A few questions about UI strings

[Andreas Pakulat]

On 11.04.09 18:57:05, Frederik Schwarzer wrote:
> Hi,
> 
> while translating, I found somme issues, I was unsure about.
> 
> ====
> kdevelop/languages/cpp/quickopen.cpp
> "This file imports the current open document<br/>"
> 
> What does this mean? A file can import something?
> Is there ony one open Document? Does it import all open documents or only the active one?

I'm not 100% sure about this, but I think this talks about #include for
C++ files. The resulting output is used in a navigation view,
telling the user something about the file he currently looks at. In this
particular case, the code tries to support singular and plural forms, so
in case the file that the user currently works in is used only by one
other file in the project, the text you mentioned will be used. Else the
other part will be used. So instead of "imports" one could also say
"include's" in the C++ sense of #include <>.
 
> ====
> kdevelop/languages/cpp/cppparsejob.cpp
> "Parsing actual file"
> 
> Any chance you meant "current file" here?

Kind of, I think it tries to explain that the parser finished parsing
all the files that are #include'd in the current file and now goes on
with parsing the actual files.

> ====
> kdevelop/languages/cpp/parser/rpp/pp-engine.cpp
> "expected ``:'' = %1"
> 
> Messages like these should have a context to make it possible to translate them.

In this case I think only the "expected" should be translatable anyway,
the rest is actual parser output and should be printed verbatim.

> In general you should use contexts (i18nc()) way more often.

Yes, problem is that the developer writing the text actually knows quite
well how he would translate it into his own language (if different from
english). So he might not be aware that a context is needed, not to
mention that we easily forget that the other 99% of world population
have no clue about programming :)

> ====
> kdevelop/languages/cpp/veritas/uutcontextaction.cpp
> "Generate Unit Under Test"
> 
> Another candidate for some context

Hmm, why? Though, again I'd say only the first word needs translation.
The rest is a standard term within the area of unit-testing :) Should we
change such things to

i18n("Generate")+" Unit Under Test", or rather use

i18n("Generate %1", "Unit Under Test")? Maybe still including context to
explain that some code will be generated from doing that action.

> ====
> "If this option is set, the Document View will organize documents\n
>         by their mimetype. If this option is not set, the Document View will 
> not\n
>         organize the open documents by type."
> 
> The second sentence ist not needed. Please do not over-document this kind of
> highly understandable features.

Feel free to remove.

> ====
> Could "Trolltech" be replaced by "Nokia" or is it ok to be as is?

Are you talking about copyrights? If so, then the answer is: I don't
know, please ask a KDE lawyer - or somebody more familiar with
copyrights. We did copy some code from Qt, but AFAIK that was all way
before Nokia took over and changed the company name. So I'm not sure
what the right thing is.

> ====
> kdevelop/formatters/indent_plugin.cpp
> "Home Page: <a href=\"http://www.gnu.org/software/indent/\">http://www.gnu.org/software/indent</a>/"
> 
> What is the point of using the trailing slash like that?

Typo I think.

> ====
> kdevelop/formatters/astyle_preferences.ui
> "Break bl&ocks all"

Actually had to lookup what it does in the code, not really god huh? And
then I needed to lookup the option in man astyle.

Ok, here's what it does, additionally to "Break blocks" (which inserts
newlines between unrelated blocks, for example between two class
declarations), "Break Blocks All" adds newlines also around closing
headers, i.e. around the "catch" for a try/catch, or the "else" for an
if/else.

So yes, "Break all blocks" is better, I think this comes from the
original astyle api/manpage which lists "--break-blocks=all".

Andreas

-- 
You are fairminded, just and loving.