Merged in shimizukawa/sphinx-multibyte-filename-fork (pull request #61)

5 files changed, 29 insertions, 12 deletions

diff --git a/sphinx/builders/epub.py b/sphinx/builders/epub.py
index b4c3b277..c30fe735 100644
--- a/sphinx/builders/epub.py
+++ b/sphinx/builders/epub.py
@@ -662,7 +662,12 @@ class EpubBuilder(StandaloneHTMLBuilder):
             zipfile.ZIP_STORED)
         for file in projectfiles:
             fp = path.join(outdir, file)
-            if isinstance(fp, unicode):
-                fp = fp.encode(sys.getfilesystemencoding())
+            if sys.version_info < (2, 6):
+                # When zipile.ZipFile.write call with unicode filename, ZipFile
+                # encode filename to 'utf-8' (only after Python-2.6).
+                if isinstance(file, unicode):
+                    # OEBPS Container Format (OCF) 2.0.1 specification require
+                    # "File Names MUST be UTF-8 encoded".
+                    file = file.encode('utf-8')
             epub.write(fp, file, zipfile.ZIP_DEFLATED)
         epub.close()
diff --git a/sphinx/cmdline.py b/sphinx/cmdline.py
index eff89b94..eb05d176 100644
--- a/sphinx/cmdline.py
+++ b/sphinx/cmdline.py
@@ -22,9 +22,17 @@ from sphinx.errors import SphinxError
 from sphinx.application import Sphinx
 from sphinx.util import Tee, format_exception_cut_frames, save_traceback
 from sphinx.util.console import red, nocolor, color_terminal
+from sphinx.util.osutil import fs_encoding
 from sphinx.util.pycompat import terminal_safe, bytes
 
 
+def abspath(pathdir):
+    pathdir = path.abspath(pathdir)
+    if isinstance(pathdir, bytes):
+        pathdir = pathdir.decode(fs_encoding)
+    return pathdir
+
+
 def usage(argv, msg=None):
     if msg:
         print >>sys.stderr, msg
@@ -65,7 +73,7 @@ def main(argv):
     try:
         opts, args = getopt.getopt(argv[1:], 'ab:t:d:c:CD:A:ng:NEqQWw:P')
         allopts = set(opt[0] for opt in opts)
-        srcdir = confdir = path.abspath(args[0])
+        srcdir = confdir = abspath(args[0])
         if not path.isdir(srcdir):
             print >>sys.stderr, 'Error: Cannot find source directory `%s\'.' % (
                                 srcdir,)
@@ -75,7 +83,7 @@ def main(argv):
             print >>sys.stderr, ('Error: Source directory doesn\'t '
                                  'contain conf.py file.')
             return 1
-        outdir = path.abspath(args[1])
+        outdir = abspath(args[1])
         if not path.isdir(outdir):
             print >>sys.stderr, 'Making output directory...'
             os.makedirs(outdir)
@@ -119,9 +127,9 @@ def main(argv):
         elif opt == '-t':
             tags.append(val)
         elif opt == '-d':
-            doctreedir = path.abspath(val)
+            doctreedir = abspath(val)
         elif opt == '-c':
-            confdir = path.abspath(val)
+            confdir = abspath(val)
             if not path.isfile(path.join(confdir, 'conf.py')):
                 print >>sys.stderr, ('Error: Configuration directory '
                                      'doesn\'t contain conf.py file.')
diff --git a/sphinx/config.py b/sphinx/config.py
index eef8d31c..3599ccfc 100644
--- a/sphinx/config.py
+++ b/sphinx/config.py
@@ -16,7 +16,7 @@ from os import path
 
 from sphinx.errors import ConfigError
 from sphinx.locale import l_
-from sphinx.util.osutil import make_filename
+from sphinx.util.osutil import make_filename, fs_encoding
 from sphinx.util.pycompat import bytes, b, convert_with_2to3
 
 nonascii_re = re.compile(b(r'[\x80-\xff]'))
@@ -211,14 +211,15 @@ class Config(object):
                     f.close()
                 try:
                     # compile to a code object, handle syntax errors
+                    config_file_enc = config_file.encode(fs_encoding)
                     try:
-                        code = compile(source, config_file, 'exec')
+                        code = compile(source, config_file_enc, 'exec')
                     except SyntaxError:
                         if convert_with_2to3:
                             # maybe the file uses 2.x syntax; try to refactor to
                             # 3.x syntax using 2to3
                             source = convert_with_2to3(config_file)
-                            code = compile(source, config_file, 'exec')
+                            code = compile(source, config_file_enc, 'exec')
                         else:
                             raise
                     exec code in config
diff --git a/sphinx/environment.py b/sphinx/environment.py
index 1adc6809..cada082d 100644
--- a/sphinx/environment.py
+++ b/sphinx/environment.py
@@ -41,7 +41,8 @@ from sphinx.util import url_re, get_matching_docs, docname_join, split_into, \
      FilenameUniqDict
 from sphinx.util.nodes import clean_astext, make_refnode, extract_messages, \
      WarningStream
-from sphinx.util.osutil import movefile, SEP, ustrftime, find_catalog
+from sphinx.util.osutil import movefile, SEP, ustrftime, find_catalog, \
+     fs_encoding
 from sphinx.util.matching import compile_matchers
 from sphinx.util.pycompat import all, class_types
 from sphinx.util.websupport import is_commentable
@@ -49,7 +50,6 @@ from sphinx.errors import SphinxError, ExtensionError
 from sphinx.locale import _, init as init_locale
 from sphinx.versioning import add_uids, merge_doctrees
 
-fs_encoding = sys.getfilesystemencoding() or sys.getdefaultencoding()
 
 orig_role_function = roles.role
 orig_directive_function = directives.directive
@@ -1373,7 +1373,7 @@ class BuildEnvironment:
         def _entries_from_toctree(toctreenode, parents,
                                   separate=False, subtree=False):
             """Return TOC entries for a toctree node."""
-            refs = [(e[0], str(e[1])) for e in toctreenode['entries']]
+            refs = [(e[0], e[1]) for e in toctreenode['entries']]
             entries = []
             for (title, ref) in refs:
                 try:
diff --git a/sphinx/util/osutil.py b/sphinx/util/osutil.py
index 8ecfe692..17619ee1 100644
--- a/sphinx/util/osutil.py
+++ b/sphinx/util/osutil.py
@@ -157,3 +157,6 @@ def find_catalog(docname, compaction):
         ret = docname
 
     return ret
+
+fs_encoding = sys.getfilesystemencoding() or sys.getdefaultencoding()
+