I'm using C and gettext to write some strings to a small display.

How do I force or at least tell translators to limit their translated string to some maximum length which could still be displayed? This max length depends on a string so it should be specified for each string separately.

Font is a monospace type, so all characters are the same length. Only the number of characters matters.

One bad and primitive solution would be to make all original (untranslated) strings of max length by adding underscores to each original string. Is there any better way, or anything that is already widely accepted and commonly used?

i'm not an expert on this but this makes sense.
i guess after gettext() does its thing, truncate the result to however many bytes:
[pseudo-code]
[/pseudo-code]
not sure why you would want to underscore-pad the input (is the probelm the input is too long or too small?)

edit: i see - your solution would make everything constant len.
(a small waste of mem)

This is already done anyway. I want the translators to know for each string how long can it be to be still shown as a whole. Then they could make a shorter translation that would not get truncated, e.g. they could use another shorter words with same meaning, or use abbreviations when needed - and only when needed.

It would be best if a translation program like KBabel would recognise this information and prevent the translator from entering too many characters for a particular translation. Or at least show him this info somehow.

i guess you would need a dictionary (table) of certain words and an appropriate abbrev.

your loop would contain a control of:
- if len is > max
- look for strings that exist in that table
- if it exists substitute it for the abbrev.

thats the only thing i can think of
______________

is it possible you can parse the output into max-len peices and then show each peice in order?