From 24d8f758e7fa94c17cb749853b3ae32a59c645ef Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Mon, 7 Mar 2011 22:08:48 -0700
Subject: [PATCH] CHM Input: If an input encoding is specified, use it rather
 than trying to detect the encoding of the text in the CHM file. Fixes #9173
 ("Input character encoding" is useless for chm file.)

---
 src/calibre/ebooks/chm/input.py  | 5 +++--
 src/calibre/ebooks/chm/reader.py | 5 ++++-
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/src/calibre/ebooks/chm/input.py b/src/calibre/ebooks/chm/input.py
index 89efa2b4d1..f55a76d67e 100644
--- a/src/calibre/ebooks/chm/input.py
+++ b/src/calibre/ebooks/chm/input.py
@@ -22,7 +22,7 @@ class CHMInput(InputFormatPlugin):
     def _chmtohtml(self, output_dir, chm_path, no_images, log):
         from calibre.ebooks.chm.reader import CHMReader
         log.debug('Opening CHM file')
-        rdr = CHMReader(chm_path, log)
+        rdr = CHMReader(chm_path, log, self.opts)
         log.debug('Extracting CHM to %s' % output_dir)
         rdr.extract_content(output_dir)
         self._chm_reader = rdr
@@ -32,13 +32,13 @@ class CHMInput(InputFormatPlugin):
     def convert(self, stream, options, file_ext, log, accelerators):
         from calibre.ebooks.chm.metadata import get_metadata_from_reader
         from calibre.customize.ui import plugin_for_input_format
+        self.opts = options
 
         log.debug('Processing CHM...')
         with TemporaryDirectory('_chm2oeb') as tdir:
             html_input = plugin_for_input_format('html')
             for opt in html_input.options:
                 setattr(options, opt.option.name, opt.recommended_value)
-            options.input_encoding = 'utf-8'
             no_images = False #options.no_images
             chm_name = stream.name
             #chm_data = stream.read()
@@ -54,6 +54,7 @@ class CHMInput(InputFormatPlugin):
 
             odi = options.debug_pipeline
             options.debug_pipeline = None
+            options.input_encoding = 'utf-8'
             # try a custom conversion:
             #oeb = self._create_oebbook(mainpath, tdir, options, log, metadata)
             # try using html converter:
diff --git a/src/calibre/ebooks/chm/reader.py b/src/calibre/ebooks/chm/reader.py
index fe8b4fdbde..34d228ef3b 100644
--- a/src/calibre/ebooks/chm/reader.py
+++ b/src/calibre/ebooks/chm/reader.py
@@ -40,13 +40,14 @@ class CHMError(Exception):
     pass
 
 class CHMReader(CHMFile):
-    def __init__(self, input, log):
+    def __init__(self, input, log, opts):
         CHMFile.__init__(self)
         if isinstance(input, unicode):
             input = input.encode(filesystem_encoding)
         if not self.LoadCHM(input):
             raise CHMError("Unable to open CHM file '%s'"%(input,))
         self.log = log
+        self.opts = opts
         self._sourcechm = input
         self._contents = None
         self._playorder = 0
@@ -151,6 +152,8 @@ class CHMReader(CHMFile):
                     break
 
     def _reformat(self, data, htmlpath):
+        if self.opts.input_encoding:
+            data = data.decode(self.opts.input_encoding)
         try:
             data = xml_to_unicode(data, strip_encoding_pats=True)[0]
             soup = BeautifulSoup(data)