Remove hardcoding of mbcs for filesystem names

src/calibre/ebooks/metadata/pdf.py

@@ -8,7 +8,7 @@ import os, subprocess, shutil, re
 from functools import partial
 
 from calibre import prints
-from calibre.constants import iswindows, ispy3
+from calibre.constants import iswindows
 from calibre.ptempfile import TemporaryDirectory
 from calibre.ebooks.metadata import (
     MetaInformation, string_to_authors, check_isbn, check_doi)
@@ -99,8 +99,6 @@ def page_images(pdfpath, outputdir='.', first=1, last=1, image_format='jpeg', pr
 
 
 def is_pdf_encrypted(path_to_pdf):
-    if not ispy3 and not isinstance(path_to_pdf, bytes):
-        path_to_pdf = path_to_pdf.encode('mbcs' if iswindows else 'utf-8')
     pdfinfo = get_tools()[0]
     raw = subprocess.check_output([pdfinfo, path_to_pdf])
     q = re.search(br'^Encrypted:\s*(\S+)', raw, flags=re.MULTILINE)

src/calibre/gui2/preferences/server.py

@@ -2,8 +2,6 @@
 # vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
 # License: GPLv3 Copyright: 2010, Kovid Goyal <kovid at kovidgoyal.net>
 
-
-
 import errno
 import json
 import numbers
@@ -50,7 +48,7 @@ if iswindows and not isportable:
         exe_base = os.path.abspath(os.path.dirname(sys.executable))
         exe = os.path.join(exe_base, 'calibre.exe')
         if isinstance(exe, bytes):
-            exe = exe.decode('mbcs')
+            exe = os.fsdecode(exe)
         return exe
 
     def startup_shortcut_path():

src/calibre/gui2/win_file_dialogs.py

@@ -30,10 +30,8 @@ def is_ok():
 
 
 try:
-    from calibre.constants import filesystem_encoding
     from calibre.utils.config import dynamic
 except ImportError:
-    filesystem_encoding = 'mbcs'
     dynamic = {}
 
 
@@ -112,7 +110,7 @@ class Loop(QEventLoop):
 
 def process_path(x):
     if isinstance(x, bytes):
-        x = x.decode(filesystem_encoding)
+        x = os.fsdecode(x)
     return os.path.abspath(os.path.expanduser(x))
 
 
@@ -167,7 +165,7 @@ def run_file_dialog(
             data.append(serialize_string('FOLDER', initial_folder))
     if filename:
         if isinstance(filename, bytes):
-            filename = filename.decode(filesystem_encoding)
+            filename = os.fsdecode(filename)
         data.append(serialize_string('FILENAME', filename))
     if only_dirs:
         file_types = ()  # file types not allowed for dir only dialogs

src/calibre/ptempfile.py

@@ -125,7 +125,7 @@ def base_dir():
 
         try:
             tempfile.gettempdir()
-        except:
+        except Exception:
             # Widows temp vars set to a path not encodable in mbcs
             # Use our temp dir
             tempfile.tempdir = _base_dir

src/calibre/srv/utils.py

@@ -20,7 +20,7 @@ from calibre.utils.shared_file import share_open, raise_winerror
 from polyglot.builtins import iteritems, map, range
 from polyglot import reprlib
 from polyglot.http_cookie import SimpleCookie
-from polyglot.builtins import is_py3, unicode_type, as_bytes, as_unicode
+from polyglot.builtins import unicode_type, as_unicode
 from polyglot.urllib import parse_qs, quote as urlquote
 from polyglot.binary import as_hex_unicode as encode_name, from_hex_unicode as decode_name
 
@@ -523,10 +523,5 @@ def get_use_roman():
     return _use_roman
 
 
-if iswindows and not is_py3:
+def fast_now_strftime(fmt):
-    def fast_now_strftime(fmt):
-        fmt = as_bytes(fmt, encoding='mbcs')
-        return time.strftime(fmt).decode('mbcs', 'replace')
-else:
-    def fast_now_strftime(fmt):
     return as_unicode(time.strftime(fmt), errors='replace')

src/calibre/startup.py

@@ -7,7 +7,7 @@ __docformat__ = 'restructuredtext en'
 Perform various initialization tasks.
 '''
 
