RTL text in line with LTR

Scrivener Wish List
1

I humbly beg for improved support for mixed-direction text, i.e., RTL languages in the same line/paragraph as LTR text. I am an academic working on biblical Hebrew, so I need to be able to write (in English) about words/phrases/sentences in Hebrew and have it all in one continuous paragraph. I find that if I begin writing LTR and switch mid-line to RTL and back again, it generally works as it should, as long as I compile to .rtf. But if I try to start with RTL and switch to LTR (which I need to do often in tables), it all becomes a jumbled mess.

I love Scrivener and really want to be able to use it for my dissertation, but the project is just too big for me to take the chance of things not working as they should. Please help!

2

Hi @jessntlee and welcome.

I’m not an RTL user, though I do use Chinese so am used to switching script types.

Is it only in tables that you have the problem of RTL → LTR or does it happen in ordinary text too?

The first thing to say is that tables in Scrivener are pretty primitive. Scrivener is based on the Apple TextKit, which is pretty deficient for that and for lists. KB might eventually get round to wring his own code for these, but that not only involves displaying in the editor but also means wwriting code for all the compile routines for the different formats.

So, if tables are a significant element of your dissertation, I’d create them in a word processor, then paste a screenshot or PDF into a Scrivener document (there are other ways of handling this but let’s keep it simple for the moment!). For choice, as Hebrew is the RTL language, I’d use Mellel or Nisus Writer Pro (NWP, though currently the future of that seems uncertain).

If you’re encountering the problem of switching RTL → LTR inline when starting in RTL, that surprises me a bit, but I’m not in a position to advise.

All that said, you are right to compile to RTF, as that is Scrivener’s native format and so requires no conversion. RTF is the Native format of NWP, but is opened perfectly well by Mellel, Word and LibreOffice. I presume the same for Pages too.

:slight_smile:
Mark

1 Like
3

Hi, Mark!

I have the problem with RTL to LTR in ordinary text as well as tables, though I more often have to begin with RTL in tables.

If tables were my only problem, I could work around it as you suggested. But I’m still having problems in ordinary text when I start with LTR. After testing it out more, I’m finding that I can switch from LTR to RTL (and back again) once or maybe twice per paragraph, but if I have multiple switches in a paragraph, it inevitably becomes jumbled at some point.

I was able to work around this for some earlier projects that didn’t incorporate as much Hebrew text, but I need to do more switching for this project, and I don’t see a way forward with Scrivener unless this gets updated.

Sadly,
Jessica

4

Hi Jessica,

I guess mixing LTR and RTL inline must create complexities that switching between 1-byte and 2-byte glyphs all in LTR doesn’t.

All I can suggest immediately is:

  1. Update your system to the latest Sequoia 15.7.1;
  2. Update Scrivener to the latest version 3.5 (though KB has just been responding to a drawing glitch noticed by another user with 3.5 and hopes he’s solved it).

Then, I just wonder if it would improve things if you used Scrivener for the organisation and research materials, but then:

  1. either write each section using NWP/Mellel/whatever, exporting/saving as RTF and importing that into Scrivener;
  2. or as an alternative, use External Folder Sync, sync’ing as RTF and actually doing the writing in NWP/Mellel/whatever.

What do you think? Worth trying?

There are, or have been, other forum members who write RTL but the last posts I remember from any of them were some years ago. I can only hope that one of them might still be around and pick up this thread.

:slight_smile:
Mark