1 files changed, 435 insertions, 0 deletions
diff --git a/src/decompress.c b/src/decompress.c
new file mode 100644
index 0000000..c483241
--- /dev/null
+++ b/src/decompress.c
@@ -0,0 +1,435 @@
+/*
+ * decompress.c: decompression abstraction layer
+ *
+ * Copyright (C) 2007, 2008 Colin Watson.
+ *
+ * This file is part of man-db.
+ *
+ * man-db 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 2 of the License, or
+ * (at your option) any later version.
+ *
+ * man-db 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 man-db; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+#ifdef HAVE_CONFIG_H
+#  include "config.h"
+#endif /* HAVE_CONFIG_H */
+
+#include <assert.h>
+#include <string.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <fcntl.h>
+
+#ifdef HAVE_LIBZ
+#  include "zlib.h"
+#endif /* HAVE_LIBZ */
+
+#include "pipeline.h"
+
+#include "attribute.h"
+#include "minmax.h"
+#include "xalloc.h"
+#include "xstrndup.h"
+#include "xvasprintf.h"
+
+#include "manconfig.h"
+
+#include "compression.h"
+#include "sandbox.h"
+
+#include "decompress.h"
+
+enum decompress_tag {
+	DECOMPRESS_PIPELINE,
+	DECOMPRESS_INPROCESS
+};
+
+struct decompress_inprocess {
+	char *buf;
+	size_t len;
+	size_t offset;
+	char *line_cache;
+};
+
+struct decompress {
+	enum decompress_tag tag;
+	union {
+		pipeline *p;
+		struct decompress_inprocess inprocess;
+	} u;
+};
+
+/* Create a new pipeline-based decompressor.  Takes ownership of p. */
+static decompress *decompress_new_pipeline (pipeline *p)
+{
+	decompress *d = XMALLOC (decompress);
+
+	d->tag = DECOMPRESS_PIPELINE;
+	d->u.p = p;
+
+	return d;
+}
+
+#ifdef HAVE_LIBZ
+
+/* Create a new in-process decompressor.  Takes ownership of buf. */
+static decompress *decompress_new_inprocess (char *buf, size_t len)
+{
+	decompress *d = XMALLOC (decompress);
+
+	d->tag = DECOMPRESS_INPROCESS;
+	d->u.inprocess.buf = buf;
+	d->u.inprocess.len = len;
+	d->u.inprocess.offset = 0;
+	d->u.inprocess.line_cache = NULL;
+
+	return d;
+}
+
+static void decompress_zlib (void *data MAYBE_UNUSED)
+{
+	gzFile zlibfile;
+	int fd;
+
+	fd = dup (STDIN_FILENO);
+	if (fd < 0)
+		return;
+
+	zlibfile = gzdopen (fd, "r");
+	if (!zlibfile) {
+		close (fd);
+		return;
+	}
+
+	for (;;) {
+		char buffer[4096];
+		int r = gzread (zlibfile, buffer, 4096);
+		if (r <= 0)
+			break;
+		if (fwrite (buffer, 1, (size_t) r, stdout) < (size_t) r)
+			break;
+	}
+
+	gzclose (zlibfile);
+	return;
+}
+
+/* The largest number of uncompressed bytes we're prepared to read into
+ * memory.  (We actually allow at most one fewer byte than this, for easy
+ * EOF detection.)
+ *
+ * At the time of writing, 11 out of 27959 (0.04%) installed manual pages on
+ * the author's system were larger than this.
+ *
+ * We could lift this restriction if we streamed in-process decompression
+ * instead, but that's a bit complicated: we'd also need to stream encoding
+ * conversion, and there's relatively little point until lexgrog can rely on
+ * preprocessor header lines rather than having to scan the whole file for
+ * preprocessor indications.  For the time being, one-shot buffering is
+ * cheap enough and much simpler.
+ */
+#define MAX_INPROCESS 1048576
+
+static decompress *decompress_try_zlib (const char *filename)
+{
+	gzFile zlibfile;
+	/* We only ever call this from the parent process (and don't
+	 * currently use threads), and this lets us skip per-file memory
+	 * allocation.
+	 */
+	static char buffer[MAX_INPROCESS];
+	int len = 0;
+
+	zlibfile = gzopen (filename, "r");
+	if (!zlibfile)
+		return NULL;
+
+	while (len < MAX_INPROCESS) {
+		/* Read one more byte than we're prepared to return, in
+		 * order to detect EOF at the right position.  The "len >=
+		 * MAX_INPROCESS" check below catches the boundary case.
