inspect-mobi: Extract resources from KF8 files

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

@@ -7,9 +7,10 @@ __license__   = 'GPL v3'
 __copyright__ = '2012, Kovid Goyal <kovid@kovidgoyal.net>'
 __docformat__ = 'restructuredtext en'
 
-import sys, os
+import sys, os, imghdr
 
 from calibre.ebooks.mobi.debug.headers import TextRecord
+from calibre.ebooks.mobi.utils import read_font_record
 
 class MOBIFile(object):
 
@@ -30,6 +31,7 @@ class MOBIFile(object):
                 first_text_record+offset+h8.number_of_text_records])]
 
         self.raw_text = b''.join(r.raw for r in self.text_records)
+        self.extract_resources()
 
     def print_header(self, f=sys.stdout):
         print (str(self.mf.palmdb).encode('utf-8'), file=f)
@@ -41,6 +43,42 @@ class MOBIFile(object):
         print (file=f)
         print (str(self.mf.mobi8_header).encode('utf-8'), file=f)
 
+    def extract_resources(self):
+        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'}
+
+        for i, rec in enumerate(self.resource_records):
+            sig = rec.raw[:4]
+            payload = rec.raw
+            ext = 'dat'
+            prefix = 'binary'
+            suffix = ''
+            if sig in {b'HUFF', b'CDIC', b'INDX'}: continue
+            # TODO: Ignore CNCX records as well
+            if sig == b'FONT':
+                font = read_font_record(rec.raw)
+                if font['err']:
+                    raise ValueError('Failed to read font record: %s Headers: %s'%(
+                        font['err'], font['headers']))
+                payload = (font['font_data'] if font['font_data'] else
+                        font['raw_data'])
+                prefix, ext = 'fonts', font['ext']
+            elif sig not in known_types:
+                q = imghdr.what(None, rec.raw)
+                if q:
+                    prefix, ext = 'images', q
+
+            if prefix == 'binary':
+                if sig == b'\xe9\x8e\r\n':
+                    suffix = '-EOF'
+                elif sig in known_types:
+                    suffix = '-' + sig.decode('ascii')
+
+            self.resource_map.append(('%s/%06d%s.%s'%(prefix, i, suffix, ext),
+                payload))
+
 
 def inspect_mobi(mobi_file, ddir):
     f = MOBIFile(mobi_file)
@@ -51,12 +89,14 @@ def inspect_mobi(mobi_file, ddir):
     with open(alltext, 'wb') as of:
         of.write(f.raw_text)
 
-    for tdir, attr in [('text_records', 'text_records'), ('images',
+    for x in ('text_records', 'images', 'fonts', 'binary'):
-        'image_records'), ('binary', 'binary_records'), ('font',
+        os.mkdir(os.path.join(ddir, x))
-            'font_records')]:
+
-        tdir = os.path.join(ddir, tdir)
+    for rec in f.text_records:
-        os.mkdir(tdir)
+        rec.dump(os.path.join(ddir, 'text_records'))
-        for rec in getattr(f, attr, []):
+
-            rec.dump(tdir)
+    for href, payload in f.resource_map:
+        with open(os.path.join(ddir, href), 'wb') as f:
+            f.write(payload)