Back to LanguageTool Homepage - Privacy - Imprint

[SR] LT stopping after 1st sentence


I am adding support for Serbian language. After closely following instructions for adding new language, I started LT server and created web page that should send text to server for check and display errors. However, it seems to me that LT is not checking text past first sentence. For example, text:

“прва реченица. за њом и друга.”

is correctly identified as wrong, because first sentence does not start with capital letter. However, SECOND sentence also starts with lowercase letter, but is NOT labelled as wrong. What might be wrong?

Test from standalone LT GUI and simple check using “curl” command also do not get TWO errors, but just one.

Please help.

Have you tried putting them on separate lines? It’s possible that LT doesn’t know wether the mid-line period is a sentence-stop or an abbreviation/initial(ism)-stop (EG: Doe, Lt. Cmdr. J.)

I tried putting them in separate lines and - lo and behold! - both sentences were labelled as incorrect (which had to happen)!!! Thanks for your help!

Now, the question is: how should I structure my texts so that they are checked from first to last sentence?

It’s best to put any new sentence on a separate line until you are certain it is fully correct.

Putting multiple sentences onto a single line should wait until they are confirmed as correct.
I personally believe it also requires that they combine into a single scene-element, but that may simply be a matter of style.

LT on shows errors in ALL sentences, and they are all in a single line. How do I achieve the same with Serbian language checker?

I’ve just checked the wiki and it seems you need to add some segmentation rules.

I guess I was stupid. Thank you for your help! I’ll read about adding segmentation rules.