File:  [Repository] / kupu / python / nationalizer.py
Revision 1.1.1.1 (vendor branch): download - view: text, annotated - select for diffs - revision graph
Thu Sep 15 13:06:00 2005 UTC (18 years, 9 months ago) by dwinter
Branches: first, MAIN
CVS tags: dwinter, HEAD
modifizierter kupu fuer webpages des instituts

    1: #!/usr/bin/python2.3
    2: 
    3: """Return the Kupu .html file with i18n applied"""
    4: 
    5: from xml.dom.minidom import parseString
    6: import os
    7: 
    8: ID = 0
    9: STR = 1
   10: 
   11: I18NNS = 'http://xml.zope.org/namespaces/i18n'
   12: 
   13: def ustr(i):
   14:     if type(i) == unicode:
   15:         return i
   16:     else:
   17:         return unicode(str(i), 'UTF-8')
   18: 
   19: def get_locale():
   20:     if os.environ.has_key('HTTP_ACCEPT_LANGUAGE'):
   21:         charsets = [l.strip() for l in 
   22:                 os.environ['HTTP_ACCEPT_LANGUAGE'].split(';')[0].split(',')]
   23:         return charsets
   24: 
   25: class Nationalizer:
   26:     """Translates string in an HTML or XML file using i18n: directives"""
   27: 
   28:     not_single = ['a', 'abbr', 'acronym', 'address', 'applet', 
   29:                     'b', 'bdo', 'big', 'blink', 'blockquote', 
   30:                     'button', 'caption', 'center', 'cite', 
   31:                     'comment', 'del', 'dfn', 'dir', 'div',
   32:                     'dl', 'dt', 'em', 'embed', 'fieldset',
   33:                     'font', 'form', 'frameset', 'h1', 'h2',
   34:                     'h3', 'h4', 'h5', 'h6', 'i', 'iframe',
   35:                     'ins', 'kbd', 'label', 'legend', 'li',
   36:                     'listing', 'map', 'marquee', 'menu',
   37:                     'multicol', 'nobr', 'noembed', 'noframes',
   38:                     'noscript', 'object', 'ol', 'optgroup',
   39:                     'option', 'p', 'pre', 'q', 's', 'script',
   40:                     'select', 'small', 'span', 'strike', 
   41:                     'strong', 'style', 'sub', 'sup', 'table',
   42:                     'tbody', 'td', 'textarea', 'tfoot',
   43:                     'th', 'thead', 'title', 'tr', 'tt', 'u',
   44:                     'ul', 'xmp']
   45: 
   46:     def __init__(self, htmlfile, locale):
   47:         self.htmlfile = htmlfile
   48:         self.locale = locale
   49: 
   50:     def translate(self):
   51:         """load and translate everything"""
   52:         popath = self.get_po_file_path(self.locale)
   53:         if popath is not None:
   54:             pofp = open(popath)
   55:             try:
   56:                 msgcat = self.parse_po_file(pofp)
   57:             finally:
   58:                 pofp.close()
   59:         else:
   60:             # if no pofile, parse anyway to get rid of those nasty i18n:
   61:             # attributes (obviously not very fast, perhaps we need to either
   62:             # cache a parsed version and send that back or just remove the
   63:             # attributes here)
   64:             msgcat = {}
   65:         xmlfp = open(self.htmlfile)
   66:         try:
   67:             xml = xmlfp.read()
   68:         finally:
   69:             xmlfp.close()
   70:         dom = parseString(xml)
   71:         self.apply_i18n(dom, msgcat)
   72:         return self.serialize(dom.documentElement)
   73: 
   74:     def parse_po_file(self, pofp):
   75:         """parse the .po file, create a mapping msgid->msgstr"""
   76:         cat = {}
   77:         state = None
   78:         msgid = None
   79:         msgstr = None
   80:         for line in pofp.readlines():
   81:             line = line.strip()
   82:             if line.startswith('#') or not line:
   83:                 continue
   84:             if line.startswith('msgid'):
   85:                 if msgid and msgstr:
   86:                     cat[msgid] = msgstr
   87:                 msgid = line[7:-1]
   88:                 state = ID
   89:             elif line.startswith('msgstr'):
   90:                 msgstr = line[8:-1]
   91:             else:
   92:                 # ignore for now, might be a multiline msgstr, if we
   93:                 # want to support those we should add some code here...
