unicodetools icon indicating copy to clipboard operation
unicodetools copied to clipboard

"age" filter for "list-unicodeset" utility does not work correctly for values >= 10

Open srl295 opened this issue 4 years ago • 7 comments

@srutzky wrote:

(I originally posted this in JIRA as CLDR-14700 but was told that this was a more appropriate forum for reporting this type of problem)

This concerns the Unicode Utilities: UnicodeSet utility, located at: https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp

Filtering on the "age" property does not work correctly. The issue started a few years ago, after version 10.0 was released (and this utility was updated for that version). Versions starting at 10.0 do not behave correctly, and are somehow "linked" (for lack of a better term) to version 1.1. The examples below should illustrate the issue.

 

Correct, but only when "=1.1"

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D1.1%3A%5D&g=age&i=

[:Age=1.1:]
33,977 Code Points

 

A value of "=10.0" or higher returns 1.1 and version 10.0 through the specified version

But, not versions 2.0 through 9.0 (which should be returned). For example, specifying version 12.0 will return versions:

  • 1.1
  • 10.0
  • 11.0
  • 12.0

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D12.0%3A%5D&g=age&i=

[:Age=12.0:]
43,733 Code Points

 

A value of "=2.0" through "=9.0" returns 1.1 through the specified version and versions 10.0 and higher

(versions 10.0 and higher should not be returned). For example, specifying version 3.1 will return versions:

  • 1.1
  • 2.0
  • 2.1
  • 3.0
  • 3.1
  • 10.0
  • 11.0
  • 12.0

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D3.1%3A%5D&g=age&i=

[:Age=3.1:]
243,477 Code Points

 

A value of "=10.0" or higher and "!=1.1" returns version 10.0 through the specified version

But, not versions 2.0 through 9.0 (which should be returned). For example, specifying version 11.0 and !1.1 will return versions:

  • 10.0
  • 11.0

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D11.0%3A%5D+%26+%5B%3AAge%21%3D1.1%3A%5D&g=age&i=

[:Age=11.0:] & [:Age!=1.1:]
9,202 Code Points

 

A value of "=10.0" or higher and "!=2.0" through "!=9.0" returns nothing / empty-set

(one version higher than the specified "!=" version through specified "=" version should be returned). For example, specifying version 11.0 and !2.0 will return versions:

  • (none / nada / zip / zilch )
  • (however, versions 2.1 through 11.0 should be returned)
  • (specifying version 9.0 and !2.0 does return the correct results)

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D11.0%3A%5D+%26+%5B%3AAge%21%3D2.0%3A%5D&g=age&i=

[:Age=11.0:] & [:Age!=2.0:]
0 Code Points

https://util.unicode.org/UnicodeJsps/list-unicodeset.jsp?a=%5B%3AAge%3D9.0%3A%5D+%26+%5B%3AAge%21%3D2.0%3A%5D&g=age&i=

[:Age=9.0:] & [:Age!=2.0:]
89,287 Code Points

srl295 avatar May 03 '21 18:05 srl295

(Related, do we have a clear link for how to give feedback on the utilities?)

srl295 avatar May 03 '21 18:05 srl295

Is it comparing String values? It should compare VersionInfo values. (I believe that the ICU UnicodeSet implementation handles Age properly.)

markusicu avatar May 03 '21 20:05 markusicu

Hello @srl295 and @markusicu . I appreciate Steve moving what was CLDR-14700 over to here as I didn't know when I posted CLDR-14700 that this was a place to report such things. However, for consistency and formatting, etc, I went ahead and re-posted the issue here as #55 (sorry if that made things confusing and/or sloppy). Since there most definitely is no clear link or instructions for reporting issues with the Unicode utilities, perhaps this issue could be adjusted to be about that?

srutzky avatar May 04 '21 04:05 srutzky

Hello @srl295 and @markusicu . I appreciate Steve moving what was CLDR-14700 over to here as I didn't know when I posted CLDR-14700 that this was a place to report such things. However, for consistency and formatting, etc, I went ahead and re-posted the issue here as #55 (sorry if that made things confusing and/or sloppy). Since there most definitely is no clear link or instructions for reporting issues with the Unicode utilities, perhaps this issue could be adjusted to be about that?

I copied your description of #55 to here. This one is already linked and has comments.

srl295 avatar May 04 '21 04:05 srl295

Hi @srl295 . Ok, if you think that is best. Is there a way for me to edit the initial comment here if I want to make an adjustment? Or, can I be made the owner of this issue (which would allow me to make changes)? This was a large part of why I reposted.

srutzky avatar May 04 '21 13:05 srutzky

(Related, do we have a clear link for how to give feedback on the utilities?)

@srl295 I just posted #56 to handle that issue.

srutzky avatar May 04 '21 14:05 srutzky

Hi @srl295 . Ok, if you think that is best. Is there a way for me to edit the initial comment here if I want to make an adjustment? Or, can I be made the owner of this issue (which would allow me to make changes)? This was a large part of why I reposted.

just add another comment…

srl295 avatar May 04 '21 16:05 srl295