+		 */
+		int r = gzread (zlibfile, buffer + len, MAX_INPROCESS - len);
+		if (r < 0) {
+			gzclose (zlibfile);
+			return NULL;
+		} else if (r == 0)
+			break;
+		else
+			len += r;
+	}
+
+	gzclose (zlibfile);
+	if (len >= MAX_INPROCESS)
+		return NULL;
+	/* Copy input data so that we don't have potential data corruption
+	 * if more than one in-process decompressor is active at once.  (An
+	 * alternative might be to use a lock to prevent that situation.)
+	 */
+	return decompress_new_inprocess (xmemdup (buffer, (size_t) len),
+					 (size_t) len);
+}
+
+#define OPEN_FLAGS_UNUSED
+#else /* !HAVE_LIBZ */
+#define OPEN_FLAGS_UNUSED MAYBE_UNUSED
+#endif /* HAVE_LIBZ */
+
+extern man_sandbox *sandbox;
+
+decompress *decompress_open (const char *filename, int flags OPEN_FLAGS_UNUSED)
+{
+	pipecmd *cmd;
+	pipeline *p;
+	struct stat st;
+#ifdef HAVE_LIBZ
+	size_t filename_len;
+#endif /* HAVE_LIBZ */
+	char *ext;
+	struct compression *comp;
+
+	if (stat (filename, &st) < 0 || S_ISDIR (st.st_mode))
+		return NULL;
+
+#ifdef HAVE_LIBZ
+	filename_len = strlen (filename);
+	if (filename_len > 3 && STREQ (filename + filename_len - 3, ".gz")) {
+		if (flags & DECOMPRESS_ALLOW_INPROCESS) {
+			decompress *d = decompress_try_zlib (filename);
+			if (d)
+				return d;
+		}
+
+		cmd = pipecmd_new_function ("zcat", &decompress_zlib, NULL,
+					    NULL);
+		pipecmd_pre_exec (cmd, sandbox_load, sandbox_free, sandbox);
+		p = pipeline_new_commands (cmd, (void *) 0);
+		goto got_pipeline;
+	}
+#endif /* HAVE_LIBZ */
+
+	ext = strrchr (filename, '.');
+	if (ext) {
+		++ext;
+
+		for (comp = comp_list; comp->ext; ++comp) {
+			if (!STREQ (comp->ext, ext))
+				continue;
+
+			cmd = pipecmd_new_argstr (comp->prog);
+			pipecmd_pre_exec (cmd, sandbox_load, sandbox_free,
+					  sandbox);
+			p = pipeline_new_commands (cmd, (void *) 0);
+			goto got_pipeline;
+		}
+	}
+
+#ifdef HAVE_GZIP
+	/* HP-UX */
+	ext = strstr (filename, ".Z/");
+	if (ext) {
+		cmd = pipecmd_new_argstr (PROG_GUNZIP);
+		pipecmd_pre_exec (cmd, sandbox_load, sandbox_free, sandbox);
+		p = pipeline_new_commands (cmd, (void *) 0);
+		goto got_pipeline;
+	}
+#endif
+
+	p = pipeline_new ();
+
+got_pipeline:
+	pipeline_want_infile (p, filename);
+	pipeline_want_out (p, -1);
+	return decompress_new_pipeline (p);
+}
+
+decompress *decompress_fdopen (int fd)
+{
+	pipeline *p;
+#ifdef HAVE_LIBZ
+	pipecmd *cmd;
+#endif /* HAVE_LIBZ */
+
+#ifdef HAVE_LIBZ
+	cmd = pipecmd_new_function ("zcat", &decompress_zlib, NULL, NULL);
+	pipecmd_pre_exec (cmd, sandbox_load, sandbox_free, sandbox);
+	p = pipeline_new_commands (cmd, (void *) 0);
+#else /* HAVE_LIBZ */
+	p = pipeline_new ();
+#endif /* HAVE_LIBZ */
+
+	pipeline_want_in (p, fd);
+	pipeline_want_out (p, -1);
+	return decompress_new_pipeline (p);
+}
+
+bool ATTRIBUTE_PURE decompress_is_pipeline (decompress *d)
+{
+	return d->tag == DECOMPRESS_PIPELINE;
+}
+
+pipeline * ATTRIBUTE_PURE decompress_get_pipeline (decompress *d)
+{
+	assert (d->tag == DECOMPRESS_PIPELINE);
+	return d->u.p;
+}
+
+const char * ATTRIBUTE_PURE decompress_inprocess_buf (decompress *d)
+{
+	assert (d->tag == DECOMPRESS_INPROCESS);
+	return d->u.inprocess.buf;
+}
+
+size_t ATTRIBUTE_PURE decompress_inprocess_len (decompress *d)
+{
+	assert (d->tag == DECOMPRESS_INPROCESS);
+	return d->u.inprocess.len;
+}
+
+void decompress_inprocess_replace (decompress *d, char *buf, size_t len)
+{
+	assert (d->tag == DECOMPRESS_INPROCESS);