-import locale, sys, os
+import locale, sys
 
 # Default translation is NOOP
 from polyglot.builtins import builtins, is_py3, unicode_type
@@ -21,7 +21,7 @@ builtins.__dict__['__'] = lambda s: s
 builtins.__dict__['dynamic_property'] = lambda func: func(None)
 
 
-from calibre.constants import iswindows, preferred_encoding, plugins, isosx, islinux, isfrozen, DEBUG, isfreebsd, ispy3
+from calibre.constants import iswindows, preferred_encoding, plugins, isosx, islinux, DEBUG, isfreebsd
 
 _run_once = False
 winutil = winutilerror = None
@@ -30,23 +30,8 @@ if not _run_once:
     _run_once = True
     from importlib import import_module
 
-    if not isfrozen and not ispy3:
-        # Prevent PyQt4 from being loaded
-        class PyQt4Ban(object):
-
-            def find_module(self, fullname, path=None):
-                if fullname.startswith('PyQt4'):
-                    return self
-
-            def load_module(self, fullname):
-                raise ImportError('Importing PyQt4 is not allowed as calibre uses PyQt5')
-
-        sys.meta_path.insert(0, PyQt4Ban())
-
     class DeVendor(object):
 
-        if ispy3:
-
         def find_spec(self, fullname, path, target=None):
             spec = None
             if fullname == 'calibre.web.feeds.feedparser':
@@ -57,17 +42,6 @@ if not _run_once:
                 spec = m.__spec__
             return spec
 
-        else:
-
-            def find_module(self, fullname, path=None):
-                if fullname == 'calibre.web.feeds.feedparser' or fullname.startswith('calibre.ebooks.markdown'):
-                    return self
-
-            def load_module(self, fullname):
-                if fullname == 'calibre.web.feeds.feedparser':
-                    return import_module('feedparser')
-                return import_module(fullname[len('calibre.ebooks.'):])
-
     sys.meta_path.insert(0, DeVendor())
 
     #
@@ -79,25 +53,6 @@ if not _run_once:
         if len(sys.argv) > 1 and not isinstance(sys.argv[1], unicode_type):
             sys.argv[1:] = winutil.argv()[1-len(sys.argv):]
 
-        if not ispy3:
-            # Python2's expanduser is broken for non-ASCII usernames
-            # and unicode paths
-
-            def expanduser(path):
-                if isinstance(path, bytes):
-                    path = path.decode('mbcs')
-                if path[:1] != '~':
-                    return path
-                i, n = 1, len(path)
-                while i < n and path[i] not in '/\\':
-                    i += 1
-                userhome = winutil.special_folder_path(winutil.CSIDL_PROFILE)
-                if i != 1:  # ~user
-                    userhome = os.path.join(os.path.dirname(userhome), path[1:i])
-
-                return userhome + path[i:]
-            os.path.expanduser = expanduser
-
     # Ensure that all temp files/dirs are created under a calibre tmp dir
     from calibre.ptempfile import base_dir
     try:

src/calibre/utils/config_base.py

@@ -1,8 +1,6 @@
 #!/usr/bin/env python
 # vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
 
-
-
 __license__   = 'GPL v3'
 __copyright__ = '2011, Kovid Goyal <kovid@kovidgoyal.net>'
 __docformat__ = 'restructuredtext en'

src/calibre/utils/filenames.py

@@ -490,9 +490,9 @@ if iswindows:
     def rename_file(a, b):
         move_file = plugins['winutil'][0].move_file
         if isinstance(a, bytes):
-            a = a.decode('mbcs')
+            a = os.fsdecode(a)
         if isinstance(b, bytes):
-            b = b.decode('mbcs')
+            b = os.fsdecode(b)
         move_file(a, b)
 
