RE: CLDR Survey bug? (currency formating)

From: Keutgen, Walter (walter.keutgen@be.unisys.com)
Date: Thu Apr 13 2006 - 03:59:58 CST

  • Next message: Mark Davis: "Re: Unicode 5.0 Character Count?"

    Philippe,

    the XXX (=no currecny) instead of EUR is also incorrect.

    It seems that the tool often fetches false values in this example display routine. If one looks at currencies sorted by priority, one finds a real mixing of ISO currency codes. See: http://unicode.org/cldr/apps/survey?_=fr&x=currencies&p_sort=interest

    CHE/displayName euro WIR Example: -10472.603 → ‎-10 472,60 DEM‎ → -10472.6

    CHE becomes DEM (Deutsche Mark) and the French number format instead of the Swiss one.

    (-: The pivotal pair of the European Union, France and Germany, rooted in Swiss finance? :-)

    The case you reported is even worse, as it yields not another but an inexisting format.

    So bug reports 1001 and 1026 are complementary.

    Best regards

    Walter

    THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY MATERIAL and is thus for use only by the intended recipient. If you received this in error, please contact the sender and delete the e-mail and its attachments from all computers.

    -----Original Message-----
    From: unicode-bounce@unicode.org [mailto:unicode-bounce@unicode.org] On Behalf Of Philippe Verdy
    Sent: Donnerstag, den 13. April 2006 06:56
    To: unicode@unicode.org
    Subject: CLDR Survey bug? (currency formating)

    http://unicode.org/cldr/apps/survey?_=fr&x=numbers

    Why does the survey tool generate this invalid output from the correct format and input string?

    //ldml/numbers/currencyFormats/currencyFormatLength/currencyFormat[@type="standard"]/pattern[@type="standard"]

    Locale Data: "#,##0.00 ¤" (correct)

    Example: "-50428.047" → ‎"-50,428,05 XXX"‎ → Unparseable number: "-50,428,05 XXX"

    My opinion is that it should be:

    Example: "-50428.047" → ‎"-50 428,05 XXX"‎ → Unparseable number: "-50428.047"

    The output uses "," from the format string, instead of honoring the group separator (a space for French), and so this produces an unparsable number (because the incorrect comma is the French decimal separator)

    It looks like a bug in the formatting library used, that incorrectly uses the LDML data (which is correct here) when formating the input number with that localized format string.

    This archive was generated by hypermail 2.1.5 : Thu Apr 13 2006 - 04:13:54 CST