libmspack 0.0.5 → 0.10.1.2

data/ext/libmspack/mspack/cabd.c

@@ -1,5 +1,5 @@
 /* This file is part of libmspack.
- * (C) 2003-2011 Stuart Caie.
+ * (C) 2003-2018 Stuart Caie.
  *
  * libmspack is free software; you can redistribute it and/or modify it under
  * the terms of the GNU Lesser General Public License (LGPL) version 2.1
@@ -23,7 +23,9 @@
 
 #include <system.h>
 #include <cab.h>
-#include <assert.h>
+#include <mszip.h>
+#include <lzx.h>
+#include <qtm.h>
 
 /* Notes on compliance with cabinet specification:
  *
@@ -72,7 +74,7 @@ static void cabd_close(
   struct mscab_decompressor *base, struct mscabd_cabinet *origcab);
 static int cabd_read_headers(
   struct mspack_system *sys, struct mspack_file *fh,
-  struct mscabd_cabinet_p *cab, off_t offset, int quiet);
+  struct mscabd_cabinet_p *cab, off_t offset, int salvage, int quiet);
 static char *cabd_read_string(
   struct mspack_system *sys, struct mspack_file *fh, int *error);
 
@@ -109,7 +111,7 @@ static int cabd_sys_write(
   struct mspack_file *file, void *buffer, int bytes);
 static int cabd_sys_read_block(
   struct mspack_system *sys, struct mscabd_decompress_state *d, int *out,
-  int ignore_cksum);
+  int ignore_cksum, int ignore_blocksize);
 static unsigned int cabd_checksum(
   unsigned char *data, unsigned int bytes, unsigned int cksum);
 static struct noned_state *noned_init(
@@ -154,9 +156,10 @@ struct mscab_decompressor *
     self->d               = NULL;
     self->error           = MSPACK_ERR_OK;
 
-    self->param[MSCABD_PARAM_SEARCHBUF] = 32768;
-    self->param[MSCABD_PARAM_FIXMSZIP]  = 0;
-    self->param[MSCABD_PARAM_DECOMPBUF] = 4096;
+    self->searchbuf_size  = 32768;
+    self->fix_mszip       = 0;
+    self->buf_size        = 4096;
+    self->salvage         = 0;
   }
   return (struct mscab_decompressor *) self;
 }
@@ -170,9 +173,9 @@ void mspack_destroy_cab_decompressor(struct mscab_decompressor *base) {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) base;
   if (self) {
     struct mspack_system *sys = self->system;
-    cabd_free_decomp(self);
     if (self->d) {
       if (self->d->infh) sys->close(self->d->infh);
+      cabd_free_decomp(self);
       sys->free(self->d);
     }
     sys->free(self);
@@ -186,7 +189,7 @@ void mspack_destroy_cab_decompressor(struct mscab_decompressor *base) {
  * opens a file and tries to read it as a cabinet file
  */
 static struct mscabd_cabinet *cabd_open(struct mscab_decompressor *base,
-					const char *filename)
+                                        const char *filename)
 {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) base;
   struct mscabd_cabinet_p *cab = NULL;
@@ -200,10 +203,10 @@ static struct mscabd_cabinet *cabd_open(struct mscab_decompressor *base,
   if ((fh = sys->open(sys, filename, MSPACK_SYS_OPEN_READ))) {
     if ((cab = (struct mscabd_cabinet_p *) sys->alloc(sys, sizeof(struct mscabd_cabinet_p)))) {
       cab->base.filename = filename;
-      error = cabd_read_headers(sys, fh, cab, (off_t) 0, 0);
+      error = cabd_read_headers(sys, fh, cab, (off_t) 0, self->salvage, 0);
       if (error) {
-	cabd_close(base, (struct mscabd_cabinet *) cab);
-	cab = NULL;
+        cabd_close(base, (struct mscabd_cabinet *) cab);
+        cab = NULL;
       }
       self->error = error;
     }
@@ -224,7 +227,7 @@ static struct mscabd_cabinet *cabd_open(struct mscab_decompressor *base,
  * frees all memory associated with a given mscabd_cabinet.
  */
 static void cabd_close(struct mscab_decompressor *base,
-		       struct mscabd_cabinet *origcab)
+                       struct mscabd_cabinet *origcab)
 {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) base;
   struct mscabd_folder_data *dat, *ndat;
@@ -252,16 +255,16 @@ static void cabd_close(struct mscab_decompressor *base,
 
       /* free folder decompression state if it has been decompressed */
       if (self->d && (self->d->folder == (struct mscabd_folder_p *) fol)) {
-	if (self->d->infh) sys->close(self->d->infh);
-	cabd_free_decomp(self);
-	sys->free(self->d);
-	self->d = NULL;
+        if (self->d->infh) sys->close(self->d->infh);
+        cabd_free_decomp(self);
+        sys->free(self->d);
+        self->d = NULL;
       }
 
       /* free folder data segments */
       for (dat = ((struct mscabd_folder_p *)fol)->data.next; dat; dat = ndat) {
-	ndat = dat->next;
-	sys->free(dat);
+        ndat = dat->next;
+        sys->free(dat);
       }
       sys->free(fol);
     }
@@ -303,11 +306,11 @@ static void cabd_close(struct mscab_decompressor *base,
  * for folders and files as necessary
  */
 static int cabd_read_headers(struct mspack_system *sys,
-			     struct mspack_file *fh,
-			     struct mscabd_cabinet_p *cab,
-			     off_t offset, int quiet)
+                             struct mspack_file *fh,
+                             struct mscabd_cabinet_p *cab,
+                             off_t offset, int salvage, int quiet)
 {
-  int num_folders, num_files, folder_resv, i, x;
+  int num_folders, num_files, folder_resv, i, x, err, fidx;
   struct mscabd_folder_p *fol, *linkfol = NULL;
   struct mscabd_file *file, *linkfile = NULL;
   unsigned char buf[64];
@@ -363,6 +366,7 @@ static int cabd_read_headers(struct mspack_system *sys,
 
