sheets/checktrans.py

   1 #!/usr/bin/python
   2 #
   3 # This quick hack gives translation statistics about the various sheets.
   4 #
   5 # Copyright (C) 2001, Cyrille Chepelov <chepelov@calixo.net>
   6 #
   7 # This quick hack is free software; you can redistribute it and/or modify
   8 # it under the terms of the GNU General Public License as published by
   9 # the Free Software Foundation; either version 2 of the License, or
  10 # (at your option) any later version.
  11 #
  12 # This quick hack is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 # GNU General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU General Public License
  18 # along with this quick hack; if not, write to the Free Software
  19 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  20 #
  21
  22
  23
  24 from xml.sax import saxexts
  25 from xml.sax import saxlib
  26 import os,sys,string
  27
  28 class CounterHandler(saxlib.DocumentHandler):
  29
  30     def __init__(self):
  31         self.elemstk = []
  32         self.langstk = []
  33         self.namestk = []
  34
  35     def setDocumentLocator(self,locator):
  36         self.locator = locator
  37         saxlib.DocumentHandler.setDocumentLocator(self,locator)
  38         self.langs = {}
  39         self.namelangs = {}
  40
  41     def warning(self,message):
  42         print "W:%s:L%d:C%d: %s" % (self.locator.getSystemId(),
  43                                     self.locator.getLineNumber(),
  44                                     self.locator.getColumnNumber(),
  45                                     message)
  46     def _countlang(self,name):
  47         locdct = self.langstk[-1]
  48         if locdct.has_key(name):
  49             self.warning("duplicate description for %s, language code %s" % \
  50                          (self.namestk[-1],name))
  51             locdct[name] = locdct[name] + 1
  52         else:
  53             locdct[name] = 1
  54
  55         if self.langs.has_key(name):
  56             self.langs[name] = self.langs[name] + 1
  57         else:
  58             self.langs[name] = 1
  59
  60     def _countnamelang(self,name):
  61         if self.namelangs.has_key(name):
  62             self.warning("duplicate name for sheet, language code %s" % name)
  63             self.namelangs[name] = self.namelangs[name] + 1
  64         else:
  65             self.namelangs[name] = 1
  66
  67         if self.langs.has_key(name):
  68             self.langs[name] = self.langs[name] + 1
  69         else:
  70             self.langs[name] = 1
  71
  72     def startElement(self,name,attrs):
  73         #print "start of ",name,attrs,attrs.map
  74         attmap = attrs.map
  75         self.elemstk.append(name)
  76         if (name == "sheet") or (name == "object"):
  77             self.langstk.append({})
  78             if attmap.has_key('name'):
  79                 name = 'Object "%s"' % attmap['name']
  80             else:
  81                 name = 'Sheet "%s"' % self.locator.getSystemId()
  82             self.namestk.append(name)
  83         elif (name == "name"):
  84             if attmap.has_key("xml:lang"):
  85                 lang = attmap["xml:lang"]
  86             else:
  87                 lang = ""
  88             self._countnamelang(lang)
  89         elif (name == "description"):
  90             if attmap.has_key("xml:lang"):
  91                 lang = attmap["xml:lang"]
  92             else:
  93                 lang = ""
  94             self._countlang(lang)
  95
  96     def endElement(self,name):
  97         popped = self.elemstk.pop()
  98         if popped != name:
  99             raise Exception("stack error somewhere...")
 100         if (name == "sheet") or (name == "object"):
 101             res = self.langstk.pop()
 102             self.namestk.pop()
 103             #print "end of",name, res
 104         else:
 105             #print "end of ",name
 106             pass
 107
 108 ##  class BasicEntityResolver(saxlib.EntityResolver):
 109 ##      def resolveEntity(name,publicId,systemId):
 110 ##          print "D:resolveEntity(%s,%s,%s)" % (name,publicId,systemId),
 111 ##          if publicId == "-//FOO//BAR//EN":
 112 ##              res = "file:../doc/sheet.dtd"
 113 ##          else:
 114 ##              res = saxlib.EntityResolver.resolveEntity(name,publicId,systemId)
 115
 116 ##          print "--> res",res,type(res)
 117
 118 ##          return res
 119
 120 if len(sys.argv)<2:
 121     print "Usage: %s <sheet.sheet>" % sys.argv[0]
 122     print
 123     print " <sheet.sheet>: file name of the sheet to check"
 124     sys.exit(1)
 125
 126 # Load parser and driver
 127
 128 p=saxexts.make_parser()
 129 #p=saxexts.XMLValParserFactory.make_parser()
 130 ch=CounterHandler()
 131 p.setDocumentHandler(ch)
 132 #p.setEntityResolver(BasicEntityResolver())
 133
 134 fnames =sys.argv[1:]
 135
 136 def make_langresult(langdict,colsize):
 137     langres = map(lambda s,colsize=colsize: string.ljust(s,colsize),
 138                   map(lambda (cc,count),total=langdict['']:
 139                       ("%s:%d%%" % (cc,100*count/total)),
 140                       filter(lambda (cc,count): cc, langdict.items())))
 141     langres.sort()
 142     return langres
 143
 144 def maxlen(a,b):
 145     if len(a)>len(b): return a
 146     return b
 147 namelen = len(reduce(maxlen,fnames,""))
 148
 149 globlangs = {}
 150
 151 for name in fnames:
 152     OK=0
 153     try:
 154         p.parse("file://localhost" + os.getcwd() + "/" + name)
 155         OK=1
 156
 157         maxlanglen = len(reduce(maxlen,ch.langs.keys(),""))
 158         langres = make_langresult(ch.langs,maxlanglen+5)
 159         columns = (79 - (4 + namelen)) / (maxlanglen+5)
 160         #print "maxlanglen = ",maxlanglen,"columns=",columns
 161
 162         langres1,langrest = langres[:columns],langres[columns:]
 163         sys.stdout.write(("I: %%%ds %%s\n" % namelen) %
 164                          (name,string.join(langres1)))
 165         while langrest:
 166             langresn,langrest = langrest[:columns],langrest[columns:]
 167             sys.stdout.write("I: %s %s\n" % (' ' * namelen,
 168                                              string.join(langresn)))
 169
 170         for (k,v) in ch.langs.items():
 171             if globlangs.has_key(k):
 172                 globlangs[k] = globlangs[k] + v
 173             else:
 174                 globlangs[k] = v
 175
 176     except IOError,e:
 177         sys.stderr.write("E: %s: %s\n" % (name,str(e)))
 178     except saxlib.SAXException,e:
 179         sys.stderr.write("E: %s\n" % str(e))
 180
 181 maxlanglen = len(reduce(maxlen,globlangs.keys(),""))
 182 langres = make_langresult(globlangs,maxlanglen+5)
 183 columns = (79 - 12) / (maxlanglen+5)
 184
 185 langres1,langrest = langres[:columns],langres[columns:]
 186 sys.stdout.write("\nI: OVERALL: %s\n" % string.join(langres1))
 187 while langrest:
 188     langresn,langrest = langrest[:columns],langrest[columns:]
 189     sys.stdout.write("I: %s %s\n" % (' ' * 8,
 190                                      string.join(langresn)))
 191
 192