When dumping joint MOBI files, properly extract all image/font resources and only put resources upto the KF8 boundary in the mobi6 part

src/calibre/ebooks/mobi/debug/headers.py

@@ -383,6 +383,13 @@ class MOBIHeader(object):  # {{{
             if hasattr(self, x) and getattr(self, x) != NULL_INDEX:
                 setattr(self, x, self.header_offset+getattr(self, x))
 
+        # Try to find the first non-text record
+        self.first_resource_record = offset + 1 + self.number_of_text_records  # Default to first record after all text records
+        pointer = min(getattr(self, 'first_non_book_record', NULL_INDEX), getattr(self, 'first_image_index', NULL_INDEX))
+        if pointer != NULL_INDEX:
+            self.first_resource_record = max(pointer, self.first_resource_record)
+        self.last_resource_record = NULL_INDEX
+
         if self.has_exth:
             self.exth_offset = 16 + self.length
 
@@ -391,6 +398,10 @@ class MOBIHeader(object):  # {{{
             self.end_of_exth = self.exth_offset + self.exth.length
             self.bytes_after_exth = self.raw[self.end_of_exth:self.fullname_offset]
 
+            if self.exth.kf8_header_index is not None and offset == 0:
+                # MOBI 6 header in a joint file, adjust self.last_resource_record
+                self.last_resource_record = self.exth.kf8_header_index - 2
+
     def __str__(self):
         ans = ['*'*20 + ' MOBI %d Header '%self.file_version+ '*'*20]
 

src/calibre/ebooks/mobi/debug/mobi6.py

@@ -492,7 +492,7 @@ class BinaryRecord(object):  # {{{
         sig = self.raw[:4]
         name = '%06d'%idx
         if sig in {b'FCIS', b'FLIS', b'SRCS', b'DATP', b'RESC', b'BOUN',
-                b'FDST', b'AUDI', b'VIDE',}:
+                b'FDST', b'AUDI', b'VIDE', b'CRES', b'CONT', b'CMET'}:
             name += '-' + sig.decode('ascii')
         elif sig == b'\xe9\x8e\r\n':
             name += '-' + 'EOF'
@@ -743,17 +743,14 @@ class MOBIFile(object):  # {{{
             self.indexing_record_nums |= set(xrange(sir+1, sir+1+numi))
 
         ntr = self.mobi_header.number_of_text_records
-        fntbr = self.mobi_header.first_non_book_record
         fii = self.mobi_header.first_image_index
-        if fntbr == NULL_INDEX:
-            fntbr = len(self.records)
         self.text_records = [TextRecord(r, self.records[r],
             self.mobi_header.extra_data_flags, mf.decompress6) for r in xrange(1,
             min(len(self.records), ntr+1))]
         self.image_records, self.binary_records = [], []
         self.font_records = []
         image_index = 0
-        for i in xrange(fntbr, len(self.records)):
+        for i in xrange(self.mobi_header.first_resource_record, min(self.mobi_header.last_resource_record, len(self.records))):
             if i in self.indexing_record_nums or i in self.huffman_record_nums:
                 continue
             image_index += 1
@@ -761,7 +758,7 @@ class MOBIFile(object):  # {{{
             fmt = None
             if i >= fii and r.raw[:4] not in {b'FLIS', b'FCIS', b'SRCS',
                     b'\xe9\x8e\r\n', b'RESC', b'BOUN', b'FDST', b'DATP',
-                    b'AUDI', b'VIDE', b'FONT'}:
+                    b'AUDI', b'VIDE', b'FONT', b'CRES', b'CONT', b'CMET'}:
                 try:
                     fmt = what(None, r.raw)
                 except:
@@ -832,7 +829,6 @@ def inspect_mobi(mobi_file, ddir):
             rec.dump(tdir)
 
 
-
 # }}}
 
 

src/calibre/ebooks/mobi/debug/mobi8.py

@@ -73,12 +73,11 @@ class MOBIFile(object):
         h, h8 = mf.mobi_header, mf.mobi8_header
         first_text_record = 1
         offset = 0
-        res_end = len(mf.records)
+        self.resource_ranges = [(h8.first_resource_record, h8.last_resource_record, h8.first_image_index)]
         if mf.kf8_type == 'joint':
             offset = h.exth.kf8_header_index
-            res_end = offset - 1
+            self.resource_ranges.insert(0, (h.first_resource_record, h.last_resource_record, h.first_image_index))
 
-        self.resource_records = mf.records[h.first_non_book_record:res_end]
         self.text_records = [TextRecord(i, r, h8.extra_data_flags,
             mf.decompress8) for i, r in
             enumerate(mf.records[first_text_record+offset:
@@ -86,7 +85,7 @@ class MOBIFile(object):
 
         self.raw_text = b''.join(r.raw for r in self.text_records)
         self.header = self.mf.mobi8_header
-        self.extract_resources()
+        self.extract_resources(mf.records)
         self.read_fdst()
         self.read_indices()
         self.build_files()
@@ -151,13 +150,21 @@ class MOBIFile(object):
             with open(os.path.join(ddir, 'flow%04d.txt'%i), 'wb') as f:
                 f.write(raw)
 
-    def extract_resources(self):
+    def extract_resources(self, records):
         self.resource_map = []
         known_types = {b'FLIS', b'FCIS', b'SRCS',
                     b'\xe9\x8e\r\n', b'RESC', b'BOUN', b'FDST', b'DATP',
-                    b'AUDI', b'VIDE'}
+                    b'AUDI', b'VIDE', b'CRES', b'CONT', b'CMET'}
 
-        for i, rec in enumerate(self.resource_records):
+        for i, rec in enumerate(records):
+            for (l, r, offset) in self.resource_ranges:
+                if l <= i <= r:
+                    resource_index = i + 1
+                    if offset is not None and resource_index >= offset:
+                        resource_index -= offset
+                    break
+            else:
+                continue
             sig = rec.raw[:4]
             payload = rec.raw
             ext = 'dat'
@@ -185,7 +192,7 @@ class MOBIFile(object):
                 elif sig in known_types:
                     suffix = '-' + sig.decode('ascii')
 
-            self.resource_map.append(('%s/%06d%s.%s'%(prefix, i, suffix, ext),
+            self.resource_map.append(('%s/%06d%s.%s'%(prefix, resource_index, suffix, ext),
                 payload))
 
     def read_tbs(self):