Thread: Bidi content

Hi colleagues! I feel like pressing the red button and throwing a nuclear bomb. But...hold! Not for bad, but because of the many paths this post could take. As a Project Manager, I'm frequently dealing with localization projects dealing with a mix of LTR and RTL languages.

The biggest problem comes when I need to mingle with both directions within the same paragraph, and I'm willing to share some tips and tricks I learned from other colleagues.

Stay tuned, and, of course, I look forward to your contributions, too!

Just to get the ball rolling, let me tell you that several applications show features that will allow you to use bidirectional text within the same content.

I bet you are MS Word heavy-users, and for that Word has to buttons you can easily add to your ribbon by using the Options menu, clicking on the item on how to customize your ribbon, selecting to show "all commands" from the pull-down, and simply scrolling down until you come across two different options "RTL run" and "LTR run." Then you go to the right hand side of the same window and create a menu item (maybe under "paragraph" style) that will bear those two commands together. You click on "accept/apply" and...voila! You are ready to deal with your next BiDi file.

Hi Gentle,
I usually deal with RTL Word documents and I always asked myself how to work with sentences that combine RTL and LTR texts. Brackets and some other symbols don't appear correctly specially when you have a line break in the middle. I use to force line breaks when a LTR block text begins but I'm afraid that´s not the correct way to solve the problem. So, any other info you can provide will be fine!

Thanks!
Santiago

First off, you can start by going back to my post above (the one with the screenshot). With that, you can practice "embedding" text in the reverse direction with the help of a couple of opening & closing invisible marks as the ones added with the help of Notepad.

I know MS Word has a similar set of marks under the menu "Insert/Symbols" but I still need to figure out the ones for embedding.

Thank you gentle for the trick. From a PM point of view, I have never handled mixed RTL+LTR projects, but I sure will use this trick if so. However I'm left with one question to you gentle: let's say we have a paragraph in English, with text in Urdu between parenthesis (within the same paragraph), will this RTL/LTR run in Word automatically recognize the English and Arabic text (and set them as LTR and RTL)?

Whether you have mixed or single RTL language projects, the output remains the same in my opinion: after the full process with software and computational tricks is done, a native of the RTL language must proofread the job done by DTP. Indeed, even though you use this or other tutorials to make the mixed/fullRTL paragraphs look "correct" (let's say), you won't be sure until you have it proofed. And I'm only talking about simple cases, you could get to have the proofreader working alongside the DTPmanager in a back-and-forth feedback kind of scope...

Originally Posted by nabylm

However I'm left with one question to you gentle: let's say we have a paragraph in English, with text in Urdu between parenthesis (within the same paragraph), will this RTL/LTR run in Word automatically recognize the English and Arabic text (and set them as LTR and RTL)?

Hi Nabyl!

I'm glad you ask because it needs to be crystal clear for everyone that they need to add these "hidden" marks (actually, they work as real "tags") that will show the application when start switching the direction of letters and words.

The question would be...does they need to be applied in the target? What if I prepare the source file instead? (What would be an "internationalization" for a later "localization" into RTL languages.)

So far, and in MemSource, I've tried by applying them in the target while translating, and it seems to work. That is, once I look at the preview the Arabic text goes RTL while the English one remains LTR.

I'll let you know what happens if I "prepare" the source by adding those marks and if it later is transferred to the target with the translation environment.

Stay tuned once more! ;-)

I understand.... But let's think of an example (we PMs like to be down-to-earth). I have an english file that needs to be translated into arabic. Let's say I am using memsource. The context contains a lot of parts that will remain in english. I upload the file, do the translation and download the completed. I would have to have someone to copy/paste (with the "hidden" marks/tags) each LTR parts back to the translated file... Internationalization....
It would be good to try putting all these tags beforehand, and upload the file to a translation software to check if they are conserved.

Forgive me if I went too fast and misunderstood a part.... but I agree it's a very important point that needs to be crystal clear.

You are almost there!

"Internationalization" implies preparing the source file to support BiDi content. In other words, it would entail the implementation of these "invisible" marks/tags in the source wherever that text should remain in the source LTR language (English or else).

My question is on how the translator may preserve those marks... I imagine her copying the source segment into the target (somewhat easy with all CAT tools nowadays, like by a simple Ctl+Insert) and later replacing all text that should go translated, without touching (maybe even getting close!) to what needs to remain LTR. Beware the same would apply for quite the reverse (ie., translating from Arabic or any RTL language into a LTR one). But that is something I still need to check.

Meanwhile, I bet the linguist will need to work with either Word or Notepad open and manage marks and text in there, and then cutting and pasting that into, let's say, MemSource. This is what I tried, and, at least so far, it worked.

Is that clear?