Decapitalisation

for each line:

• If it contains '<num>' do nothing
• If it contains '<DOCID' or '<DOCNO' or any case variant do nothing.
• Else:
  • map all upper case letters to lower case according to the table below

Finnish and French

We implemented a general ISO 8859-1 case mapping:

Russian

Please note when reading the following table that characters are not in Russian alphabetic order in KOI8-R. See Wikipedia article on KOI8-R for details. Note also that we did not map Latin letters that might appear in proper nouns.