+
+	free (d->u.inprocess.line_cache);
+	free (d->u.inprocess.buf);
+
+	d->u.inprocess.buf = buf;
+	d->u.inprocess.len = len;
+	d->u.inprocess.offset = 0;
+	d->u.inprocess.line_cache = NULL;
+}
+
+void decompress_start (decompress *d)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		pipeline_start (d->u.p);
+}
+
+const char *decompress_read (decompress *d, size_t *len)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		return pipeline_read (d->u.p, len);
+	else {
+		const char *ret;
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		*len = MIN (*len, d->u.inprocess.len - d->u.inprocess.offset);
+		ret = d->u.inprocess.buf + d->u.inprocess.offset;
+		d->u.inprocess.offset += *len;
+		return ret;
+	}
+}
+
+const char *decompress_peek (decompress *d, size_t *len)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		return pipeline_peek (d->u.p, len);
+	else {
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		*len = MIN (*len, d->u.inprocess.len - d->u.inprocess.offset);
+		return d->u.inprocess.buf + d->u.inprocess.offset;
+	}
+}
+
+void decompress_peek_skip (decompress *d, size_t len)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		pipeline_peek_skip (d->u.p, len);
+	else {
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		assert (len <= d->u.inprocess.len - d->u.inprocess.offset);
+		d->u.inprocess.offset += len;
+	}
+}
+
+const char *decompress_readline (decompress *d)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		return pipeline_readline (d->u.p);
+	else {
+		const char *cur, *end;
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		/* This isn't on the hot path (only called for a few lines
+		 * at the start of the file), so we can afford to
+		 * reallocate.
+		 */
+		if (d->u.inprocess.line_cache) {
+			free (d->u.inprocess.line_cache);
+			d->u.inprocess.line_cache = NULL;
+		}
+		cur = d->u.inprocess.buf + d->u.inprocess.offset;
+		end = memchr (cur, '\n',
+			      d->u.inprocess.len - d->u.inprocess.offset);
+		if (end) {
+			d->u.inprocess.line_cache = xstrndup
+				(cur, end - cur + 1);
+			d->u.inprocess.offset += end - cur + 1;
+			return d->u.inprocess.line_cache;
+		} else
+			return NULL;
+	}
+}
+
+const char *decompress_peekline (decompress *d)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		return pipeline_peekline (d->u.p);
+	else {
+		const char *cur, *end;
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		/* This isn't on the hot path (only called for a few lines
+		 * at the start of the file), so we can afford to
+		 * reallocate.
+		 */
+		if (d->u.inprocess.line_cache) {
+			free (d->u.inprocess.line_cache);
+			d->u.inprocess.line_cache = NULL;
+		}
+		cur = d->u.inprocess.buf + d->u.inprocess.offset;
+		end = memchr (cur, '\n',
+			      d->u.inprocess.len - d->u.inprocess.offset);
+		if (end) {
+			d->u.inprocess.line_cache = xstrndup
+				(cur, end - cur + 1);
+			return d->u.inprocess.line_cache;
+		} else
+			return NULL;
+	}
+}
+
+int decompress_wait (decompress *d)
+{
+	if (d->tag == DECOMPRESS_PIPELINE)
+		return pipeline_wait (d->u.p);
+	else {
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		return 0;
+	}
+}
+
+void decompress_free (decompress *d)
+{
+	if (!d)
+		return;
+	if (d->tag == DECOMPRESS_PIPELINE)
+		pipeline_free (d->u.p);
+	else {
+		assert (d->tag == DECOMPRESS_INPROCESS);
+		free (d->u.inprocess.line_cache);
+		free (d->u.inprocess.buf);
+	}
+	free (d);
+}