 

src/calibre/utils/img.py

@@ -17,7 +17,7 @@ from PyQt5.QtCore import QBuffer, QByteArray, Qt
 from PyQt5.QtGui import QColor, QImage, QImageReader, QImageWriter, QPixmap, QTransform
 
 from calibre import fit_image, force_unicode
-from calibre.constants import iswindows, plugins, ispy3
+from calibre.constants import iswindows, plugins
 from calibre.ptempfile import TemporaryDirectory
 from calibre.utils.config_base import tweaks
 from calibre.utils.filenames import atomic_rename
@@ -53,8 +53,8 @@ def get_exe_path(name):
 
 def load_jxr_data(data):
     with TemporaryDirectory() as tdir:
-        if iswindows and isinstance(tdir, unicode_type):
+        if isinstance(tdir, bytes):
-            tdir = tdir.encode('mbcs')
+            tdir = os.fsdecode(tdir)
         with lopen(os.path.join(tdir, 'input.jxr'), 'wb') as f:
             f.write(data)
         cmd = [get_exe_path('JxrDecApp'), '-i', 'input.jxr', '-o', 'output.tif']
@@ -542,13 +542,6 @@ def run_optimizer(file_path, cmd, as_filter=False, input_data=None):
             cmd[cmd.index(q)] = r
         if not as_filter:
             repl(True, iname), repl(False, oname)
-        if iswindows and not ispy3:
-            # subprocess in python 2 cannot handle unicode strings that are not
-            # encodeable in mbcs, so we fail here, where it is more explicit,
-            # instead.
-            cmd = [x.encode('mbcs') if isinstance(x, unicode_type) else x for x in cmd]
-            if isinstance(cwd, unicode_type):
-                cwd = cwd.encode('mbcs')
         stdin = subprocess.PIPE if as_filter else None
         stderr = subprocess.PIPE if as_filter else subprocess.STDOUT
         creationflags = 0x08 if iswindows else 0

src/calibre/utils/lock.py

@@ -2,8 +2,6 @@
 # vim:fileencoding=utf-8
 # License: GPLv3 Copyright: 2017, Kovid Goyal <kovid at kovidgoyal.net>
 
-
-
 import atexit
 import errno
 import os
@@ -50,7 +48,7 @@ def unix_retry(err):
 
 def windows_open(path):
     if isinstance(path, bytes):
-        path = path.decode('mbcs')
+        path = os.fsdecode(path)
     try:
         h = win32file.CreateFileW(
             path,

src/polyglot/builtins.py

@@ -2,8 +2,6 @@
 # vim:fileencoding=utf-8
 # License: GPL v3 Copyright: 2018, Kovid Goyal <kovid at kovidgoyal.net>
 
-
-
 import os
 import sys
 
@@ -46,12 +44,15 @@ def only_unicode_recursive(x, encoding='utf-8', errors='strict'):
     if isinstance(x, (set, list, tuple, frozenset)):
         return type(x)(only_unicode_recursive(i, encoding, errors) for i in x)
     if isinstance(x, dict):
-        return {only_unicode_recursive(k, encoding, errors): only_unicode_recursive(v, encoding, errors) for k, v in iteritems(x)}
+        return {
+            only_unicode_recursive(k, encoding, errors):
+            only_unicode_recursive(v, encoding, errors)
+            for k, v in iteritems(x)
+        }
     return x
 
 
-if is_py3:
+def reraise(tp, value, tb=None):
-    def reraise(tp, value, tb=None):
     try:
         if value is None:
             value = tp()
@@ -62,141 +63,61 @@ if is_py3:
         value = None
         tb = None
 
-    import builtins
 
-    zip = builtins.zip
+import builtins
-    map = builtins.map
-    filter = builtins.filter
-    range = builtins.range
 
-    codepoint_to_chr = chr
+zip = builtins.zip
-    unicode_type = str
+map = builtins.map
-    string_or_bytes = str, bytes
+filter = builtins.filter
-    string_or_unicode = str
+range = builtins.range
-    long_type = int
-    raw_input = input
-    getcwd = os.getcwd
-    getenv = os.getenv
 
