From 393fd790b03708585f64b6646c28a09fed44b610 Mon Sep 17 00:00:00 2001
From: Moinak Ghosh <moinakg@gmail.com>
Date: Sun, 8 Dec 2013 23:24:06 +0530
Subject: [PATCH] Add more robust checks for Jpeg and packJPG format files in
 filter routine. Use case-insensitive checks for extension names. Enable more
 features based on compression level, when archiving.

---
 archive/pc_arc_filter.c | 25 ++++++++++++++++++-------
 archive/pc_archive.c    | 10 ++++++++--
 archive/pjpg_helper.cpp |  3 +++
 pcompress.c             |  8 ++++++++
 4 files changed, 37 insertions(+), 9 deletions(-)

diff --git a/archive/pc_arc_filter.c b/archive/pc_arc_filter.c
index 3236f80..c691f75 100644
--- a/archive/pc_arc_filter.c
+++ b/archive/pc_arc_filter.c
@@ -44,6 +44,8 @@
 
 #define	PACKJPG_DEF_BUFSIZ	(512 * 1024)
 #define	JPG_SIZE_LIMIT		(8 * 1024 * 1024)
+#define	PJG_APPVERSION1		(25)
+#define	PJG_APPVERSION2		(25)
 
 struct packjpg_filter_data {
 	uchar_t *in_buff;
@@ -126,9 +128,12 @@ write_archive_data(struct archive *aw, uchar_t *out_buf, size_t len, int block_s
 	return (tot);
 }
 
-/*
- * Helper routine to bridge to packJPG C++ lib, without changing packJPG itself.
- */
+int
+pjg_version_supported(char ver)
+{
+	return (ver >= PJG_APPVERSION1 && ver <= PJG_APPVERSION2);
+}
+
 ssize_t
 packjpg_filter(struct filter_info *fi, void *filter_private)
 {
@@ -150,9 +155,14 @@ packjpg_filter(struct filter_info *fi, void *filter_private)
 		}
 
 		/*
-		 * We are trying to compress and this is not a jpeg. Skip.
+		 * We are trying to compress and this is not a proper jpeg. Skip.
 		 */
-		if (mapbuf[0] != 0xFF && mapbuf[1] != 0xD8) {
+		if (mapbuf[0] != 0xFF || mapbuf[1] != 0xD8) {
+			munmap(mapbuf, len);
+			return (FILTER_RETURN_SKIP);
+		}
+		if (strncmp((char *)&mapbuf[6], "Exif", 4) != 0 &&
+		    strncmp((char *)&mapbuf[6], "JFIF", 4) != 0) {
 			munmap(mapbuf, len);
 			return (FILTER_RETURN_SKIP);
 		}
@@ -187,9 +197,10 @@ packjpg_filter(struct filter_info *fi, void *filter_private)
 
 		/*
 		 * We are trying to decompress and this is not a packJPG file.
-		 * Write the raw data and skip.
+		 * Write the raw data and skip. Third byte in PackJPG file is
+		 * version number. We also check if it is supported.
 		 */
-		if (mapbuf[0] != 'J' && mapbuf[1] != 'S') {
+		if (mapbuf[0] != 'J' || mapbuf[1] != 'S' || !pjg_version_supported(mapbuf[2])) {
 			return (write_archive_data(fi->target_arc, pjdat->in_buff,
 			    len, fi->block_size));
 		}
diff --git a/archive/pc_archive.c b/archive/pc_archive.c
index b8c0133..7f97841 100644
--- a/archive/pc_archive.c
+++ b/archive/pc_archive.c
@@ -1393,13 +1393,15 @@ detect_type_by_ext(const char *path, int pathlen)
 	ub4 slot;
 	int i, len;
 	uint64_t extnum;
+	char extl[8];
 
 	for (i = pathlen-1; i > 0 && path[i] != '.' && path[i] != PATHSEP_CHAR; i--);
 	if (i == 0 || path[i] != '.') goto out; // If extension not found give up
 	len = pathlen - i - 1;
-	if (len == 0) goto out; // If extension is empty give up
+	if (len == 0 || len > 8) goto out; // If extension is empty give up
 	ext = &path[i+1];
-	slot = phash(ext, len);
+	for (i = 0; i < len; i++) extl[i] = tolower(ext[i]);
+	slot = phash(extl, len);
 	if (slot >= PHASHNKEYS) goto out; // Extension maps outside hash table range, give up
 	extnum = 0;
 
@@ -1454,6 +1456,10 @@ out:
 
 /*
  * Detect a few file types from looking at magic signatures.
+ * NOTE: Jpeg files must be detected via '.jpg' or '.jpeg' (case-insensitive)
+ *	extensions. Do not add Jpeg header detection here. it will break
+ *	context based PackJPG processing. Jpeg files not have proper
+ *	extension must not be processed via PackJPG.
  */
 static int
 detect_type_by_data(uchar_t *buf, size_t len)
diff --git a/archive/pjpg_helper.cpp b/archive/pjpg_helper.cpp
index e7fc996..b82c32d 100644
--- a/archive/pjpg_helper.cpp
+++ b/archive/pjpg_helper.cpp
@@ -65,6 +65,9 @@ typedef unsigned char uchar_t;
 
 #define	POLAROID_LE 0x64696f72616c6f50
 
+/*
+ * Helper routine to bridge to packJPG C++ lib, without changing packJPG itself.
+ */
 size_t
 packjpg_filter_process(uchar_t *in_buf, size_t len, uchar_t **out_buf)
 {
diff --git a/pcompress.c b/pcompress.c
index 2237fb5..d36f906 100644
--- a/pcompress.c
+++ b/pcompress.c
@@ -3230,6 +3230,14 @@ init_pc_context(pc_ctx_t *pctx, int argc, char *argv[])
 			if (pctx->level > 8) pctx->dispack_preprocess = 1;
 			if (pctx->level > 4) pctx->enable_delta2_encode = 1;
 			if (pctx->level > 9) pctx->lzp_preprocess = 1;
+			if (pctx->level > 3) {
+				pctx->enable_rabin_global = 1;
+				pctx->enable_rabin_scan = 1;
+				pctx->enable_rabin_split = 1;
+				pctx->rab_blk_size = 2;
+				if (pctx->level > 4) pctx->rab_blk_size = 1;
+				if (pctx->level > 8) pctx->rab_blk_size = 0;
+			}
 		}
 		if (pctx->lzp_preprocess || pctx->enable_delta2_encode || pctx->dispack_preprocess) {
 			pctx->preprocess_mode = 1;