[pology] Extrange behaviour with validation subdirectives

mvillarino mvillarino at kde-espana.es
Wed Nov 16 06:48:53 PST 2011


Good afternoon,

All of the below mentioned rules available at:
http://websvn.kde.org/trunk/l10n-support/pology/lang/gl/rules/terminology.rules?view=markup

133 {\bcheck(s|ed|ing|)?}i
134 valid after="\bspell "
135 valid before="box"
136 valid before=" spell"
137 id="noPT-2011_check"
138 valid msgstr="\bcomprob"
139 valid msgstr="\brevis"
140 valid msgstr="\bmar(quear|ado|ou)"
141 valid msgstr="\bescoll"
142 hint="«check» tradúcese como

Line 134 will match according to what is expected from the documentation:
if the "after" text is placed as a "prefix" (it will match, for instance, "*
spell* *checking*")

But:

156 # «marcar» traduce a «check»
157 [\bmar(c|que)]i
158 valid !after="e\s"
159 valid !after="ador"
160 id="PT-2011-dual_check3"
161 valid msgid="\bcheck"
162 hint="«marcar» traduce a «check»"

[\bactiv]i
320 valid !after="idade"
321 id="PT-2011-dual_enable"
322 valid msgid="\benable"
323 hint="«activar» traduce a «enable»"

# «substituir» traduce a «replace»
707 [\bsubstitu]i
708 valid after="car(a|á)cter(es)?\sde\s"
709 id="PT-2011-dual_replace"
710 valid msgid="\breplace"
711 hint="«substituir» traduce a «replace»"

Line 159 ( or 320) will math just the opposite, if the "after" text goes as
a "postfixed" element (it will (not) match, for instance, "*marc**ador*",
or "*activ**idade*", neither it will match "*caracter de **substitu*ción ))
But after the documentation I would expect 320 to pass a message wich
contains "*idade**activ*" or "*e ****marque*", that is if the trigger is
not after the subdirective.

Short version: It seems that the semantics of the validations subdirectives
"after", "before", "not after", and "not before" depend upon wich part of
the message (source or target) acts as trigger.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.nedohodnik.net/pipermail/pology-nedohodnik.net/attachments/20111116/f2e005ed/attachment.htm>


More information about the pology mailing list