storage/ts.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3 #
   4 # Copyright 2004-2007 Zuza Software Foundation
   5 #
   6 # This file is part of translate.
   7 #
   8 # translate is free software; you can redistribute it and/or modify
   9 # it under the terms of the GNU General Public License as published by
  10 # the Free Software Foundation; either version 2 of the License, or
  11 # (at your option) any later version.
  12 #
  13 # translate is distributed in the hope that it will be useful,
  14 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 # GNU General Public License for more details.
  17 #
  18 # You should have received a copy of the GNU General Public License
  19 # along with translate; if not, write to the Free Software
  20 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  21 #
  22
  23 """Module for parsing Qt .ts files for translation.
  24
  25 Currently this module supports the old format of .ts files. Some applictaions
  26 use the newer .ts format which are documented here:
  27 U{TS file format 4.3<http://doc.trolltech.com/4.3/linguist-ts-file-format.html>},
  28 U{Example<http://svn.ez.no/svn/ezcomponents/trunk/Translation/docs/linguist-format.txt>}
  29
  30 U{Specification of the valid variable entries <http://doc.trolltech.com/4.3/qstring.html#arg>},
  31 U{2 <http://doc.trolltech.com/4.3/qstring.html#arg-2>}
  32 """
  33
  34 from translate.misc import ourdom
  35
  36 class QtTsParser:
  37     contextancestors = dict.fromkeys(["TS"])
  38     messageancestors = dict.fromkeys(["TS", "context"])
  39     def __init__(self, inputfile=None):
  40         """make a new QtTsParser, reading from the given inputfile if required"""
  41         self.filename = getattr(inputfile, "filename", None)
  42         self.knowncontextnodes = {}
  43         self.indexcontextnodes = {}
  44         if inputfile is None:
  45             self.document = ourdom.parseString("<!DOCTYPE TS><TS></TS>")
  46         else:
  47             self.document = ourdom.parse(inputfile)
  48             assert self.document.documentElement.tagName == "TS"
  49
  50     def addtranslation(self, contextname, source, translation, comment=None, transtype=None, createifmissing=False):
  51         """adds the given translation (will create the nodes required if asked). Returns success"""
  52         contextnode = self.getcontextnode(contextname)
  53         if contextnode is None:
  54             if not createifmissing:
  55                 return False
  56             # construct a context node with the given name
  57             contextnode = self.document.createElement("context")
  58             namenode = self.document.createElement("name")
  59             nametext = self.document.createTextNode(contextname)
  60             namenode.appendChild(nametext)
  61             contextnode.appendChild(namenode)
  62             self.document.documentElement.appendChild(contextnode)
  63         if not createifmissing:
  64             return False
  65         messagenode = self.document.createElement("message")
  66         sourcenode = self.document.createElement("source")
  67         sourcetext = self.document.createTextNode(source)
  68         sourcenode.appendChild(sourcetext)
  69         messagenode.appendChild(sourcenode)
  70         if comment:
  71             commentnode = self.document.createElement("comment")
  72             commenttext = self.document.createTextNode(comment)
  73             commentnode.appendChild(commenttext)
  74             messagenode.appendChild(commentnode)
  75         translationnode = self.document.createElement("translation")
  76         translationtext = self.document.createTextNode(translation)
  77         translationnode.appendChild(translationtext)
  78         if transtype:
  79             translationnode.setAttribute("type", transtype)
  80         messagenode.appendChild(translationnode)
  81         contextnode.appendChild(messagenode)
  82         return True
  83
  84     def getxml(self):
  85         """return the ts file as xml"""
  86         xml = self.document.toprettyxml(indent="    ", encoding="utf-8")
  87         #This line causes empty lines in the translation text to be removed (when there are two newlines)
  88         xml = "\n".join([line for line in xml.split("\n") if line.strip()])
  89         return xml
  90
  91     def getcontextname(self, contextnode):
  92         """returns the name of the given context"""
  93         namenode = ourdom.getFirstElementByTagName(contextnode, "name")
  94         return ourdom.getnodetext(namenode)
  95
  96     def getcontextnode(self, contextname):
  97         """finds the contextnode with the given name"""
  98         contextnode = self.knowncontextnodes.get(contextname, None)
  99         if contextnode is not None:
 100             return contextnode
 101         contextnodes = self.document.searchElementsByTagName("context", self.contextancestors)
 102         for contextnode in contextnodes:
 103             if self.getcontextname(contextnode) == contextname:
 104                 self.knowncontextnodes[contextname] = contextnode
 105                 return contextnode
 106         return None
 107
 108     def getmessagenodes(self, context=None):
 109         """returns all the messagenodes, limiting to the given context (name or node) if given"""
 110         if context is None:
 111             return self.document.searchElementsByTagName("message", self.messageancestors)
 112         else:
 113             if isinstance(context, (str, unicode)):
 114                 # look up the context node by name
 115                 context = self.getcontextnode(context)
 116                 if context is None:
 117                     return []
 118             return context.searchElementsByTagName("message", self.messageancestors)
 119
 120     def getmessagesource(self, message):
 121         """returns the message source for a given node"""
 122         sourcenode = ourdom.getFirstElementByTagName(message, "source")
 123         return ourdom.getnodetext(sourcenode)
 124
 125     def getmessagetranslation(self, message):
 126         """returns the message translation for a given node"""
 127         translationnode = ourdom.getFirstElementByTagName(message, "translation")
 128         return ourdom.getnodetext(translationnode)
 129
 130     def getmessagetype(self, message):
 131         """returns the message translation attributes for a given node"""
 132         translationnode = ourdom.getFirstElementByTagName(message, "translation")
 133         return translationnode.getAttribute("type")
 134
 135     def getmessagecomment(self, message):
 136         """returns the message comment for a given node"""
 137         commentnode = ourdom.getFirstElementByTagName(message, "comment")
 138         # NOTE: handles only one comment per msgid (OK)
 139         # and only one-line comments (can be VERY wrong) TODO!!!
 140         return ourdom.getnodetext(commentnode)
 141
 142     def iteritems(self):
 143         """iterates through (contextname, messages)"""
 144         for contextnode in self.document.searchElementsByTagName("context", self.contextancestors):
 145             yield self.getcontextname(contextnode), self.getmessagenodes(contextnode)
 146
 147     def __del__(self):
 148         """clean up the document if required"""
 149         if hasattr(self, "document"):
 150             self.document.unlink()
 151