storage/tmx.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3 #
   4 # Copyright 2005-2007 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 TMX translation memeory files"""
  24
  25 from translate.storage import lisa
  26 from lxml import etree
  27
  28 from translate import __version__
  29
  30 class tmxunit(lisa.LISAunit):
  31     """A single unit in the TMX file."""
  32     rootNode = "tu"
  33     languageNode = "tuv"
  34     textNode = "seg"
  35
  36     def createlanguageNode(self, lang, text, purpose):
  37         """returns a langset xml Element setup with given parameters"""
  38         if isinstance(text, str):
  39             text = text.decode("utf-8")
  40         langset = etree.Element(self.languageNode)
  41         lisa.setXMLlang(langset, lang)
  42         seg = etree.SubElement(langset, self.textNode)
  43         seg.text = text
  44         return langset
  45
  46     def getid(self):
  47         """Returns the identifier for this unit. The optional tuid property is
  48         used if available, otherwise we inherit .getid(). Note that the tuid
  49         property is only mandated to be unique from TMX 2.0."""
  50         id = self.xmlelement.get("tuid", "")
  51         return id or super(tmxunit, self).getid()
  52
  53     def istranslatable(self):
  54         return bool(self.source)
  55
  56     def addnote(self, text, origin=None):
  57         """Add a note specifically in a "note" tag.
  58
  59         The origin parameter is ignored"""
  60         if isinstance(text, str):
  61             text = text.decode("utf-8")
  62         note = etree.SubElement(self.xmlelement, self.namespaced("note"))
  63         note.text = text.strip()
  64
  65     def getnotelist(self, origin=None):
  66         """Private method that returns the text from notes.
  67
  68         The origin parameter is ignored.."""
  69         note_nodes = self.xmlelement.findall(".//%s" % self.namespaced("note"))
  70         note_list = [lisa.getText(note) for note in note_nodes]
  71
  72         return note_list
  73
  74     def getnotes(self, origin=None):
  75         return '\n'.join(self.getnotelist(origin=origin))
  76
  77     def removenotes(self):
  78         """Remove all the translator notes."""
  79         notes = self.xmlelement.findall(".//%s" % self.namespaced("note"))
  80         for note in notes:
  81             self.xmlelement.remove(note)
  82
  83     def adderror(self, errorname, errortext):
  84         """Adds an error message to this unit."""
  85         #TODO: consider factoring out: some duplication between XLIFF and TMX
  86         text = errorname + ': ' + errortext
  87         self.addnote(text, origin="pofilter")
  88
  89     def geterrors(self):
  90         """Get all error messages."""
  91         #TODO: consider factoring out: some duplication between XLIFF and TMX
  92         notelist = self.getnotelist(origin="pofilter")
  93         errordict = {}
  94         for note in notelist:
  95             errorname, errortext = note.split(': ')
  96             errordict[errorname] = errortext
  97         return errordict
  98
  99     def copy(self):
 100         """Make a copy of the translation unit.
 101
 102         We don't want to make a deep copy - this could duplicate the whole XML
 103         tree. For now we just serialise and reparse the unit's XML."""
 104         #TODO: check performance
 105         new_unit = self.__class__(None, empty=True)
 106         new_unit.xmlelement = etree.fromstring(etree.tostring(self.xmlelement))
 107         return new_unit
 108
 109
 110 class tmxfile(lisa.LISAfile):
 111     """Class representing a TMX file store."""
 112     UnitClass = tmxunit
 113     rootNode = "tmx"
 114     bodyNode = "body"
 115     XMLskeleton = '''<?xml version="1.0" encoding="utf-8"?>
 116 <!DOCTYPE tmx SYSTEM "tmx14.dtd">
 117 <tmx version="1.4">
 118 <header></header>
 119 <body></body>
 120 </tmx>'''
 121
 122     def addheader(self):
 123         headernode = self.document.find("//%s" % self.namespaced("header"))
 124         headernode.set("creationtool", "Translate Toolkit - po2tmx")
 125         headernode.set("creationtoolversion", __version__.ver)
 126         headernode.set("segtype", "sentence")
 127         headernode.set("o-tmf", "UTF-8")
 128         headernode.set("adminlang", "en")
 129         #TODO: consider adminlang. Used for notes, etc. Possibly same as targetlanguage
 130         headernode.set("srclang", self.sourcelanguage)
 131         headernode.set("datatype", "PlainText")
 132         #headernode.set("creationdate", "YYYYMMDDTHHMMSSZ"
 133         #headernode.set("creationid", "CodeSyntax"
 134
 135     def addtranslation(self, source, srclang, translation, translang):
 136         """addtranslation method for testing old unit tests"""
 137         unit = self.addsourceunit(source)
 138         unit.target = translation
 139         tuvs = unit.xmlelement.findall('.//%s' % self.namespaced('tuv'))
 140         lisa.setXMLlang(tuvs[0], srclang)
 141         lisa.setXMLlang(tuvs[1], translang)
 142
 143     def translate(self, sourcetext, sourcelang=None, targetlang=None):
 144         """method to test old unit tests"""
 145         return getattr(self.findunit(sourcetext), "target", None)
 146