new updated Irish dictionary hash files and other dictionary stuff [fjf]

From: Alan Horkan (horkana@tcd.ie)
Date: Sat Aug 03 2002 - 16:23:28 EDT

  • Next message: Patrick Lam: "Re: commit: fl_BlockLayout.h/cpp"

    the maintainer of the irish dictionaries contacted me to say he had
    updated the dictionary and it was substantially more accurate than the one
    older one we got that had been packaged for Mandrake.
    he very generously did this all in English so i did not have to try very
    hard.
    http://borel.slu.edu/ispell/index-en.html

    gabriel said on that fjf might want to build/package the hashes in a
    particular way but i had already built them myself

    they are here, at least temporarily (not all files are have read
    permission)
    http://www.maths.tcd.ie/~horkana/testing/dict/ispell-gaeilge-2.0/

    i recommend we used gaeilgemor.hash (literally irish big .hash)
    and rename it to irish.hash for back compatibility.

    I have no idea how to make PPC compatible dictionaries, please advise.
    Aspell dictionaries are also available.

    while i am at it on other dictionary related matters
    http://cancerweb.ncl.ac.uk/omd/spell.html
    these guys have a English medical words list that could be converted into
    a .hash file although there may be licensing issues (but they are already
    effectively giving away the words as text list). I cannot remember why i
    never properly finished looking into this.

    the words could probably be converted to American Medical words using a
    well thought out find and replace script, from my limited understanding
    American Medical terms tend to be less like latin
    English American
    Paediatrician Pediatrician
    and a few other regional variations.

    Dom might know someone to ask ;) or failing that I could ask my sister &
    brother in law.
    Im slightly amazed i never filed a bug report in bugzilla about
    requests for a medical dictionary.

    I had some other thoughts about why we provide .rpm, .zip, .tar.gz because
    just about any system that has the tools to extract zip has the tools to
    extract tar.gz and abiword has built in unzip code so could we save the
    user the effort of extracting these files and simply extract any gzipped
    files placed in the dictionary folder (but then i remember the developers
    have better things to do).

    In some cases we also provide NSIS based self extracting exe files, and i
    wonder why they cannot optionally be extracted using a standard zip
    program like certain other zip compressed SFX files, i guess i will have
    to ask the nice people at Nullsoft about that one.

    I would really like to have this done in time for 1.0.3

    Im tired, rambling, need sleep ...

    Sincerely
    Alan Horkan



    This archive was generated by hypermail 2.1.4 : Sat Aug 03 2002 - 16:30:39 EDT