Add thesearchpage application to nextcloud transifex

[callmemagnus]

👤 App repository

To be done by you

• Link to your repository: https://github.com/callmemagnus/nextcloud-searchpage
• If the app is not in the Nextcloud organisation
  • Invite nextcloud-bot with write permissions
  • Make sure all translated branches do not have branch protection enabled, see https://docs.nextcloud.com/server/latest/developer_manual/basics/front-end/l10n.html#setup-of-the-transifex-sync for more details
• Create file l10n/.gitkeep with empty content
• Add .l10nignore to exclude compiled JS files and thirdparty code, e.g. ignoring compiled javascript assets and composer PHP dependencies in the vendor/ directory:

js/
vendor/

• Create file .tx/config with the following content, replace {{APPID}} with your app id 3 times:

[main]
host     = https://www.transifex.com
lang_map = hu_HU: hu, nb_NO: nb, sk_SK: sk, th_TH: th, ja_JP: ja, bg_BG: bg, cs_CZ: cs, fi_FI: fi

[o:nextcloud:p:nextcloud:r:{{APPID}}]
file_filter = translationfiles/<lang>/{{APPID}}.po
source_file = translationfiles/templates/{{APPID}}.pot
source_lang = en
type        = PO

🏗️ Sysadmin team

• 👀 Ensure access:
  • Nextcloud organisation repository: Add nextcloud-bot with admin permissions
  • Other repositories:
    • Ensure the invite was accepted
    • Ensure the app certificate is linked to the same repository/author https://github.com/nextcloud/app-certificate-requests
• ⚙️ Ensure repository setup:
  • .tx/config
  • l10n/.gitkeep
  • .l10ignore
• ➕ Add "Owner Repository", into https://github.com/nextcloud/docker-ci/edit/master/translations/config.json
  • Pull request:
• 🔑 SSH into translation machine: ssh [email protected]
• ↩️ Change dir: cd /srv/docker-ci
• ⬇️ Pull: git pull origin master
• 🧑‍💻 Run the first sync manually:

docker run -v /srv/cronie-data/transifexrc:/root/.transifexrc \
    -v /srv/cronie-data/gpg:/gpg \
    -v /srv/cronie-data/ssh/id_rsa:/root/.ssh/id_rsa \
    --rm -i ghcr.io/nextcloud/continuous-integration-translations-app \
    AUTHOR APPID

🔣 Translation team

• Transifex: New resource should appear
• Transifex: Do translations and check sync to repo
• Add resource to wiki

[nickvergessen]

Will have a look, mostlikely on friday

[callmemagnus]

I merged your PR. Is there anything else I need to do ?

[nickvergessen]

So I synced the translations, but just for your information, the svelte seem to not be handled properly:
This string is not in the list of translated strings:
https://github.com/callmemagnus/nextcloud-searchpage/blob/main/src/components/Header/SearchBox.svelte#L91

[callmemagnus]

So I synced the translations, but just for your information, the svelte seem to not be handled properly:

Indeed, none of them are correctly handled :-(

Is there a way for me to investigate it further ?

[nickvergessen]

Maybe #475 (comment) can give you a trace?

[callmemagnus]

Ok. So I need to generate a fake .js file so that the translation CI can parse it ?

Is this a correct understanding ?

[nickvergessen]

Yes, that would work

[callmemagnus]

What is the function name that is search in the files ?

I see translate is exported from @nextcloud/l10n but the comment you mentioned talks about t.

[callmemagnus]

I pushed the script and the fake label file to develop.

Are you able to trigger the job to test it on that branch ? (Or can I do it ?)

[callmemagnus]

Ok I used the translationtool.phar and it expects t(

[callmemagnus]

Pushed a new version to develop that "works" with a local translationtool.phar

[nickvergessen]

callmemagnus/nextcloud-searchpage@3039ad5

Works now. Also the new strings are found:

[callmemagnus]

Thank you!

[callmemagnus]

@nickvergessen sorry to bother you again but I have the feeling that only develop is retrieved.

E.g. my app description is different on develop than on main but I only see the one from develop on transifex.

[nickvergessen]

all strings are combined, it starts with the master/main branch, so yeah the alter branches would win on the description it seems.

Hmm actually it should be the otherway around:
https://github.com/nextcloud/docker-ci/blob/master/translations-app/handleAppTranslations.sh#L37

[nickvergessen]

Ah, but here we say use-first:
https://github.com/nextcloud/docker-ci/blob/master/translations-app/handleAppTranslations.sh#L76

So that should be it.

[nickvergessen]

      --use-first             use first available translation for each
                              message, don't merge several translations

I guess we should turn it around, so that master/main wins

[nickvergessen]

So we fixed that in the code, but can't roll out a new version of the container until next week

[callmemagnus]

all strings are combined, it starts with the master/main branch, so yeah the alter branches would win on the description it seems.

Hmm actually it should be the otherway around: https://github.com/nextcloud/docker-ci/blob/master/translations-app/handleAppTranslations.sh#L37

Are the labels from the info.xml somewhat handled differently than labels from the code ?

I would have guessed that the scrapper just returned all values from main and develop in my case and made them available on the translation site.

My issues are:

• info.xml labels (only description from develop is uploaded to transifex)
• label updates in develop do not appear (what is the scan interval ?)

[nickvergessen]

Are the labels from the info.xml somewhat handled differently than labels from the code ?

Hmm after checking the code, the answer should be no:
https://github.com/nextcloud/docker-ci/blob/master/translations/translationtool/src/translationtool.php#L293-L314

It should work the same way, so not sure why the one from your master is not picked up