   /* read the reserved-sizes part of header, if present */
   cab->base.flags = EndGetI16(&buf[cfhead_Flags]);
+
   if (cab->base.flags & cfheadRESERVE_PRESENT) {
     if (sys->read(fh, &buf[0], cfheadext_SIZEOF) != cfheadext_SIZEOF) {
       return MSPACK_ERR_READ;
@@ -378,7 +382,7 @@ static int cabd_read_headers(struct mspack_system *sys,
     /* skip the reserved header */
     if (cab->base.header_resv) {
       if (sys->seek(fh, (off_t) cab->base.header_resv, MSPACK_SYS_SEEK_CUR)) {
-	return MSPACK_ERR_SEEK;
+        return MSPACK_ERR_SEEK;
       }
     }
   }
@@ -390,14 +394,18 @@ static int cabd_read_headers(struct mspack_system *sys,
 
   /* read name and info of preceeding cabinet in set, if present */
   if (cab->base.flags & cfheadPREV_CABINET) {
-    cab->base.prevname = cabd_read_string(sys, fh, &x); if (x) return x;
-    cab->base.previnfo = cabd_read_string(sys, fh, &x); if (x) return x;
+    cab->base.prevname = cabd_read_string(sys, fh, &err);
+    if (err) return err;
+    cab->base.previnfo = cabd_read_string(sys, fh, &err);
+    if (err) return err;
   }
 
   /* read name and info of next cabinet in set, if present */
   if (cab->base.flags & cfheadNEXT_CABINET) {
-    cab->base.nextname = cabd_read_string(sys, fh, &x); if (x) return x;
-    cab->base.nextinfo = cabd_read_string(sys, fh, &x); if (x) return x;
+    cab->base.nextname = cabd_read_string(sys, fh, &err);
+    if (err) return err;
+    cab->base.nextinfo = cabd_read_string(sys, fh, &err);
+    if (err) return err;
   }
 
   /* read folders */
@@ -407,7 +415,7 @@ static int cabd_read_headers(struct mspack_system *sys,
     }
     if (folder_resv) {
       if (sys->seek(fh, (off_t) folder_resv, MSPACK_SYS_SEEK_CUR)) {
-	return MSPACK_ERR_SEEK;
+        return MSPACK_ERR_SEEK;
       }
     }
 
@@ -446,45 +454,44 @@ static int cabd_read_headers(struct mspack_system *sys,
     file->offset   = EndGetI32(&buf[cffile_FolderOffset]);
 
     /* set folder pointer */
-    x = EndGetI16(&buf[cffile_FolderIndex]);
-    if (x < cffileCONTINUED_FROM_PREV) {
-      /* normal folder index; count up to the correct folder. the folder
-       * pointer will be NULL if folder index is invalid */
-      struct mscabd_folder *ifol = cab->base.folders; 
-      while (x--) if (ifol) ifol = ifol->next;
-      file->folder = ifol;
-
-      if (!ifol) {
-	sys->free(file);
-	D(("invalid folder index"))
-	return MSPACK_ERR_DATAFORMAT;
+    fidx = EndGetI16(&buf[cffile_FolderIndex]);
+    if (fidx < cffileCONTINUED_FROM_PREV) {
+      /* normal folder index; count up to the correct folder */
+      if (fidx < num_folders) {
+        struct mscabd_folder *ifol = cab->base.folders;
+        while (fidx--) if (ifol) ifol = ifol->next;
+        file->folder = ifol;
+      }
+      else {
+        D(("invalid folder index"))
+        file->folder = NULL;
       }
     }
     else {
       /* either CONTINUED_TO_NEXT, CONTINUED_FROM_PREV or
        * CONTINUED_PREV_AND_NEXT */
-      if ((x == cffileCONTINUED_TO_NEXT) ||
-	  (x == cffileCONTINUED_PREV_AND_NEXT))
+      if ((fidx == cffileCONTINUED_TO_NEXT) ||
+          (fidx == cffileCONTINUED_PREV_AND_NEXT))
       {
-	/* get last folder */
-	struct mscabd_folder *ifol = cab->base.folders;
-	while (ifol->next) ifol = ifol->next;
-	file->folder = ifol;
-
-	/* set "merge next" pointer */
-	fol = (struct mscabd_folder_p *) ifol;
-	if (!fol->merge_next) fol->merge_next = file;
+        /* get last folder */
+        struct mscabd_folder *ifol = cab->base.folders;
+        while (ifol->next) ifol = ifol->next;
+        file->folder = ifol;
+
+        /* set "merge next" pointer */
+        fol = (struct mscabd_folder_p *) ifol;
+        if (!fol->merge_next) fol->merge_next = file;
       }
 
-      if ((x == cffileCONTINUED_FROM_PREV) ||
-	  (x == cffileCONTINUED_PREV_AND_NEXT))
+      if ((fidx == cffileCONTINUED_FROM_PREV) ||
+          (fidx == cffileCONTINUED_PREV_AND_NEXT))
       {
-	/* get first folder */
-	file->folder = cab->base.folders;
+        /* get first folder */
+        file->folder = cab->base.folders;
 
-	/* set "merge prev" pointer */
-	fol = (struct mscabd_folder_p *) file->folder;
-	if (!fol->merge_prev) fol->merge_prev = file;
+        /* set "merge prev" pointer */
+        fol = (struct mscabd_folder_p *) file->folder;
+        if (!fol->merge_prev) fol->merge_prev = file;
       }
     }
 
