[ARCHIVED] Supporting multiple languages for course content

[maguire]

I am a member of the KTH Royal Institute of Technology reference group on language issues (Språkkommittén (Links to an external site.)) and KTH wishes to have English and Swedish language content in parallel. I have tried using  enhanceable_content tabs to support multiple languages in a page (see examples at Using tabs for content in three languages: Chip sandbox). It seems to work, but I expect that there is probably a better way.

Moreover, there is a problem in that if a user has selected their user interface language in their user settings, then the content in this language should appear as the "top" (i.e., visible) tab when navigating to a given page. Similar once the user selects a language by selecting a tab, it would be good if the "Next" and "Previous" buttons took the user to the relevant page and continued to show the selected language version of the page.

Are there any good examples or best practices that others have used for supporting multiple languages in parallel?

[johan_holmberg]

Gerald, did you ever find out a good way of doing this?

[maguire]

I've used the method above, you can see the results at Using tabs for content in three languages: Chip sandbox  This has been used a current course for Master's degree projects to provide material in both Swedish and English.

Recently, I have been thinking about what could be done if the check for the user's language setting and the choice of what to render on top was done in the institution wide javascript that is loaded before the Canvas course page is loaded. One way to check the user's language choice is describe by @Fede_Arbelaez in the posting Identify the language of user with JS . It should also be possible to check the user's language setting in Canvas via the Canvas API.

[johan_holmberg]

Thanks! It's a rather elegant solution, but not very friendly for non-techies.

[maguire]

Actually, it is not too hard to do. I often make a partial page template
and then it is just a matter of pasting it in as HTML and then filling
in the content. The result is that the person filling in the content
really does not have to understand what is behind it, they simply
replace existing dummy strings with their content. For example, I made
the following and then copied with just cut and paste in the RCE editing
mode the first EEEEE line and pasted it in and modified it as EEqqqqqEEE
and this work (but RCE wrapped a <a href="#fragment-1">English</a>

<a href="#fragment-2">På svenska</a>

EEEEE

EEqqqqqEEE

EEEEE.

SSSSSS

SSSSSSSS

Alternatively one could use some other markers in the page to indicate
which is English and Swedish and then post process it with a program to
generate the correct HTML. During the summer I did this while converting
some assignments that had javascript in the pages and used the fact that
the HTML supports UTF characters so I used a pair of characters that
were very unlikely to appear in any page to mark the regions of the page
that I wanted to process further. I used the following pairs:
a Caution sign U2621 followed by UMBRELLA WITH RAIN DROPS (U2614)
a WARNING SIGN (U26A0) and UNIVERSAL RECYCLING SYMBOL (U2672)
and a WHITE FROWNING FACE (U+2639) followed by a UNIVERSAL RECYCLING
SYMBOL (U+2672).

[maguire]

I have done an example with markers to make it easier to post-process the text. I used one pair of markers for the English text and another pair for the Swedish text. These can later be turned into language tags for the indicated text.

See Copying two language content: Chip sandbox  

One could probably use emoji symbols for national flags, but I could not get them to appear in my old chrome browser and this is not a good indication of the language versus an indication of a country. Are there standard unicode indications for languages as opposed to countries?