<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">Andre Heinecke skreiv 25.07.2024 22:11:<span
style="white-space: pre-wrap">
</span></div>
<blockquote type="cite" cite="mid:6047388.GXAFRqVoOG@esus">
<pre class="moz-quote-pre" wrap="">I wrote a small script using polib to do that and tried to find out where the
best place would be to add this to the tooling. But then I noticed that in
pology/sieve there is already a script called:
unfuzzy_context_only.py
which does this. But I have not yet figured out where this should be added or
how this could be activated for the German language.
Since this affects not only German or KMail and probably will continue in the
future I think this script should be run when the messages are collected/
merged automatically.</pre>
</blockquote>
<p>I would strongly caution against this. If a context is added or
changed, the strings *should* be looked at by the translator, as
the new context might change what’s the correct translation.</p>
<p>For example, a msgid string might originally be "View". But then
context markers are added, splitting it into two strings, one
"View" string for a button you click to view an email (or a
certificate or some such), and one "View" string for changing the
‘view’ (e.g., changing how emails are displayed), or the name of
the ‘View’ menu (having display-related settings/actions). In many
languages, these strings must have different translations.
Similarly, "Forward" might mean forwarding an email, or going to
the next item (in KOrganizer, the "Forward" string means ‘to move
to the next week/month’ in the calendar view).</p>
<p>Also note that if you use a translation compendium, translations
that differ only in context might have come from *other*
applications (e.g., a "Forward" string in KMail might have come
from KOrganizer).<br>
</p>
<p>It’s usually very quick and easy to manually go through all fuzzy
strings and manually unfuzzy them (in Lokalize: ‘Ctrl + Page Down’
followed by ‘Ctrl + U’), checking that the translations are
correct in the process.</p>
<p>But *if* you choose to use the ‘unfuzzy-context-only’ (or
‘unfuzzy-ctxmark-only’), I would recommend going through all
strings manually afterwards (by searching for the
‘unreviewed-context’ translator comment that the sieves add by
default).<br>
</p>
<br>
<pre class="moz-signature" cols="72">--
Karl Ove Hufthammer</pre>
</body>
</html>