doc/translator.py unified for Python 2.6+ and Python 3.0+

- Open() replaced by xopen() that internally uses the encoding argument
  only for Python 3. The earlier usage of the codecs module and of the
  prefixed unicode string literals was removed.
- Some lists of prototypes were sorted to get the same translator report
  for different versions of Python.
- The local dedent() definition was replaced by textwrap.dedent().
  (The older versions of Python did not have the module.)
- Tested on Windows for Python 2.6.0, 2.7.7, 3.0.1, and 3.4.1.

doc/translator.py

@@ -64,8 +64,11 @@
   2013/06/25 - TranslatorDecoder checks removed after removing the class.
   2013/09/04 - Coloured status in langhowto. *ALMOST up-to-date* category
                of translators introduced.
+  2014/06/16 - unified for Python 2.6+ and 3.0+
   """
 
+from __future__ import print_function
+
 import os
 import platform
 import re
@@ -74,23 +77,17 @@ import textwrap
 
 
 def xopen(fname, mode='r', encoding='utf-8-sig'):
-    '''Unified open of text files with UTF-8 default encoding.
+    '''Unified file opening for Python 2 an Python 3.
 
-    The 'utf-8-sig' skips the BOM automatically.
+    Python 2 does not have the encoding argument. Python 3 has one, and
+    the default 'utf-8-sig' is used (skips the BOM automatically).
     '''
 
-    # Use UTF-8 without BOM when writing to a text file.
-    if encoding == 'utf-8-sig' and mode == 'w':
-        encoding = 'utf-8'
-
     major, minor, patch = (int(e) for e in platform.python_version_tuple())
     if major == 2:
-        if mode == 'w':
-            mode = 'wU'
-        import codecs
-        return codecs.open(fname, mode=mode, encoding=encoding) # Python 2
+        return open(fname, mode=mode) # Python 2 without encoding
     else:
-        return open(fname, mode=mode, encoding=encoding)        # Python 3
+        return open(fname, mode=mode, encoding=encoding) # Python 3 with encoding
 
 
 def fill(s):
@@ -1074,6 +1071,7 @@ class Transl:
         for p in myDic:
             if p not in reqDic:
                 self.obsoleteMethods.append(p)
+        self.obsoleteMethods.sort()
 
         # Build the list of missing methods and the list of implemented
         # required methods.
@@ -1084,6 +1082,8 @@ class Transl:
                 self.implementedMethods.append(p)
             else:
                 self.missingMethods.append(p)
+        self.missingMethods.sort()
+        self.implementedMethods.sort()
 
         # Check whether adapter must be used or suggest the newest one.
         # Change the status and set the note accordingly.
@@ -1788,7 +1788,7 @@ class TrManager:
         doctpl = f.read()
         f.close()
 
-        pos = doctpl.find(u'/***')
+        pos = doctpl.find('/***')
         assert pos != -1
         doctpl = doctpl[pos:]
 
@@ -1796,21 +1796,21 @@ class TrManager:
         # document template.
         tplDic = {}
 
-        s = u'Do not edit this file. It was generated by the %s script. * Instead edit %s and %s' % (self.script_name, self.languageTplFileName, self.maintainersFileName)
+        s = 'Do not edit this file. It was generated by the %s script. * Instead edit %s and %s' % (self.script_name, self.languageTplFileName, self.maintainersFileName)
         tplDic['editnote'] = s
 
         tplDic['doxVersion'] = self.doxVersion
         tplDic['supportedLangReadableStr'] = self.supportedLangReadableStr
         tplDic['translatorReportFileName'] = self.translatorReportFileName
 
-        ahref = u'<a href="../doc/' + self.translatorReportFileName
-        ahref += u'"\n><code>doxygen/doc/'  + self.translatorReportFileName
-        ahref += u'</code></a>'
+        ahref = '<a href="../doc/' + self.translatorReportFileName
+        ahref += '"\n><code>doxygen/doc/'  + self.translatorReportFileName
+        ahref += '</code></a>'
         tplDic['translatorReportLink'] = ahref
         tplDic['numLangStr'] = str(self.numLang)
 
         # Define templates for HTML table parts of the documentation.
-        htmlTableTpl = u'''\
+        htmlTableTpl = '''\
             \\htmlonly
             <table align="center" cellspacing="0" cellpadding="0" border="0">
             <tr bgcolor="#000000">
@@ -1833,9 +1833,9 @@ class TrManager:
             \\endhtmlonly
             '''
         htmlTableTpl = textwrap.dedent(htmlTableTpl)
-        htmlTrTpl = u'\n  <tr bgcolor="#ffffff">%s\n  </tr>'
-        htmlTdTpl = u'\n    <td>%s</td>'
-        htmlTdStatusColorTpl = u'\n    <td bgcolor="%s">%s</td>'
+        htmlTrTpl = '\n  <tr bgcolor="#ffffff">%s\n  </tr>'
+        htmlTdTpl = '\n    <td>%s</td>'
+        htmlTdStatusColorTpl = '\n    <td bgcolor="%s">%s</td>'
 