   94:                 pass
   95:         if msgid and msgstr:
   96:             cat[msgid] = msgstr
   97:         return cat
   98: 
   99:     def apply_i18n(self, dom, msgcat):
  100:         """apply nationalization of the full dom"""
  101:         nodes = dom.documentElement.getElementsByTagName('*')
  102:         for node in nodes:
  103:             if node.hasAttributeNS(I18NNS, 'translate'):
  104:                 self.apply_translate(node, msgcat)
  105:             if node.hasAttributeNS(I18NNS, 'attributes'):
  106:                 self.apply_attributes(node, msgcat)
  107: 
  108:     def apply_translate(self, node, msgcat):
  109:         """handle Zope-style i18n:translate"""
  110:         buf = []
  111:         msgid = msgstr = node.getAttributeNS(I18NNS, 'translate').strip()
  112:         if not msgid:
  113:             # no msgid in the attribute, use the node value
  114:             for child in node.childNodes:
  115:                 if child.nodeType == 3:
  116:                     buf.append(child.nodeValue)
  117:                 else:
  118:                     raise TypeError, \
  119:                         ('illegal element %s in i18n:translate element' % 
  120:                             child.nodeName)
  121:             msgid = msgstr = self.reduce_whitespace(u''.join(buf).strip())
  122:         if msgcat.has_key(msgid):
  123:             msgstr = msgcat[msgid]
  124:         # now replace the contents of the node with the new contents
  125:         while node.hasChildNodes():
  126:             node.removeChild(node.firstChild)
  127:         node.removeAttributeNS(I18NNS, 'translate')
  128:         node.appendChild(node.ownerDocument.createTextNode(msgstr))
  129: 
  130:     def apply_attributes(self, node, msgcat):
  131:         """handle Zope-style i18n:attributes"""
  132:         attrnames = node.getAttributeNS(I18NNS, 'attributes').split(' ')
  133:         for attr in attrnames:
  134:             value = node.getAttribute(attr)
  135:             if value and msgcat.has_key(value):
  136:                 node.setAttribute(attr, unicode(msgcat[value], 'UTF-8'))
  137:         node.removeAttributeNS(I18NNS, 'attributes')
  138: 
  139:     def reduce_whitespace(self, string):
  140:         for char in ['\n', '\t', '\r']:
  141:             string  = string.replace(char, ' ')
  142:         while string.find('  ') > -1:
  143:             string = string.replace('  ', ' ')
  144:         return string
  145: 
  146:     def get_po_file_path(self, locale):
  147:         for language in locale:
  148:             startdir = '../i18n'
  149:             language = language.split('-')
  150:             pathstart = '%s/kupu-%s' % (startdir, language[0])
  151:             paths = []
  152:             if len(language) == 2:
  153:                 paths.append('%s-%s.po' % (pathstart, language[1]))
  154:             paths += [
  155:                 '%s-default.po' % pathstart,
  156:                 '%s.po' % pathstart,
  157:                 ]
  158:             for path in paths:
  159:                 if os.path.isfile(path):
  160:                     return path
  161: 
  162:     def serialize(self, el):
  163:         buf = []
  164:         if el.nodeType == 1:
  165:             buf.append('<%s' % el.nodeName)
  166:             if len(el.attributes):
  167:                 for attr, value in el.attributes.items():
  168:                     if value is not None:
  169:                         buf.append(' %s="%s"' % (attr, self.entitize(value)))
  170:             if el.hasChildNodes() or el.nodeName in self.not_single:
  171:                 buf.append('>')
  172:                 for child in el.childNodes:
  173:                     buf += self.serialize(child)
  174:                 buf.append('</%s>' % el.nodeName)
  175:             else:
  176:                 buf.append(' />')
  177:         elif el.nodeType == 3:
  178:             buf.append(el.nodeValue)
  179:         else:
  180:             print 'ignoring node of type', node.nodeType
  181:         return ''.join([ustr(b) for b in buf])
  182: 
  183:     def entitize(self, string):
  184:         string = string.replace('&', '&amp;')
  185:         string = string.replace('<', '&lt;')
  186:         string = string.replace('>', '&gt;')
  187:         string = string.replace('"', '&quot;')
  188:         return string
  189:         
  190: if __name__ == '__main__':
  191:     # test code
  192:     os.chdir(os.path.abspath(os.path.dirname(__file__)))
  193:     i = Nationalizer('../common/kupu.html', ['nl'])
  194:     print i.translate().encode('UTF-8')

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>