[Localization] Add a Named-Entity recognition middleware to perform localization of tutorials

[jibe-b]

My motivation

Let's let tutorial authors write in English.

Automated recognition of mentions of things like UI buttons, tools, standard workflows and so on may enable to add markup in the markdown file that will then be localized in the html result.

The learner will eventually read the tutorial with mentions Galaxy standard entities localized to their languages, possibly with a on_mouseover() mention in English. (Await: localization working group)

This process may encourage use of standard ids/label of entities instead of

Use of standard ids/labels may enable further features, such as stats on tutorials… (Await: ideas ;) )

Technically speaking

Mining tutorial markdown files

Needs:

• Galaxy terminology resources: lists of Galaxy UI buttons, etc, which are already used by the translation/localization working group -> enabling use of basic regex (level0)
• manual annotation of some tutorials -> enabling recognition of standard language (level1 :slightly_smiling_face:)

Adding a middleware…

• means having the "markdown to html" continuous integration script executing this script

Eating your own dogs' food

When Text-mining workflows will be made available by the OpenMinTeD community, the Named-Entity recognition step will be carried on by a Galaxy workflow :)

Disclaimer: I do work in the OpenMinTeD project.

Disclaimer

Now just let's do it… Please ping me whenever you think I should realize I said "just do it" and never did…

[jibe-b]

--> Alvis workflows

[jibe-b]

--> Makefile + Jekyll : https://github.com/galaxyproject/training-material/blob/master/Makefile

[jibe-b]

--> https://github.com/galaxyproject/galaxy/blob/dev/locale/en/LC_MESSAGES/ginga.po

[jibe-b]

.po https://www.icanlocalize.com/site/tutorials/how-to-translate-with-gettext-po-and-pot-files/

PHP style