@@ -501,10 +508,14 @@ static int cabd_read_headers(struct mspack_system *sys,
     file->date_y = (x >> 9) + 1980;
 
     /* get filename */
-    file->filename = cabd_read_string(sys, fh, &x);
-    if (x) { 
+    file->filename = cabd_read_string(sys, fh, &err);
+
+    /* if folder index or filename are bad, either skip it or fail */
+    if (err || !file->folder) {
+      sys->free(file->filename);
       sys->free(file);
-      return x;
+      if (salvage) continue;
+      return err ? err : MSPACK_ERR_DATAFORMAT;
     }
 
     /* link file entry into file list */
@@ -513,21 +524,34 @@ static int cabd_read_headers(struct mspack_system *sys,
     linkfile = file;
   }
 
+  if (cab->base.files == NULL) {
+    /* We never actually added any files to the file list.  Something went wrong.
+     * The file header may have been invalid */
+    D(("No files found, even though header claimed to have %d files", num_files))
+    return MSPACK_ERR_DATAFORMAT;
+  }
+
   return MSPACK_ERR_OK;
 }
 
 static char *cabd_read_string(struct mspack_system *sys,
-			      struct mspack_file *fh, int *error)
+                              struct mspack_file *fh, int *error)
 {
   off_t base = sys->tell(fh);
   char buf[256], *str;
-  unsigned int len, i, ok;
+  int len, i, ok;
 
   /* read up to 256 bytes */
-  len = sys->read(fh, &buf[0], 256);
+  if ((len = sys->read(fh, &buf[0], 256)) <= 0) {
+    *error = MSPACK_ERR_READ;
+    return NULL;
+  }
 
   /* search for a null terminator in the buffer */
   for (i = 0, ok = 0; i < len; i++) if (!buf[i]) { ok = 1; break; }
+  /* reject empty strings */
+  if (i == 0) ok = 0;
+
   if (!ok) {
     *error = MSPACK_ERR_DATAFORMAT;
     return NULL;
@@ -563,7 +587,7 @@ static char *cabd_read_string(struct mspack_system *sys,
  * break out of the loop and be sure that all resources are freed
  */
 static struct mscabd_cabinet *cabd_search(struct mscab_decompressor *base,
-					  const char *filename)
+                                          const char *filename)
 {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) base;
   struct mscabd_cabinet_p *cab = NULL;
@@ -576,7 +600,7 @@ static struct mscabd_cabinet *cabd_search(struct mscab_decompressor *base,
   sys = self->system;
 
   /* allocate a search buffer */
-  search_buf = (unsigned char *) sys->alloc(sys, (size_t) self->param[MSCABD_PARAM_SEARCHBUF]);
+  search_buf = (unsigned char *) sys->alloc(sys, (size_t) self->searchbuf_size);
   if (!search_buf) {
     self->error = MSPACK_ERR_NOMEMORY;
     return NULL;
@@ -586,21 +610,21 @@ static struct mscabd_cabinet *cabd_search(struct mscab_decompressor *base,
   if ((fh = sys->open(sys, filename, MSPACK_SYS_OPEN_READ))) {
     if (!(self->error = mspack_sys_filelen(sys, fh, &filelen))) {
       self->error = cabd_find(self, search_buf, fh, filename,
-			      filelen, &firstlen, &cab);
+                              filelen, &firstlen, &cab);
     }
 
     /* truncated / extraneous data warning: */
     if (firstlen && (firstlen != filelen) &&
-	(!cab || (cab->base.base_offset == 0)))
+        (!cab || (cab->base.base_offset == 0)))
     {
       if (firstlen < filelen) {
-	sys->message(fh, "WARNING; possible %" LD
-		     " extra bytes at end of file.",
-		     filelen - firstlen);
+        sys->message(fh, "WARNING; possible %" LD
+                     " extra bytes at end of file.",
+                     filelen - firstlen);
       }
       else {
-	sys->message(fh, "WARNING; file possibly truncated by %" LD " bytes.",
-		     firstlen - filelen);
+        sys->message(fh, "WARNING; file possibly truncated by %" LD " bytes.",
+                     firstlen - filelen);
       }
     }
     
@@ -617,8 +641,8 @@ static struct mscabd_cabinet *cabd_search(struct mscab_decompressor *base,
 }
 
 static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
-		     struct mspack_file *fh, const char *filename, off_t flen,
-		     off_t *firstlen, struct mscabd_cabinet_p **firstcab)
+                     struct mspack_file *fh, const char *filename, off_t flen,
+                     off_t *firstlen, struct mscabd_cabinet_p **firstcab)
 {
   struct mscabd_cabinet_p *cab, *link = NULL;
   off_t caboff, offset, length;
@@ -627,7 +651,7 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
   unsigned int cablen_u32 = 0, foffset_u32 = 0;
   int false_cabs = 0;
 
-#ifndef LARGEFILE_SUPPORT
+#if !LARGEFILE_SUPPORT
   /* detect 32-bit off_t overflow */
   if (flen < 0) {
     sys->message(fh, largefile_msg);
@@ -640,8 +664,8 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
     /* search length is either the full length of the search buffer, or the
      * amount of data remaining to the end of the file, whichever is less. */
     length = flen - offset;
-    if (length > self->param[MSCABD_PARAM_SEARCHBUF]) {
-      length = self->param[MSCABD_PARAM_SEARCHBUF];
+    if (length > self->searchbuf_size) {
+      length = self->searchbuf_size;
     }
 
     /* fill the search buffer with data from disk */
@@ -651,22 +675,21 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
 
