Posted by KayL on October 5, 2009 at 5:21pm
Hi,
l.d.o managed 5xxx projects.
I found that many sentence/Source text have very similar structure.
Estimated around 5000 and maybe a half them only 1/2 words difference.
example:
"must be a positive integer" - only 1/2 words difference.
http://groups.drupal.org/files/mustbeatips.png
more example:
has been added
has been created
has been updated
has been edited
has been saved
has been removed
has been approved
has been duplicated
"have been"...
must be a number
must be a string
must be a ...
must be a between
Please enter a
must enter a
enter a valid
hope it can help your team. Please share if your found some patterns.
regards,
KayL
| Attachment | Size |
|---|---|
| mustbeatips.png | 40.8 KB |
Comments
Application?
I am not sure how your observersion can be used for making the life of translators easier.
A possible application would be to try to crossreference (long) strings in order to unify them. This would decrease the number of strings to translate, but only if they have not already been translated. The best time to check for existing similar strings would be before developers release a new module.
example: This is an
example:
This is an apple
This is an orange
This is a dog
This is XXX
we only need to translate "This is an/a" once and then copy them to next string. so we only to translate apple/orange/dog/XXX. at least it can save some typing time. also easier to keep translation in order to unify.
It's only for a large translation community. if you only focus on ONE modules. It will not help.