AZW3: Preserve the page-progression-direction property when converting/editing/polishing AZW3 files

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

@@ -32,6 +32,7 @@ class EXTHHeader(object):  # {{{
         left = self.num_items
         self.kf8_header = None
         self.uuid = self.cdetype = None
+        self.page_progression_direction = None
 
         self.decode = lambda x : clean_ascii_chars(x.decode(codec, 'replace'))
 
@@ -81,6 +82,13 @@ class EXTHHeader(object):  # {{{
                         self.mi.language = lang
                 except:
                     pass
+            elif idx == 527:
+                try:
+                    ppd = content.decode(codec)
+                    if ppd:
+                        self.page_progression_direction = ppd
+                except Exception:
+                    pass
             # else:
             #    print 'unknown record', idx, repr(content)
         if title:

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

@@ -20,7 +20,7 @@ from calibre.ebooks.mobi.reader.ncx import read_ncx, build_toc
 from calibre.ebooks.mobi.reader.markup import expand_mobi8_markup
 from calibre.ebooks.metadata.opf2 import Guide, OPFCreator
 from calibre.ebooks.metadata.toc import TOC
-from calibre.ebooks.mobi.utils import read_font_record, read_resc_record
+from calibre.ebooks.mobi.utils import read_font_record
 from calibre.ebooks.oeb.parse_utils import parse_html
 from calibre.ebooks.oeb.base import XPath, XHTML, xml2text
 from calibre.utils.imghdr import what
@@ -66,7 +66,6 @@ class Mobi8Reader(object):
         self.mobi6_reader, self.log = mobi6_reader, log
         self.header = mobi6_reader.book_header
         self.encrypted_fonts = []
-        self.resc_data = {}
 
     def __call__(self):
         self.mobi6_reader.check_for_drm()
@@ -398,10 +397,8 @@ class Mobi8Reader(object):
             typ = data[:4]
             href = None
             if typ in {b'FLIS', b'FCIS', b'SRCS', b'\xe9\x8e\r\n', b'BOUN',
-                       b'FDST', b'DATP', b'AUDI', b'VIDE'}:
+                       b'FDST', b'DATP', b'AUDI', b'VIDE', b'RESC', b'CMET'}:
                 pass  # Ignore these records
-            elif typ == b'RESC':
-                self.resc_data = read_resc_record(data)
             elif typ == b'FONT':
                 font = read_font_record(data)
                 href = "fonts/%05d.%s" % (fname_idx, font['ext'])
@@ -480,8 +477,8 @@ class Mobi8Reader(object):
                 entry.mime_type = 'application/xhtml+xml'
         opf.create_spine(spine)
         opf.set_toc(toc)
-        ppd = self.resc_data.get('page-progression-direction', None)
+        ppd = getattr(self.header.exth, 'page_progression_direction', None)
-        if ppd:
+        if ppd in {'ltr', 'rtl', 'default'}:
             opf.page_progression_direction = ppd
 
         with open('metadata.opf', 'wb') as of, open('toc.ncx', 'wb') as ncx:

src/calibre/ebooks/mobi/utils.py

@@ -7,7 +7,7 @@ __license__   = 'GPL v3'
 __copyright__ = '2011, Kovid Goyal <kovid@kovidgoyal.net>'
 __docformat__ = 'restructuredtext en'
 
-import struct, string, zlib, os, re
+import struct, string, zlib, os
 from collections import OrderedDict
 from io import BytesIO
 
@@ -395,15 +395,6 @@ def mobify_image(data):
         data = im.export('gif')
     return data
 
-def read_resc_record(data):
-    ans = {}
-    match = re.search(br'''<spine [^>]*page-progression-direction=['"](.+?)['"]''', data)
-    if match is not None:
-        ppd = match.group(1).lower()
-        if ppd in {b'ltr', b'rtl'}:
-            ans['page-progression-direction'] = ppd.decode('ascii')
-    return ans
-
 # Font records {{{
 def read_font_record(data, extent=1040):
     '''

src/calibre/ebooks/mobi/writer8/exth.py

@@ -40,6 +40,7 @@ EXTH_CODES = {
     'lastupdatetime': 502,
     'title': 503,
     'language': 524,
+    'page_progression_direction': 527,
 }
 
 COLLAPSE_RE = re.compile(r'[ \t\r\n\v]+')
@@ -47,7 +48,8 @@ COLLAPSE_RE = re.compile(r'[ \t\r\n\v]+')
 def build_exth(metadata, prefer_author_sort=False, is_periodical=False,
         share_not_sync=True, cover_offset=None, thumbnail_offset=None,
         start_offset=None, mobi_doctype=2, num_of_resources=None,
-        kf8_unknown_count=0, be_kindlegen2=False, kf8_header_index=None):
+        kf8_unknown_count=0, be_kindlegen2=False, kf8_header_index=None,
+        page_progression_direction=None):
     exth = BytesIO()
     nrecs = 0
 
@@ -205,6 +207,12 @@ def build_exth(metadata, prefer_author_sort=False, is_periodical=False,
             kf8_unknown_count))
         nrecs += 1
 
+    if page_progression_direction in {'rtl', 'ltr', 'default'}:
+        ppd = bytes(page_progression_direction)
+        exth.write(pack(b'>II', EXTH_CODES['page_progression_direction'], len(ppd) + 8))
+        exth.write(ppd)
+        nrecs += 1
+
     exth = exth.getvalue()
     trail = len(exth) % 4
     pad = b'\0' * (4 - trail)  # Always pad w/ at least 1 byte

src/calibre/ebooks/mobi/writer8/mobi.py

@@ -219,6 +219,7 @@ class KF8Book(object):
     def __init__(self, writer, for_joint=False):
         self.build_records(writer, for_joint)
         self.used_images = writer.used_images
+        self.page_progression_direction = writer.oeb.spine.page_progression_direction
 
     def build_records(self, writer, for_joint):
         metadata = writer.oeb.metadata
@@ -297,15 +298,18 @@ class KF8Book(object):
         code to customize various values after build_records() has been
         called'''
         opts = self.opts
-        self.exth = build_exth(self.metadata,
+        self.exth = build_exth(
-                prefer_author_sort=opts.prefer_author_sort,
+            self.metadata,
-                is_periodical=opts.mobi_periodical,
+            prefer_author_sort=opts.prefer_author_sort,
-                share_not_sync=opts.share_not_sync,
+            is_periodical=opts.mobi_periodical,
-                cover_offset=self.cover_offset,
+            share_not_sync=opts.share_not_sync,
-                thumbnail_offset=self.thumbnail_offset,
+            cover_offset=self.cover_offset,
-                num_of_resources=self.num_of_resources,
+            thumbnail_offset=self.thumbnail_offset,
-                kf8_unknown_count=self.kuc, be_kindlegen2=True,
+            num_of_resources=self.num_of_resources,
-                start_offset=self.start_offset, mobi_doctype=self.book_type)
+            kf8_unknown_count=self.kuc, be_kindlegen2=True,
+            start_offset=self.start_offset, mobi_doctype=self.book_type,
+            page_progression_direction=self.page_progression_direction
+        )
 
         kwargs = {field:getattr(self, field) for field in HEADER_FIELDS}
         return MOBIHeader()(**kwargs)