     /* FAQ avoidance strategy */
     if ((offset == 0) && (EndGetI32(&buf[0]) == 0x28635349)) {
-      sys->message(fh, "WARNING; found InstallShield header. "
-		   "This is probably an InstallShield file. "
-		   "Use UNSHIELD from www.synce.org to unpack it.");
+      sys->message(fh, "WARNING; found InstallShield header. Use unshield "
+                   "(https://github.com/twogood/unshield) to unpack this file"); 
     }
 
     /* read through the entire buffer. */
     for (p = &buf[0], pend = &buf[length]; p < pend; ) {
       switch (state) {
-	/* starting state */
+        /* starting state */
       case 0:
-	/* we spend most of our time in this while loop, looking for
-	 * a leading 'M' of the 'MSCF' signature */
-	while (p < pend && *p != 0x4D) p++;
-	/* if we found tht 'M', advance state */
-	if (p++ < pend) state = 1;
-	break;
+        /* we spend most of our time in this while loop, looking for
+         * a leading 'M' of the 'MSCF' signature */
+        while (p < pend && *p != 0x4D) p++;
+        /* if we found tht 'M', advance state */
+        if (p++ < pend) state = 1;
+        break;
 
       /* verify that the next 3 bytes are 'S', 'C' and 'F' */
       case 1: state = (*p++ == 0x53) ? 2 : 0; break;
@@ -688,70 +711,71 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
       case 17: foffset_u32 |= *p++ << 8;  state++; break;
       case 18: foffset_u32 |= *p++ << 16; state++; break;
       case 19: foffset_u32 |= *p++ << 24;
-	/* now we have recieved 20 bytes of potential cab header. work out
-	 * the offset in the file of this potential cabinet */
-	caboff = offset + (p - &buf[0]) - 20;
-
-	/* should reading cabinet fail, restart search just after 'MSCF' */
-	offset = caboff + 4;
-
-	/* capture the "length of cabinet" field if there is a cabinet at
-	 * offset 0 in the file, regardless of whether the cabinet can be
-	 * read correctly or not */
-	if (caboff == 0) *firstlen = (off_t) cablen_u32;
-
-	/* check that the files offset is less than the alleged length of
-	 * the cabinet, and that the offset + the alleged length are
-	 * 'roughly' within the end of overall file length */
-	if ((foffset_u32 < cablen_u32) &&
-	    ((caboff + (off_t) foffset_u32) < (flen + 32)) &&
-	    ((caboff + (off_t) cablen_u32)  < (flen + 32)) )
-	{
-	  /* likely cabinet found -- try reading it */
-	  if (!(cab = (struct mscabd_cabinet_p *) sys->alloc(sys, sizeof(struct mscabd_cabinet_p)))) {
-	    return MSPACK_ERR_NOMEMORY;
-	  }
-	  cab->base.filename = filename;
-	  if (cabd_read_headers(sys, fh, cab, caboff, 1)) {
-	    /* destroy the failed cabinet */
-	    cabd_close((struct mscab_decompressor *) self,
-		       (struct mscabd_cabinet *) cab);
-	    false_cabs++;
-	  }
-	  else {
-	    /* cabinet read correctly! */
-
-	    /* link the cab into the list */
-	    if (!link) *firstcab = cab;
-	    else link->base.next = (struct mscabd_cabinet *) cab;
-	    link = cab;
-
-	    /* cause the search to restart after this cab's data. */
-	    offset = caboff + (off_t) cablen_u32;
-
-#ifndef LARGEFILE_SUPPORT
-	    /* detect 32-bit off_t overflow */
-	    if (offset < caboff) {
-	      sys->message(fh, largefile_msg);
-	      return MSPACK_ERR_OK;
-	    }
-#endif	      
-	  }
-	}
-
-	/* restart search */
-	if (offset >= flen) return MSPACK_ERR_OK;
-	if (sys->seek(fh, offset, MSPACK_SYS_SEEK_START)) {
-	  return MSPACK_ERR_SEEK;
-	}
-	length = 0;
-	p = pend;
-	state = 0;
-	break;
+        /* now we have recieved 20 bytes of potential cab header. work out
+         * the offset in the file of this potential cabinet */
+        caboff = offset + (p - &buf[0]) - 20;
+
+        /* should reading cabinet fail, restart search just after 'MSCF' */
+        offset = caboff + 4;
+
+        /* capture the "length of cabinet" field if there is a cabinet at
+         * offset 0 in the file, regardless of whether the cabinet can be
+         * read correctly or not */
+        if (caboff == 0) *firstlen = (off_t) cablen_u32;
+
+        /* check that the files offset is less than the alleged length of
+         * the cabinet, and that the offset + the alleged length are
+         * 'roughly' within the end of overall file length. In salvage
+         * mode, don't check the alleged length, allow it to be garbage */
+        if ((foffset_u32 < cablen_u32) &&
+            ((caboff + (off_t) foffset_u32) < (flen + 32)) &&
+            (((caboff + (off_t) cablen_u32)  < (flen + 32)) || self->salvage))
+        {
+          /* likely cabinet found -- try reading it */
+          if (!(cab = (struct mscabd_cabinet_p *) sys->alloc(sys, sizeof(struct mscabd_cabinet_p)))) {
+            return MSPACK_ERR_NOMEMORY;
+          }
+          cab->base.filename = filename;
+          if (cabd_read_headers(sys, fh, cab, caboff, self->salvage, 1)) {
+            /* destroy the failed cabinet */
+            cabd_close((struct mscab_decompressor *) self,
+                       (struct mscabd_cabinet *) cab);
+            false_cabs++;
+          }
+          else {
+            /* cabinet read correctly! */
+
+            /* link the cab into the list */
+            if (!link) *firstcab = cab;
+            else link->base.next = (struct mscabd_cabinet *) cab;
+            link = cab;
+
+            /* cause the search to restart after this cab's data. */
+            offset = caboff + (off_t) cablen_u32;
+
+#if !LARGEFILE_SUPPORT
+            /* detect 32-bit off_t overflow */
+            if (offset < caboff) {
+              sys->message(fh, largefile_msg);
+              return MSPACK_ERR_OK;
+            }
+#endif        
+          }
+        }
+
+        /* restart search */
+        if (offset >= flen) return MSPACK_ERR_OK;
+        if (sys->seek(fh, offset, MSPACK_SYS_SEEK_START)) {
+          return MSPACK_ERR_SEEK;
+        }
+        length = 0;
+        p = pend;
+        state = 0;
+        break;
 
