shouls support search without diacritics about clir

[ptakopysk]

Plan:

• each file has diacritics and non-diacritics version -- let's say filename.utf8 and filename.ascii
• these are concatenated (together with some metadata such as document name) into a txt file (cat $n.mtea $n.utf8 $n.ascii > $n.txt
• these txt files are indexed (or maybe they should be called filename.index but I am not sure if solr would like that, maybe it relies in the file extension (???)
• so the retrieved filename ends in .txt (which is not part of the file id anyway)
• but in the viewer we want to view the .utf8 file
• and the highlighting... that's hard, this probably won't work... we would need to do it inversely, as I cannot define the match, we would have to do the highlighting in the .ascii file and then port it into the .utf8 file... which I don't want to do

[ptakopysk]

Also: https://stackoverflow.com/questions/17162163/solr-accent-removal