Does TypeQL support query/sort of right-to-left languages? In particular, interested in biblical (also known as pointed) Hebrew (not modern Hebrew). If not, are there plans for such support in the future?
That’s in interesting question! We don’t currently support this – but I’d definitely like us to. Can you provide some links on how other databases handle these cases & any ideas you have on how you would like the feature to look in TypeDB?
There are two separate questions, the first being language support (Unicode) for entry and display, the second question is searching biblical Hebrew.
Language support: Your average web browser does a decent job of displaying Hebrew. I’m not sure which libraries they are using but you can’t go wrong with the libraries at: https://icu.unicode.org/ It’s been around for a long time and I am still surprised at data projects that don’t offer Unicode support and/or roll their own language handling. It’s really not necessary. But this would be a big step in supporting a wide variety of language, modern Arabic comes to mind (although usually printed without pointing), CJK and Hindi for example.
Searching Hebrew: A very difficult problem!
What first comes to mind is HebMorph - HebMorph - Hebrew Search Done Right which works with Lucene and ElasticSearch, although I don’t know the current status of the project.
See next post for 2 more links.
OpenScriptures: OpenScriptures Hebrew Bible has rather remarkable abilities.
Do be aware that most biblical data is available with verbal forms already recorded for all of the text. So the burden of encoding what is already know can be put upon the creator of the database. It’s sort of a cheat but in lieu of searching for wildcards (see next resource) it would meet a large number of needs.
I don’t know if source is available, but Hebrew Bible Software - search module for the vocalized text has a subsection that covers searching with pointed text and wildcards. I haven’t personally used this software but I take it as a good example of what higher end searching could offer.
TypeDB itself doesn’t do anything special with strings. It will store the sequence of UTF-8 bytes provided and fetch them to you on request. There is no special handling for sorting or search, I’m afraid.
Regarding display and entry, are you referring to TypeDB Studio? The editing of RTL text isn’t currently fully supported, but the display seems fine as far as I can tell.