       /* for bytes 4-7 and 12-15, just advance state/pointer */
       default:
-	p++, state++;
+        p++, state++;
       } /* switch(state) */
     } /* for (... p < pend ...) */
   } /* for (... offset < length ...) */
@@ -762,7 +786,7 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
 
   return MSPACK_ERR_OK;
 }
-					     
+                                             
 /***************************************
  * CABD_MERGE, CABD_PREPEND, CABD_APPEND
  ***************************************
@@ -772,22 +796,22 @@ static int cabd_find(struct mscab_decompressor_p *self, unsigned char *buf,
  * merged folder's data parts list.
  */
 static int cabd_prepend(struct mscab_decompressor *base,
-			struct mscabd_cabinet *cab,
-			struct mscabd_cabinet *prevcab)
+                        struct mscabd_cabinet *cab,
+                        struct mscabd_cabinet *prevcab)
 {
   return cabd_merge(base, prevcab, cab);
 }
 
 static int cabd_append(struct mscab_decompressor *base,
-			struct mscabd_cabinet *cab,
-			struct mscabd_cabinet *nextcab)
+                        struct mscabd_cabinet *cab,
+                        struct mscabd_cabinet *nextcab)
 {
   return cabd_merge(base, cab, nextcab);
 }
 
 static int cabd_merge(struct mscab_decompressor *base,
-		      struct mscabd_cabinet *lcab,
-		      struct mscabd_cabinet *rcab)
+                      struct mscabd_cabinet *lcab,
+                      struct mscabd_cabinet *rcab)
 {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) base;
   struct mscabd_folder_data *data, *ndata;
@@ -877,7 +901,7 @@ static int cabd_merge(struct mscab_decompressor *base,
      * instead */
     lfol->base.num_blocks += rfol->base.num_blocks - 1;
     if ((rfol->merge_next == NULL) ||
-	(rfol->merge_next->folder != (struct mscabd_folder *) rfol))
+        (rfol->merge_next->folder != (struct mscabd_folder *) rfol))
     {
       lfol->merge_next = rfol->merge_next;
     }
@@ -900,9 +924,9 @@ static int cabd_merge(struct mscab_decompressor *base,
       rfi = fi->next;
       /* if file's folder matches the merge folder, unlink and free it */
       if (fi->folder == (struct mscabd_folder *) rfol) {
-	if (lfi) lfi->next = rfi; else lcab->files = rfi;
-	sys->free(fi->filename);
-	sys->free(fi);
+        if (lfi) lfi->next = rfi; else lcab->files = rfi;
+        sys->free(fi->filename);
+        sys->free(fi);
       }
       else lfi = fi;
     }
@@ -937,6 +961,12 @@ static int cabd_can_merge_folders(struct mspack_system *sys,
         return 0;
     }
 
+    /* check there are not too many data blocks after merging */
+    if ((lfol->base.num_blocks + rfol->base.num_blocks) > CAB_FOLDERMAX) {
+        D(("folder merge: too many data blocks in merged folders"))
+        return 0;
+    }
+
     if (!(lfi = lfol->merge_next) || !(rfi = rfol->merge_prev)) {
         D(("folder merge: one cabinet has no files to merge"))
         return 0;
@@ -947,10 +977,10 @@ static int cabd_can_merge_folders(struct mspack_system *sys,
      * should be identical in number and order. to verify this, check the
      * offset and length of each file. */
     for (l=lfi, r=rfi; l; l=l->next, r=r->next) {
-	if (!r || (l->offset != r->offset) || (l->length != r->length)) {
-	    matching = 0;
-	    break;
-	}
+        if (!r || (l->offset != r->offset) || (l->length != r->length)) {
+            matching = 0;
+            break;
+        }
     }
 
     if (matching) return 1;
@@ -960,9 +990,9 @@ static int cabd_can_merge_folders(struct mspack_system *sys,
      * the merge with a warning about missing files. */
     matching = 0;
     for (l = lfi; l; l = l->next) {
-	for (r = rfi; r; r = r->next) {
-	    if (l->offset == r->offset && l->length == r->length) break;
-	}
+        for (r = rfi; r; r = r->next) {
+            if (l->offset == r->offset && l->length == r->length) break;
+        }
         if (r) matching = 1; else sys->message(NULL,
             "WARNING; merged file %s not listed in both cabinets", l->filename);
     }
@@ -982,6 +1012,7 @@ static int cabd_extract(struct mscab_decompressor *base,
   struct mscabd_folder_p *fol;
   struct mspack_system *sys;
   struct mspack_file *fh;
+  off_t filelen;
 
   if (!self) return MSPACK_ERR_ARGS;
   if (!file) return self->error = MSPACK_ERR_ARGS;
@@ -989,15 +1020,43 @@ static int cabd_extract(struct mscab_decompressor *base,
   sys = self->system;
   fol = (struct mscabd_folder_p *) file->folder;
 
