Markdown image format segmentation/processing

We have some formatted images in our MKDocs markdown files (Translating The Combine to Spanish, Mexico language - Crowdin), such as:

  • ![Login](images/login.png){ .center }
  • ![BasicSettings](images/projectSettings1Basic.png){ width=750 .center }

Crowdin’s default handling of these is inconsistent:

  1. On upload, { .center } is not parsed as a string to translate, but { width=750 .center } is, and thus has to be hidden.
  2. When translations are downloaded, ){ width=750 .center } is fine, but ){ .center } has messed up spacing: ) {.center}, and the new space between the ) and { messes up the MKDocs build.

Hi imnasnainaec,

Can you please share with us the example of your file to, so we can look closer into this issue?


Wanted to say that we’ve released a few improvements and bug fixes regarding MD and MDX files, you’re welcome to check out the current performance!

We are always grateful for the honest feedback!

@Dima Thanks for the notice. I just re-uploaded a source .md file and re-download a translation file and the above two issues are still present.

Hi @imnasnainaec !

We are deeply sorry to hear that!

Let us please consult with our tech team on the matter!

Hi @imnasnainaec ! we checked the project once again and back with some updates
the file has the old version of parser (35) and we do not support it anymore. With the new one (36) everything should work well. To update the version, the source file should be uploaded to the project from scratch (if you have the duplicate option, then translations from the old file will migrate to the new one). Let us know if you need help with that
As for the for example, the syntax !!! note “Note” is not typical for this file format so our tech team will have to investigate this (task created)

Thanks @Ira. I don’t have the “duplicate” option, but that would be helpful. Is there a way to enable that?

I uploaded a new copy of one of the files and am now seeing consistent behavior for both #1 and #2 in my original post. Thanks for fixing that.

#2: this was resolved in the preferable way, not adding a breaking space. Thank you!

#1: this was resolved in a less preferable way: I now have to manually hide both {.center} and {width=750 .center} strings.

@Ira The !!! note "Note" is a bit of an odd case, since I don’t want !!! note translated, but do want Note translated. The current set-up presents the whole string, which is fine.

I’ve just encountered a new minor issue in an .md file I just re-uploaded.

For example, the text paragraph inside

!!! note "Nota"

If the project has glosses in multiple languages, those languages must be added here for all the glosses to show up
in [Data Cleanup]( Click the magnifying glass icon to see all language codes present in the project.

has two sentences, which previously were presented as separate sentences to be translated. Now they are not split, so the whole paragraph has to be translated together.

Hello @imnasnainaec

It’s great to hear that the improvements worked fine for you!

As for duplicates, you can enable them on the project level → Project Settings | Crowdin Documentation

As for splitting the string in 2, you might try some pre/post processor to customize the system importer and enhance your experience: