Apache OpenOffice (AOO) Bugzilla – Issue 52062
Addition of ANY_NON_NUMBER (or equivalent)
Last modified: 2017-05-20 11:27:50 UTC
When using i18n::XChracterClassification::parseAnyToken or parsePredefinedToken methdos, one can use ANY_NUMBER token type to specify any number, whether it is ascii or unicode. This is very convenient. However, to do the opposite is not equally convenient, i.e., to pick up all characters until a character of ANY_NUMBER is found. An addition of a new parse token that qualifies anything that is not an ANY_NUMBER would be very, very useful. ANY_LETTER, for example, does not pick up characters such as '/', '\', ')' ,'(', and so on. But sometimes you want to pick up everything but numbers including those that ANY_LETTER doesn't pick up. So, I propose an addition of a new parse token, which can be named ANY_NON_NUMBER (or whatever sounds better), that qualifies any character that is not ANY_NUMBER. In other words, ANY_NON_NUMBER || ANY_NUMBER will qualify any character in existence. This will certainly ease my work on the natural sort implementation (Issue 26565). Thanks, Kohei
Hi Kohei, Good reasoning, though I think I would implement it differently by introducing a "negate" flag, e.g. passing NEGATE | ANY_NUMBER would produce the result you desire. This could also be used in other combinations as in NEGATE | ANY_LETTER. Will have to think about that again though for the return value to use and other details. Thanks Eike
Targeting to 2.x
set target from 2.x to 3.x according to http://wiki.services.openoffice.org/wiki/Target_3x
Reset assigne to the default "issues@openoffice.apache.org".