Localisation test for invalid Strings

[LinusDietz]

@mlep suggested this in #3184 (comment).

There should be another localisation consistency check that detects invalid characters like spaces in the localisation strings.

[steph37tours]

Hi, I am a very beginner. I would like to try to help. Could you give more details about this issue to help me to find a solution ? Thank you very much.

[Siedlerchr]

Hi,
thanks for your interest in JabRef. The problem is about the localization keys and their translations in the l10n properties file, e.g. here for English:
https://github.com/JabRef/jabref/blob/master/src/main/resources/l10n/JabRef_en.properties

A solution would be to add a consistency Test to make sure that there are no whitespaces in the keys and the translations. Whitespaces have to be replaced with underscores.
We have already a Test class with consistency checks:
https://github.com/JabRef/jabref/blob/master/src/test/java/org/jabref/logic/l10n/LocalizationConsistencyTest.java

You can simply expand that for a new test case.
Some general information about contribution can be found in our contribution guide https://github.com/JabRef/jabref/blob/master/CONTRIBUTING.md

[LinusDietz]

Hi, @steph37tours!
I guess the first step is to check if there are spaces in the localisation strings.
src/main/resources/l10n contains all localisation files. A possible check should detect any spaces in these files that are not commented out (using the #).

EDIT:

For example

Finished_automatically_setting_external_links.=Selesai_pengaturan_otomatis_tautan_eksternal. is fine

Finished_automatically_setting_external_links.=Selesai_pengaturan_otomatis tautan_eksternal. should not pass your test

Make sure to open your pull request early, we are happy to help you!

[steph37tours]

Thank you for your quick answers ! I look at this and I come back.

[steph37tours]

Firstable, thank you very much for your help !
So I write the beginning of the method. But I'm not sure for instance how to parse all the properties object. And after that, how to replace the spaces by an underscore.
It was too my first pull request.

[LinusDietz]

Well, I think you shouldn't replace the spaces, just let the test fail, so the author will be notified that there are invalid characters in the localization properties file.

Regarding the parsing, just read in all properties as done in the other tests of https://github.com/JabRef/jabref/blob/master/src/test/java/org/jabref/logic/l10n/LocalizationConsistencyTest.java and then check if there are spaces in the strings.
Make sure to include a meaningful message to the failing test, so the invalid string can be found!

Good luck, and give us a shout if you need further help!

[steph37tours]

I just push again. I think it's good. Does I have to make a new pull request ?

[steph37tours]

I use properties.replace(" ",""). It was not possible to use replaceAll. I think it's not a trouble as long as the assertEquals compare the String object, with the modified. Just one replace make a difference.

[steph37tours]

I try to test by put a space in properties file, but I didn't have error after launch...

[LinusDietz]

Let's discuss the PR in the PR-Thread.

You can run the tests locally in your IDE or using gradle check.

[aayush2610]

Hi Guys, I would like to help if the issue is not yet solved. It would be great if I could get some context. Thanks

…

On 5 September 2017 at 20:29, Stéphane Curet ***@***.***> wrote: I try to test by put a space in properties file, but I didn't have error after launch... — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3185 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AHARASgsPnrv-8lBASXvaGM8HVGInWr_ks5sfWG7gaJpZM4PKM29> .

[LinusDietz]

Hey @aayush2610, Mr. @steph37tours is working on this at #3198, and I'm sure he will resolve this soon.

Feel free to jump into any other issue. There are some labeled with beginner which we think are easier to resolve, even if you are new to JabRef.

[LinusDietz]

fixed in #3198