-  /* check if file can be extracted */
-  if ((!fol) || (fol->merge_prev) ||
-      (((file->offset + file->length) / CAB_BLOCKMAX) > fol->base.num_blocks))
-  {
-    sys->message(NULL, "ERROR; file \"%s\" cannot be extracted, "
-		 "cabinet set is incomplete.", file->filename);
+  /* if offset is beyond 2GB, nothing can be extracted */
+  if (file->offset > CAB_LENGTHMAX) {
     return self->error = MSPACK_ERR_DATAFORMAT;
   }
 
+  /* if file claims to go beyond 2GB either error out,
+   * or in salvage mode reduce file length so it fits 2GB limit
+   */
+  filelen = file->length;
+  if (filelen > CAB_LENGTHMAX || (file->offset + filelen) > CAB_LENGTHMAX) {
+    if (self->salvage) {
+      filelen = CAB_LENGTHMAX - file->offset;
+    }
+    else {
+      return self->error = MSPACK_ERR_DATAFORMAT;
+    }
+  }
+
+  /* extraction impossible if no folder, or folder needs predecessor */
+  if (!fol || fol->merge_prev) {
+    sys->message(NULL, "ERROR; file \"%s\" cannot be extracted, "
+                 "cabinet set is incomplete", file->filename);
+    return self->error = MSPACK_ERR_DECRUNCH;
+  }
+
+  /* if file goes beyond what can be decoded, given an error.
+   * In salvage mode, don't assume block sizes, just try decoding
+   */
+  if (!self->salvage) {
+    off_t maxlen = fol->base.num_blocks * CAB_BLOCKMAX;
+    if ((file->offset + filelen) > maxlen) {
+      sys->message(NULL, "ERROR; file \"%s\" cannot be extracted, "
+                   "cabinet set is incomplete", file->filename);
+      return self->error = MSPACK_ERR_DECRUNCH;
+    }
+  }
+
   /* allocate generic decompression state */
   if (!self->d) {
     self->d = (struct mscabd_decompress_state *) sys->alloc(sys, sizeof(struct mscabd_decompress_state));
@@ -1013,14 +1072,19 @@ static int cabd_extract(struct mscab_decompressor *base,
   }
 
   /* do we need to change folder or reset the current folder? */
-  if ((self->d->folder != fol) || (self->d->offset > file->offset)) {
+  if ((self->d->folder != fol) || (self->d->offset > file->offset) ||
+      !self->d->state)
+  {
+    /* free any existing decompressor */
+    cabd_free_decomp(self);
+
     /* do we need to open a new cab file? */
     if (!self->d->infh || (fol->data.cab != self->d->incab)) {
       /* close previous file handle if from a different cab */
       if (self->d->infh) sys->close(self->d->infh);
       self->d->incab = fol->data.cab;
       self->d->infh = sys->open(sys, fol->data.cab->base.filename,
-				MSPACK_SYS_OPEN_READ);
+                                MSPACK_SYS_OPEN_READ);
       if (!self->d->infh) return self->error = MSPACK_ERR_OPEN;
     }
     /* seek to start of data blocks */
@@ -1038,6 +1102,7 @@ static int cabd_extract(struct mscab_decompressor *base,
     self->d->data   = &fol->data;
     self->d->offset = 0;
     self->d->block  = 0;
+    self->d->outlen = 0;
     self->d->i_ptr = self->d->i_end = &self->d->input[0];
 
     /* read_error lasts for the lifetime of a decompressor */
@@ -1052,7 +1117,7 @@ static int cabd_extract(struct mscab_decompressor *base,
   self->error = MSPACK_ERR_OK;
 
   /* if file has more than 0 bytes */
-  if (file->length) {
+  if (filelen) {
     off_t bytes;
     int error;
     /* get to correct offset.
@@ -1062,14 +1127,14 @@ static int cabd_extract(struct mscab_decompressor *base,
      */
     self->d->outfh = NULL;
     if ((bytes = file->offset - self->d->offset)) {
-      error = self->d->decompress(self->d->state, bytes);
-      self->error = (error == MSPACK_ERR_READ) ? self->read_error : error;
+        error = self->d->decompress(self->d->state, bytes);
+        self->error = (error == MSPACK_ERR_READ) ? self->read_error : error;
     }
 
     /* if getting to the correct offset was error free, unpack file */
     if (!self->error) {
       self->d->outfh = fh;
-      error = self->d->decompress(self->d->state, (off_t) file->length);
+      error = self->d->decompress(self->d->state, filelen);
       self->error = (error == MSPACK_ERR_READ) ? self->read_error : error;
     }
   }
@@ -1095,34 +1160,27 @@ static int cabd_init_decomp(struct mscab_decompressor_p *self, unsigned int ct)
 {
   struct mspack_file *fh = (struct mspack_file *) self;
 
