lnbuflen = *len;
+		lnbuf = grep_realloc(lnbuf, ++lnbuflen);
+	}
+
+	if (BZ2_bzRead(&bzerror, f, lnbuf, *len) > 0)
+		return (lnbuf);
+	else
+		return (NULL);
+}
+
+static char *
 gzfgetln(gzFile *f, size_t *len)
 {
 	size_t	 n;
@@ -113,6 +133,11 @@
 		f->noseek = lseek(fd, 0L, SEEK_SET) == -1;
 		if ((f->gzf = gzdopen(fd, mode)) != NULL)
 			return (f);
+	} else if (Jflag) {
+		f->type = FILE_BZIP;
+		f->noseek = lseek(fd, 0L, SEEK_SET) == -1;
+		if ((f->bzf = BZ2_bzdopen(fd, mode)) != NULL)
+			return (f);
 	} else
 	{
 		f->type = FILE_STDIO;
@@ -139,8 +164,11 @@
 		f->type = FILE_GZIP;
 		if ((f->gzf = gzopen(fname, mode)) != NULL)
 			return (f);
-	} else
-	{
+	} else if (Jflag) {
+		f->type = FILE_BZIP;
+		if ((f->bzf = BZ2_bzopen(fname, mode)) != NULL)
+			return (f);
+	} else {
 		/* try mmap first; if it fails, try stdio */
 		if ((f->mmf = mmopen(fname, mode)) != NULL) {
 			f->type = FILE_MMAP;
@@ -168,6 +196,10 @@
 		return (mmbin_file(f->mmf));
 	case FILE_GZIP:
 		return (gzbin_file(f->gzf));
+	case FILE_BZIP:
+/*XXX
+		return (bzbin_file(f->bzf)); */
+		return (0);
 	default:
 		/* NOTREACHED */
 		errx(2, "invalid file type");
@@ -184,6 +216,8 @@
 		return (mmfgetln(f->mmf, l));
 	case FILE_GZIP:
 		return (gzfgetln(f->gzf, l));
+	case FILE_BZIP:
+		return (bzfgetln(f->bzf, l));
 	default:
 		/* NOTREACHED */
 		errx(2, "invalid file type");
@@ -203,6 +237,9 @@
 	case FILE_GZIP:
 		gzclose(f->gzf);
 		break;
+	case FILE_BZIP:
+		BZ2_bzclose(f->bzf);
+		break;
 	default:
 		/* NOTREACHED */
 		errx(2, "invalid file type");

==== //depot/projects/soc2008/gabor_textproc/grep/grep.c#16 (text+ko) ====

@@ -72,12 +72,13 @@
 int	 Fflag;		/* -F: interpret pattern as list of fixed strings */
 int	 Gflag;		/* -G: interpret pattern as basic regexp */
 int	 Hflag;		/* -H: always print file name */
+int	 Jflag;		/* -J: grep in bzipped file */
 int	 Lflag;		/* -L: only show names of files with no matches */
 int	 Oflag;		/* -O: if -R, follow explicitly listed symlinks */
 int	 Pflag;		/* -P: if -R, no symlinks are followed */
 int	 Rflag;		/* -R: recursively search directory trees */
 int	 Sflag;		/* -S: if -R, follow all symlinks */
-int	 Zflag;		/* -Z: decompress input before processing */
+int	 Zflag;		/* -Z: grep in gzipped file */
 int	 bflag;		/* -b: show block numbers for each match */
 int	 cflag;		/* -c: only show a count of matching lines */
 int	 hflag;		/* -h: don't print filename headers */
@@ -123,7 +124,7 @@
 usage(void)
 {
 	fprintf(stderr,
-	    "usage: %s [-abcDEFGHhIiLlnOoPqRSsUVvwxZ] [-A num] [-B num] [-C[num]]\n"
+	    "usage: %s [-abcDEFGHhIiJLlnOoPqRSsUVvwxZ] [-A num] [-B num] [-C[num]]\n"
 	    "\t[-e pattern] [-f file] [--binary-files=value] [--color=when]\n"
 	    "\t[--context[=num]] [--directories=action] [--label] [--line-buffered]\n"
 	    "\t[--null] [pattern] [file ...]\n"
@@ -131,7 +132,7 @@
 	exit(2);
 }
 
-static char	*optstr = "0123456789A:B:CD:EFGHILOPSRUVZabcd:e:f:hilnoqrsuvwxy";
+static char	*optstr = "0123456789A:B:CD:EFGHIJLOPSRUVZabcd:e:f:hilnoqrsuvwxy";
 
 struct option long_options[] =
 {
@@ -159,8 +160,7 @@
 	{"no-filename",		no_argument,		NULL, 'h'},
 	{"with-filename",	no_argument,		NULL, 'H'},
 	{"ignore-case",		no_argument,		NULL, 'i'},
-/* XXX: UNIMPLEMENTED
-	{"bz2decompress",	no_argument,		NULL, 'J'}, */
+	{"bz2decompress",	no_argument,		NULL, 'J'},
 	{"files-with-matches",	no_argument,		NULL, 'l'},
 	{"files-without-match", no_argument,            NULL, 'L'},
 /* XXX: UNIMPLEMENTED
@@ -377,6 +377,10 @@
 			iflag = 1;
 			cflags |= REG_ICASE;
 			break;
+		case 'J':
+			Zflag = 0;
+			Jflag++;
+			break;
 		case 'L':
 			lflag = 0;
 			Lflag = qflag = 1;

==== //depot/projects/soc2008/gabor_textproc/grep/grep.h#12 (text+ko) ====

@@ -60,7 +60,7 @@
 extern int	 cflags, eflags;
 
 /* Command line flags */
-extern int	 Aflag, Bflag, Dflag, Eflag, Fflag, Gflag, Hflag, Lflag, Pflag,
+extern int	 Aflag, Bflag, Dflag, Eflag, Fflag, Gflag, Hflag, Jflag, Lflag, Pflag,
 		 Sflag, Rflag, Zflag,
 		 bflag, cflag, hflag, iflag, lflag, nflag, Oflag, oflag, qflag, sflag,
 		 vflag, wflag, xflag,