Question about Unicode Ranges in TrueType fonts

From: Elisha Berns (e.berns@computer.org)
Date: Thu Jun 26 2003 - 00:58:28 EDT

  • Next message: souravm: "Uquivalence of some Japanes charcaters in Unicode"

    Hi,

    Some weeks back there were a number of postings about software for
    viewing Unicode Ranges in TrueType fonts and I had a few questions about
    that. Most viewers listed seemed to only check the Unicode Range bits of
    the fonts which can be misleading in certain cases.

    Anyways I wanted to ask the following: for a given, specific language
    how can you tell if a font supports that language (as opposed to the
    Unicode Range)? What I have noticed is that some fonts may support a
    Unicode Range but may have only a few characters from that range and
    thus cannot really be used for writing in that language. For example,
    the font Estrangello Edessa supports Unicode Ranges for Arabic and
    Syriac, but only has glyphs for Arabic numerals (and vowels?), but no
    glyphs for the Arabic alphabet. It apparently however fully supports
    the Syriac character set. So I would say that really it only supports
    Syriac in any practical sense.

    So If I were to algorithmically search for fonts supporting Arabic by
    checking the Unicode Subset Range bits, I would assume, mistakenly, this
    font supports Arabic writing. But if you want to search algorithmically
    and find fonts that actually support a *language* and not just a partial
    Unicode Range, how would you do it?

    I did notice a posting regarding fontconfig:

    "If you are on a system that uses fontconfig, such as most recent linux
    distros, you can find out which fonts support a particular language
    using fc-list; for example:"

      $ fc-list :lang=ko

    However, fontconfig apparently compiles in all of its .orth
    (orthography) files for each supported language so it can do sort of a
    manual computation of which font supports all the characters in a given
    .orth file. (Did I get that right?) Short of compiling all this
    information together manually and searching every font for complete
    coverage of one of these sets of characters is there another way to go
    about this?

    Many thanks for any help here in this baffling world of Unicode,

    Yours truly,

    Elisha Berns
    e.berns@computer.org
    tel. (310) 556 - 8332
    fax (310) 556 - 2839

    This archive was generated by hypermail 2.1.5 : Thu Jun 26 2003 - 01:41:53 EDT