David Williams

David Williams said:

How can my customer import the translation back into WordPress without overwriting the original German version of the site?

This is probably the first question to answer.  Your customer needs to make sure they have created a multilingual Wordpress site. This page may be a good starting point:

https://www.wpbeginner.com/beginners-guide/how-to-easily-create-a-multilingual-wordpress-site/

David Williams said:

Can they check and/or correct the source file after exporting it from WP and before sending it to me?

I would suggest they make sure the source is correct in Wordpress and then when they export you are dealing with the corrected file.  I would not change the exported file as you risk trying to import a target file that won't import correctly.  I don't really have any experience to speak of here but this seems logical to me.  I did set up a multilngual wordpress site once some time ago just to play with itand it was quite straightforward.  The only problem was really the exported files which brings me onto the next point:

David Williams said:

Can I use the WordPress "wp_posts" table from the database, or which export file format is best (SQL, CVS, XML, etc.)?

WP does have an XLIFF export.  But their export is really poor.  However, you can deal with it quite (relatively speaking) easily using the Multilingual XML filetype rather than the XLIFF filetype as it will handle the large chunks of html WP stick into single CDATA sections much more comfortably.

I'd recommend you work with your client to set up the multilingual site in a test space on their server, learn how it works and what the options are, and test a few workflows.

Alternatively, if you have something like Trados Enterprise then you could use a connector, but you'd still need the multilingual wordpress site in the first place:

https://www.trados.com/solutions/connectors/

Someone with more experience here may be better to answer, but this would be my view.

Very many thanks, Paul! I would expect that the different language versions of the site would need to be in different subdirectories, but ideally with the pages in each language somehow interlinked, eventually.

Hi Paul, I have now received the XML export from WordPress and was able to load it into Trados, but several attempts to open it have failed, possibly due to the file size (An exception of the tpye 'System.OutOfMemoryExcpetion' is triggered), and now I am unable to restart Trados, even after restarting the computer, as it is unable to start the service 'Sdl.TranslationStudio.Api.ProjectManagement.IProjectManagmentService'. Any ideas on how to overcome this? I have also opening the XML file in Notepad++ and saved it as an HTML file, but obviously can't try that out until Trados will open again.

I deleted the projects.xml file and now Trados opens again, but with a "blank slate".

I have now also been able to successfully import and open a slimmed-down XML and HTML of the source, but they both contain lots of code in addition to the source text, so aren't terribly useful as source files.

David Williams

David Williams said:

but they both contain lots of code in addition to the source text, so aren't terribly useful as source files.

I guess it would help to see how you configured the fletype?  Wordpress doesn't make good files for translation, but if you configure the filetype properly you can normally work around them.  If you share an example I'll be happy to take a look.

From my experience over the last six years, with our WordPress setup using WPML, I have found the most efficient way has been to use code view to copy the entire post/page and then save it in Notepad++ as HTML files. The WPML export to XLIFF proved itself not worth bothering with. Over the years I have used various RegExes to handle the items in square brackets - it isn't perfect yet, but does the job.

Regarding "the items in square brackets" - I am referring to shortcodes relating to plug-ins - e.g. items in our DMS, tooltip glossary entries, accessibility related items (e.g. abbreviations, indications about the language of a word for a screen reader etc.)

Michael Bailey

Michael Bailey said:

The WPML export to XLIFF proved itself not worth bothering with.

Indeed it is very poor.  You might like to try the Multilingual XML filetype for this though... does a pretty decent job on their files now.

Michael Bailey said:

Over the years I have used various RegExes to handle the items in square brackets

Using the Multilingual XML filetype gives you the ability to use an embedded content processor for the html AND regex for the items in square brackets.  So the overall process becomes quite a lot easier once you have your filetype set up correctly.

Paul Thanks for this input - that gives me something nice to look into on the "Fenstertag" on Friday.

It might provide very timely as we are due to have quite an intensive time overhauling our website in the summer! Thanks!