Fix bug 3313 (unicode error when invoking lyx2lyx) by Anders Ekberg

[lyx.git] / lib / lyx2lyx / LyX.py

diff --git a/lib/lyx2lyx/LyX.py b/lib/lyx2lyx/LyX.py
index c9c5b38021a35b8bc88fc4a45830959ea571144f..0a9e44c6d892a43a613c172b27b5ffdb1cf113b5 100644 (file)
--- a/lib/lyx2lyx/LyX.py
+++ b/lib/lyx2lyx/LyX.py
@@ -21,12 +21,16 @@ from parser_tools import get_value, check_token, find_token,\
      find_tokens, find_end_of
 import os.path
 import gzip
+import locale
 import sys
 import re
 import time
 
-import lyx2lyx_version
-version_lyx2lyx = lyx2lyx_version.version
+try:
+    import lyx2lyx_version
+    version_lyx2lyx = lyx2lyx_version.version
+except: # we are running from build directory so assume the last version
+    version_lyx2lyx = '1.5.0svn'
 
 default_debug_level = 2
 
@@ -72,8 +76,8 @@ format_relation = [("0_06",    [200], generate_minor_versions("0.6" , 4)),
                    ("1_1_6_3", [218], ["1.1.6.3","1.1.6.4","1.1"]),
                    ("1_2",     [220], generate_minor_versions("1.2" , 4)),
                    ("1_3",     [221], generate_minor_versions("1.3" , 7)),
-                   ("1_4", range(222,246), generate_minor_versions("1.4" , 3)),
-                   ("1_5", range(246,249), generate_minor_versions("1.5" , 0))]
+                   ("1_4", range(222,246), generate_minor_versions("1.4" , 4)),
+                   ("1_5", range(246,277), generate_minor_versions("1.5" , 0))]
 
 
 def formats_list():
@@ -108,11 +112,21 @@ def trim_eol(line):
         return line[:-1]
 
 
-def get_encoding(language, inputencoding):
+def get_encoding(language, inputencoding, format, cjk_encoding):
+    if format > 248:
+        return "utf8"
+    # CJK-LyX encodes files using the current locale encoding.
+    # This means that files created by CJK-LyX can only be converted using
+    # the correct locale settings unless the encoding is given as commandline
+    # argument.
+    if cjk_encoding == 'auto':
+        return locale.getpreferredencoding()
+    elif cjk_encoding != '':
+        return cjk_encoding
     from lyx2lyx_lang import lang
-    if inputencoding == "auto":        
+    if inputencoding == "auto" or inputencoding == "default":
         return lang[language][3]
-    if inputencoding == "default" or inputencoding == "":
+    if inputencoding == "":
         return "latin1"
     # python does not know the alias latin9
     if inputencoding == "latin9":
@@ -124,7 +138,11 @@ def get_encoding(language, inputencoding):
 #
 class LyX_Base:
     """This class carries all the information of the LyX file."""
-    def __init__(self, end_format = 0, input = "", output = "", error = "", debug = default_debug_level, try_hard = 0):
+    
+    def __init__(self, end_format = 0, input = "", output = "", error
+                 = "", debug = default_debug_level, try_hard = 0, cjk_encoding = '',
+                 language = "english", encoding = "auto"):
+
         """Arguments:
         end_format: final format that the file should be converted. (integer)
         input: the name of the input source, if empty resort to standard input.
@@ -141,6 +159,7 @@ class LyX_Base:
 
         self.debug = debug
         self.try_hard = try_hard
+        self.cjk_encoding = cjk_encoding
 
         if end_format:
             self.end_format = self.lyxformat(end_format)
@@ -157,7 +176,8 @@ class LyX_Base:
         self.preamble = []
         self.body = []
         self.status = 0
-        self.encoding = "auto"
+        self.encoding = encoding
+        self.language = language
 
 
     def warning(self, message, debug_level= default_debug_level):
@@ -219,7 +239,7 @@ class LyX_Base:
         self.format  = self.read_format()
         self.language = get_value(self.header, "\\language", 0, default = "english")
         self.inputencoding = get_value(self.header, "\\inputencoding", 0, default = "auto")
-        self.encoding = get_encoding(self.language, self.inputencoding)
+        self.encoding = get_encoding(self.language, self.inputencoding, self.format, self.cjk_encoding)
         self.initial_version = self.read_version()
 
         # Second pass over header and preamble, now we know the file encoding
@@ -240,6 +260,9 @@ class LyX_Base:
         " Writes the LyX file to self.output."
         self.set_version()
         self.set_format()
+        self.set_textclass()
+        if self.encoding == "auto":
+            self.encoding = get_encoding(self.language, self.encoding, self.format, self.cjk_encoding)
 
         if self.preamble:
             i = find_token(self.header, '\\textclass', 0) + 1
@@ -345,6 +368,11 @@ class LyX_Base:
         self.header[i] = "\\lyxformat %s" % format
 
 
+    def set_textclass(self):
+        i = find_token(self.header, "\\textclass", 0)
+        self.header[i] = "\\textclass %s" % self.textclass
+
+
     def set_parameter(self, param, value):
         " Set the value of the header parameter."
         i = find_token(self.header, '\\' + param, 0)
@@ -523,8 +551,8 @@ class LyX_Base:
 
 class File(LyX_Base):
     " This class reads existing LyX files."
-    def __init__(self, end_format = 0, input = "", output = "", error = "", debug = default_debug_level, try_hard = 0):
-        LyX_Base.__init__(self, end_format, input, output, error, debug, try_hard)
+    def __init__(self, end_format = 0, input = "", output = "", error = "", debug = default_debug_level, try_hard = 0, cjk_encoding = ''):
+        LyX_Base.__init__(self, end_format, input, output, error, debug, try_hard, cjk_encoding)
         self.read()