-    def error_message(exc):
+codepoint_to_chr = chr
+unicode_type = str
+string_or_bytes = str, bytes
+string_or_unicode = str
+long_type = int
+raw_input = input
+getcwd = os.getcwd
+getenv = os.getenv
+
+
+def error_message(exc):
     args = getattr(exc, 'args', None)
     if args and isinstance(args[0], unicode_type):
         return args[0]
     return unicode_type(exc)
 
-    def iteritems(d):
+
+def iteritems(d):
     return iter(d.items())
 
-    def itervalues(d):
+
+def itervalues(d):
     return iter(d.values())
 
-    def environ_item(x):
+
+def environ_item(x):
     if isinstance(x, bytes):
         x = x.decode('utf-8')
     return x
 
-    def exec_path(path, ctx=None):
+
+def exec_path(path, ctx=None):
     ctx = ctx or {}
     with open(path, 'rb') as f:
         code = f.read()
     code = compile(code, f.name, 'exec')
     exec(code, ctx)
 
-    def cmp(a, b):
+
+def cmp(a, b):
     return (a > b) - (a < b)
 
-    def int_to_byte(x):
-        return bytes((x,))
 
-    def reload(module):
+def int_to_byte(x):
+    return bytes((x, ))
+
+
+def reload(module):
     import importlib
     return importlib.reload(module)
-
-else:
-    exec("""def reraise(tp, value, tb=None):
-    try:
-        raise tp, value, tb
-    finally:
-        tb = None
-""")
-
-    from future_builtins import zip, map, filter  # noqa
-    range = xrange
-    import __builtin__ as builtins
-
-    codepoint_to_chr = unichr
-    unicode_type = unicode
-    string_or_bytes = unicode, bytes
-    string_or_unicode = str, unicode
-    long_type = long
-    exec_path = execfile
-    raw_input = builtins.raw_input
-    cmp = builtins.cmp
-    int_to_byte = chr
-    getcwd = os.getcwdu
-
-    def error_message(exc):
-        ans = exc.message
-        if isinstance(ans, bytes):
-            ans = ans.decode('utf-8', 'replace')
-        return ans
-
-    def iteritems(d):
-        return d.iteritems()
-
-    def itervalues(d):
-        return d.itervalues()
-
-    def environ_item(x):
-        if isinstance(x, unicode_type):
-            x = x.encode('utf-8')
-        return x
-
-    if hasattr(sys, 'getwindowsversion'):
-        def getenv(x, default=None):
-            if isinstance(x, bytes):
-                x = x.decode('mbcs', 'replace')
-
-            if getenv.buf is None:
-                import ctypes
-                import ctypes.wintypes as w
-                getenv.cub = ctypes.create_unicode_buffer
-                getenv.buf = getenv.cub(1024)
-                getenv.gev = ctypes.windll.kernel32.GetEnvironmentVariableW
-                getenv.gev.restype = w.DWORD
-                getenv.gev.argtypes = [w.LPCWSTR, w.LPWSTR, w.DWORD]
-            res = getenv.gev(x, getenv.buf, len(getenv.buf))
-            if res == 0:
-                return default
-            if res > len(getenv.buf) - 4:
-                getenv.buf = getenv.cub(res + 8)
-                res = getenv.gev(x, getenv.buf, len(getenv.buf))
-                if res == 0:
-                    return default
-            return getenv.buf.value
-        getenv.buf = None
-    else:
-        def getenv(x, default=None):
-            ans = os.getenv(x, default)
-            if isinstance(ans, bytes):
-                ans = ans.decode('utf-8', 'replace')
-            return ans
-
-    def reload(module):
-        return builtins.reload(module)
-
-
-def print_to_binary_file(fileobj, encoding='utf-8'):
-
-    def print(*a, **kw):
-        f = kw.get('file', fileobj)
-        if a:
-            sep = as_bytes(kw.get('sep', ' '), encoding)
-            for x in a:
-                x = as_bytes(x, encoding)
-                f.write(x)
-                if x is not a[-1]:
-                    f.write(sep)
-        f.write(as_bytes(kw.get('end', '\n')))
-
-    return print