         # Loop through transl objects in the order of sorted readable names
         # and add generate the content of the HTML table.
@@ -1848,7 +1848,7 @@ class TrManager:
             if obj.readableStatus.startswith('1.4'):
                 bkcolor = self.getBgcolorByReadableStatus('1.4')
             else:
-                bkcolor = u'#ffffff'
+                bkcolor = '#ffffff'
 
             lst = [ htmlTdStatusColorTpl % (bkcolor, obj.langReadable) ]
 
@@ -1863,8 +1863,8 @@ class TrManager:
                 classId = obj.classId[:-2]
                 if classId in self.__translDic:
                     lang = self.__translDic[classId].langReadable
-                    mm = u'see the %s language' % lang
-                    ee = u'&nbsp;'
+                    mm = 'see the %s language' % lang
+                    ee = '&nbsp;'
 
             if not mm and obj.classId in self.__maintainersDic:
                 # Build a string of names separated by the HTML break element.
@@ -1873,24 +1873,24 @@ class TrManager:
                 for maintainer in self.__maintainersDic[obj.classId]:
                     name = maintainer[0]
                     if name.startswith('--'):
-                        name = u'<span style="color: red; background-color: yellow">'\
-                               + name + u'</span>'
+                        name = '<span style="color: red; background-color: yellow">'\
+                               + name + '</span>'
                     lm.append(name)
                 mm = '<br/>'.join(lm)
 
                 # The marked adresses (they start with the mark '[unreachable]',
                 # '[resigned]', whatever '[xxx]') will not be displayed at all.
                 # Only the mark will be used instead.
-                rexMark = re.compile(u'(?P<mark>\\[.*?\\])')
+                rexMark = re.compile('(?P<mark>\\[.*?\\])')
                 le = []
                 for maintainer in self.__maintainersDic[obj.classId]:
                     address = maintainer[1]
                     m = rexMark.search(address)
                     if m is not None:
-                        address = u'<span style="color: brown">'\
-                                  + m.group('mark') + u'</span>'
+                        address = '<span style="color: brown">'\
+                                  + m.group('mark') + '</span>'
                     le.append(address)
-                ee = u'<br/>'.join(le)
+                ee = '<br/>'.join(le)
 
             # Append the maintainer and e-mail elements.
             lst.append(htmlTdTpl % mm)
@@ -1907,7 +1907,7 @@ class TrManager:
         htmlTable = htmlTableTpl % (''.join(trlst))
 
         # Define templates for LaTeX table parts of the documentation.
-        latexTableTpl = b'''
+        latexTableTpl = r'''
             \latexonly
             \footnotesize
             \begin{longtable}{|l|l|l|l|}
@@ -1919,9 +1919,9 @@ class TrManager:
             \end{longtable}
             \normalsize
             \endlatexonly
-            '''.decode('utf_8')
+            '''
         latexTableTpl = textwrap.dedent(latexTableTpl)
-        latexLineTpl = u'\n  %s & %s & {\\tt\\tiny %s} & %s \\\\'
+        latexLineTpl = '\n' + r'  %s & %s & {\tt\tiny %s} & %s \\'
 
         # Loop through transl objects in the order of sorted readable names
         # and add generate the content of the LaTeX table.
@@ -1956,9 +1956,9 @@ class TrManager:
             # Use the template to produce the line of the table and insert
             # the hline plus the constructed line into the table content.
             # The underscore character must be escaped.
-            trlst.append(u'\n  \\hline')
+            trlst.append('\n  \\hline')
             s = latexLineTpl % (lang, maintainer, email, status)
-            s = s.replace(u'_', u'\\_')
+            s = s.replace('_', '\\_')
             trlst.append(s)
 
             # List the other maintainers for the language. Do not set
@@ -1969,14 +1969,14 @@ class TrManager:
                 maintainer = m[0]
                 email = m[1]
                 s = latexLineTpl % (lang, maintainer, email, status)
-                s = s.replace(u'_', u'\\_')
+                s = s.replace('_', '\\_')
                 trlst.append(s)
 
         # Join the table lines and insert into the template.
-        latexTable = latexTableTpl % (u''.join(trlst))
+        latexTable = latexTableTpl % (''.join(trlst))
 
         # Put the HTML and LaTeX parts together and define the dic item.
-        tplDic['informationTable'] = htmlTable + u'\n' + latexTable
+        tplDic['informationTable'] = htmlTable + '\n' + latexTable
 
         # Insert the symbols into the document template and write it down.
         f = xopen(fDocName, 'w')
@@ -1986,9 +1986,10 @@ class TrManager:
 if __name__ == '__main__':
 
     # The Python 2.6+ or 3.3+ is required.
-    major, minor, patch = (int(e) for e in platfor      m.python_version_tuple())
-    if (major == 2 and minor < 6) or (major == 3 and minor < 3):
-        print('Python 2.6+ or Python 3.3+ are required for the script')
+    major, minor, patch = (int(e) for e in platform.python_version_tuple())
+    print(major, minor, patch)
+    if (major == 2 and minor < 6) or (major == 3 and minor < 0):
+        print('Python 2.6+ or Python 3.0+ are required for the script')
         sys.exit(1)
 
     # The translator manager builds the transl objects, parses the related