-  assert(self && self->d);
-
-  /* free any existing decompressor */
-  cabd_free_decomp(self);
-
   self->d->comp_type = ct;
 
   switch (ct & cffoldCOMPTYPE_MASK) {
   case cffoldCOMPTYPE_NONE:
     self->d->decompress = (int (*)(void *, off_t)) &noned_decompress;
-    self->d->state = noned_init(&self->d->sys, fh, fh,
-				self->param[MSCABD_PARAM_DECOMPBUF]);
+    self->d->state = noned_init(&self->d->sys, fh, fh, self->buf_size);
     break;
   case cffoldCOMPTYPE_MSZIP:
     self->d->decompress = (int (*)(void *, off_t)) &mszipd_decompress;
-    self->d->state = mszipd_init(&self->d->sys, fh, fh,
-				 self->param[MSCABD_PARAM_DECOMPBUF],
-				 self->param[MSCABD_PARAM_FIXMSZIP]);
+    self->d->state = mszipd_init(&self->d->sys, fh, fh, self->buf_size,
+                                 self->fix_mszip);
     break;
   case cffoldCOMPTYPE_QUANTUM:
     self->d->decompress = (int (*)(void *, off_t)) &qtmd_decompress;
     self->d->state = qtmd_init(&self->d->sys, fh, fh, (int) (ct >> 8) & 0x1f,
-			       self->param[MSCABD_PARAM_DECOMPBUF]);
+                               self->buf_size);
     break;
   case cffoldCOMPTYPE_LZX:
     self->d->decompress = (int (*)(void *, off_t)) &lzxd_decompress;
     self->d->state = lzxd_init(&self->d->sys, fh, fh, (int) (ct >> 8) & 0x1f, 0,
-			       self->param[MSCABD_PARAM_DECOMPBUF], (off_t)0,0);
+                               self->buf_size, (off_t)0,0);
     break;
   default:
     return self->error = MSPACK_ERR_DATAFORMAT;
@@ -1131,7 +1189,7 @@ static int cabd_init_decomp(struct mscab_decompressor_p *self, unsigned int ct)
 }
 
 static void cabd_free_decomp(struct mscab_decompressor_p *self) {
-  if (!self || !self->d || !self->d->folder || !self->d->state) return;
+  if (!self || !self->d || !self->d->state) return;
 
   switch (self->d->comp_type & cffoldCOMPTYPE_MASK) {
   case cffoldCOMPTYPE_NONE:    noned_free((struct noned_state *) self->d->state);   break;
@@ -1159,10 +1217,12 @@ static int cabd_sys_read(struct mspack_file *file, void *buffer, int bytes) {
   struct mscab_decompressor_p *self = (struct mscab_decompressor_p *) file;
   unsigned char *buf = (unsigned char *) buffer;
   struct mspack_system *sys = self->system;
-  int avail, todo, outlen, ignore_cksum;
+  int avail, todo, outlen, ignore_cksum, ignore_blocksize;
 
-  ignore_cksum = self->param[MSCABD_PARAM_FIXMSZIP] &&
-    ((self->d->comp_type & cffoldCOMPTYPE_MASK) == cffoldCOMPTYPE_MSZIP);
+  ignore_cksum = self->salvage ||
+    (self->fix_mszip && 
+     ((self->d->comp_type & cffoldCOMPTYPE_MASK) == cffoldCOMPTYPE_MSZIP));
+  ignore_blocksize = self->salvage;
 
   todo = bytes;
   while (todo > 0) {
@@ -1182,37 +1242,35 @@ static int cabd_sys_read(struct mspack_file *file, void *buffer, int bytes) {
 
       /* check if we're out of input blocks, advance block counter */
       if (self->d->block++ >= self->d->folder->base.num_blocks) {
-	self->read_error = MSPACK_ERR_DATAFORMAT;
-	break;
+        if (!self->salvage) {
+          self->read_error = MSPACK_ERR_DATAFORMAT;
+        }
+        else {
+          D(("Ran out of CAB input blocks prematurely"))
+        }
+        break;
       }
 
       /* read a block */
-      self->read_error = cabd_sys_read_block(sys, self->d, &outlen, ignore_cksum);
+      self->read_error = cabd_sys_read_block(sys, self->d, &outlen,
+        ignore_cksum, ignore_blocksize);
       if (self->read_error) return -1;
+      self->d->outlen += outlen;
 
       /* special Quantum hack -- trailer byte to allow the decompressor
        * to realign itself. CAB Quantum blocks, unlike LZX blocks, can have
        * anything from 0 to 4 trailing null bytes. */
       if ((self->d->comp_type & cffoldCOMPTYPE_MASK)==cffoldCOMPTYPE_QUANTUM) {
-	*self->d->i_end++ = 0xFF;
+        *self->d->i_end++ = 0xFF;
       }
 
       /* is this the last block? */
       if (self->d->block >= self->d->folder->base.num_blocks) {
-	/* last block */
-	if ((self->d->comp_type & cffoldCOMPTYPE_MASK) == cffoldCOMPTYPE_LZX) {
-	  /* special LZX hack -- on the last block, inform LZX of the
-	   * size of the output data stream. */
-	  lzxd_set_output_length((struct lzxd_stream *) self->d->state, (off_t)
-				 ((self->d->block-1) * CAB_BLOCKMAX + outlen));
-	}
-      }
-      else {
-	/* not the last block */
-	if (outlen != CAB_BLOCKMAX) {
-	  self->system->message(self->d->infh,
-				"WARNING; non-maximal data block");
-	}
+        if ((self->d->comp_type & cffoldCOMPTYPE_MASK) == cffoldCOMPTYPE_LZX) {
+          /* special LZX hack -- on the last block, inform LZX of the
+           * size of the output data stream. */
+          lzxd_set_output_length((struct lzxd_stream *) self->d->state, self->d->outlen);
+        }
       }
     } /* if (avail) */
   } /* while (todo > 0) */
@@ -1235,12 +1293,13 @@ static int cabd_sys_write(struct mspack_file *file, void *buffer, int bytes) {
  * one cab file, if it does then the fragments will be reassembled
  */
 static int cabd_sys_read_block(struct mspack_system *sys,
-			       struct mscabd_decompress_state *d,
-			       int *out, int ignore_cksum)
+                               struct mscabd_decompress_state *d,
+                               int *out, int ignore_cksum,
+                               int ignore_blocksize)
 {
   unsigned char hdr[cfdata_SIZEOF];
   unsigned int cksum;
-  int len;
+  int len, full_len;
 
