NEW in Rise: Export for Translation

Plus one to Pedro comment -- 
purple tags in the middle of the sentence confuse the translators and cause extra tress as they are worried of removing them occasionaly and destroy the formatting in the output course.

Well, tags in the middle of sentences are probably formatting that needs to be retained in the translation, so that's fair enough. But tags at the beginning and end of every paragraph I do have a problem with.

Thanks for that Jeff. I'm actually using MemoQ which does support regex, and also custom search/replace on import/export. The problem I'm thinking is that if you remove the tags, then there still needs to be something at the location in the file so that we know where to restore the tags, so there's no point. And they are not predictable - they actually do not occur on every paragraph, just most (I think it's headings where they don't), nor are they identical (different IDs each time). Also multiple sentences in a paragraph mean that opening and closing tags may be in different segments.

I continue to get import failed error with no explanation when I am trying to import the translated XLF file. Please let me know what else I can try. I also opened the XLF file in word to translate manually .

Hi, Pedro. I'll share your suggestion with our team, and we'll let you know if we make any changes that will help.

Hi Jeff,

Do you know if this problem has been resolved?

Mike

mshaw@silkweb.com

888.618.1118

https://silkweb.com/

Hi, we are experiencing a very strange problems with our courses. The content is in Finnish and so are the lables. Still, the html-language is in English, and we are not able to change it. There is some script in the package that prevents it. 

Earlier we found out that screen readers are not able to read the text properly (as our languages are pronounced very differently). 

Now (during past couple of months) some users complain that our courses are difficult to understand, that the texts are just bad translations! And we use tens of hours to make sure we use simple, coherent and understandable language. 

Our guess is that some user's browser (Google Chrome?) auto-translates Finnish text into Finnish as it thinks the content is in English. And the result is bad Finnish - English - Finnish. See the attachment that shows how. Not all is wrong, but much of the text is unreadable.  

Could you please fix this problem urgently! This is not the user experience we want to have :(

Hi Chrystal, We've tried that and unfortunately it doesn't work.

Here's a link to a video that shows how main.bundle.js excecutes a script, that changes the language back to en, no matter what we write there.

https://drive.google.com/file/d/1-uZw1w2yAJ21xLg6beIHp_m6SmnWWoOr/view?usp=sharing

Thank you so much! This works. It also fixes the problem we had with accessibility with screen readers. What a relief for our whole team. Have a great day!

Hello all,

This post is about using memoQ for translating Rise files.

I took a look at this forum because one of our customers was experiencing difficulties with all the tags included in their files, and the fact that the XLIFF content contains paragraph-based segmentation as opposed to sentence-based segmentation.

First, it must be said that the Rise developers made a big mistake when they architected the product output to include HTML embedded in XLIFF. The two file formats do not play nicely with each other. While it is true that XLIFF is a translation industry standard file format, embedding of HTML was not envisioned in the specification.  The Rise developers would have been much smarter to output in plain XML with embedded HTML. This really should be rectified in a future version of Rise.

With that little rant behind me, I have some good news. memoQ can do a pretty nice job of separating the translatable content from the surrounding tags while also separating the paragraph-based segments into their individual sentence-based component segments. To do so, it is necessary to use a cascading filter, which is a special feature of memoQ. A cascading filter essentially filters the input file for one format first, then for a succeeding format.

In this case, the first filter to be used would be the built-in XLIFF filter. That provides the basic parsing of the content. Then, it is necessary to protect the embedded HTML tags by appending on a second filter using the Regex Tagger. The Regex Tagger uses regular expressions to create generic protection for similar HTML tags that contain differing attributes. To separate out the individual sentences from their parent paragraphs, the option "Segment text if no <seg-source> is present for a 'trans-unit'" should be selected.

This results in nicely segmented source content in the translation editor. But it also produces a lot of superfluous segments that contain only tags. memoQ can help here, too. These segments are superfluous for translation, but must be kept intact so that the ultimate structure of the output file is retained. They need to be protected from inadvertent alteration.

Once the translation environment is open, the special memoQ regex expression "^\itag$" (without the quote marks) can be used in a search to isolate all segments that contain only tags. Then, the CTRL-L key combination can be used to lock the segments. A further setting in the "Go to next segment" can be used to jump only from unlocked segment to unlocked segment, thus ease of use for the translator is ensured.

This has been a pretty technical contribution. If the folks from Articulate would like, perhaps we could cooperate on a webinar to present this solution to interested users who have been struggling with using translation tools  for processing the Rise XLIFF file output.

Best regards,
Richard Sikes
Senior Solution Architect, memoQ