   /* reset the input block pointer and end of block pointer */
   d->i_ptr = d->i_end = &d->input[0];
@@ -1253,24 +1312,27 @@ static int cabd_sys_read_block(struct mspack_system *sys,
 
     /* skip any reserved block headers */
     if (d->data->cab->block_resv &&
-	sys->seek(d->infh, (off_t) d->data->cab->block_resv,
-		  MSPACK_SYS_SEEK_CUR))
+        sys->seek(d->infh, (off_t) d->data->cab->block_resv,
+                  MSPACK_SYS_SEEK_CUR))
     {
       return MSPACK_ERR_SEEK;
     }
 
     /* blocks must not be over CAB_INPUTMAX in size */
     len = EndGetI16(&hdr[cfdata_CompressedSize]);
-    if (((d->i_end - d->i_ptr) + len) > CAB_INPUTMAX) {
-      D(("block size > CAB_INPUTMAX (%ld + %d)",
-          (long)(d->i_end - d->i_ptr), len))
-      return MSPACK_ERR_DATAFORMAT;
+    full_len = (d->i_end - d->i_ptr) + len; /* include cab-spanning blocks */
+    if (full_len > CAB_INPUTMAX) {
+      D(("block size %d > CAB_INPUTMAX", full_len));
+      /* in salvage mode, blocks can be 65535 bytes but no more than that */
+      if (!ignore_blocksize || full_len > CAB_INPUTMAX_SALVAGE) {
+          return MSPACK_ERR_DATAFORMAT;
+      }
     }
 
      /* blocks must not expand to more than CAB_BLOCKMAX */
     if (EndGetI16(&hdr[cfdata_UncompressedSize]) > CAB_BLOCKMAX) {
       D(("block size > CAB_BLOCKMAX"))
-      return MSPACK_ERR_DATAFORMAT;
+      if (!ignore_blocksize) return MSPACK_ERR_DATAFORMAT;
     }
 
     /* read the block data */
@@ -1282,8 +1344,8 @@ static int cabd_sys_read_block(struct mspack_system *sys,
     if ((cksum = EndGetI32(&hdr[cfdata_CheckSum]))) {
       unsigned int sum2 = cabd_checksum(d->i_end, (unsigned int) len, 0);
       if (cabd_checksum(&hdr[4], 4, sum2) != cksum) {
-	if (!ignore_cksum) return MSPACK_ERR_CHECKSUM;
-	sys->message(d->infh, "WARNING; bad block checksum found");
+        if (!ignore_cksum) return MSPACK_ERR_CHECKSUM;
+        sys->message(d->infh, "WARNING; bad block checksum found");
       }
     }
 
@@ -1308,14 +1370,14 @@ static int cabd_sys_read_block(struct mspack_system *sys,
 
     /* advance to next member in the cabinet set */
     if (!(d->data = d->data->next)) {
-      D(("ran out of splits in cabinet set"))
+      sys->message(d->infh, "WARNING; ran out of cabinets in set. Are any missing?");
       return MSPACK_ERR_DATAFORMAT;
     }
 
     /* open next cab file */
     d->incab = d->data->cab;
     if (!(d->infh = sys->open(sys, d->incab->base.filename,
-			      MSPACK_SYS_OPEN_READ)))
+                              MSPACK_SYS_OPEN_READ)))
     {
       return MSPACK_ERR_OPEN;
     }
@@ -1331,7 +1393,7 @@ static int cabd_sys_read_block(struct mspack_system *sys,
 }
 
 static unsigned int cabd_checksum(unsigned char *data, unsigned int bytes,
-				  unsigned int cksum)
+                                  unsigned int cksum)
 {
   unsigned int len, ul = 0;
 
@@ -1340,8 +1402,8 @@ static unsigned int cabd_checksum(unsigned char *data, unsigned int bytes,
   }
 
   switch (bytes & 3) {
-  case 3: ul |= *data++ << 16;
-  case 2: ul |= *data++ <<  8;
+  case 3: ul |= *data++ << 16; /*@fallthrough@*/
+  case 2: ul |= *data++ <<  8; /*@fallthrough@*/
   case 1: ul |= *data;
   }
   cksum ^= ul;
@@ -1363,9 +1425,9 @@ struct noned_state {
 };
 
 static struct noned_state *noned_init(struct mspack_system *sys,
-				      struct mspack_file *in,
-				      struct mspack_file *out,
-				      int bufsize)
+                                      struct mspack_file *in,
+                                      struct mspack_file *out,
+                                      int bufsize)
 {
   struct noned_state *state = (struct noned_state *) sys->alloc(sys, sizeof(struct noned_state));
   unsigned char *buf = (unsigned char *) sys->alloc(sys, (size_t) bufsize);
@@ -1417,14 +1479,17 @@ static int cabd_param(struct mscab_decompressor *base, int param, int value) {
   switch (param) {
   case MSCABD_PARAM_SEARCHBUF:
     if (value < 4) return MSPACK_ERR_ARGS;
-    self->param[MSCABD_PARAM_SEARCHBUF] = value;
+    self->searchbuf_size = value;
     break;
   case MSCABD_PARAM_FIXMSZIP:
-    self->param[MSCABD_PARAM_FIXMSZIP] = value;
+    self->fix_mszip = value;
     break;
   case MSCABD_PARAM_DECOMPBUF:
     if (value < 4) return MSPACK_ERR_ARGS;
-    self->param[MSCABD_PARAM_DECOMPBUF] = value;
+    self->buf_size = value;
+    break;
+  case MSCABD_PARAM_SALVAGE:
+    self->salvage = value;
     break;
   default:
     return MSPACK_ERR_ARGS;