I’m writing an article that is primarily in English, but includes Hebrew text sprinkled through it. When I compile a draft and export into several formats (rtf, doc) the Hebrew words are reversed. Also, when I try to correct the spelling in the exported file, the characters shift around oddly.
The Hebrew characters export correctly in odt and docx form, but then I lose other formatting. Is there a way to address the exporting problem with doc files?
You might try moving the Hebrew sections to their own documents surrounded by the English documents. I think, in OS X, a document can only be left-to-right or right-to-left. It cannot have sections that are one way or the other. So try to moving the Hebrew out by itself, and then right clicking and setting the writing direction right-to-left for those snippets themselves.
Actually, it can, 'cos Nisus and Mellel can both do it. Mellel uses its own text engine, of course, and in Nisus you can mark individual stretches as being in a particular language, automatically associating font, direction and keyboard. I think I’ve read on the Nisus forum that there are direction-markers that get inserted in the text — in what form I don’t know — because it seems that if you copy text in from a different source, you may need to change the direction.
But I still think you’re right in Scrivener to put the RTL text in a separate document.
I can’t help with things shifting around in the exported files as presumably that is more to do with the programs in which they are opened, but as for the text getting reversed:
Are you using “override text formatting” in Compile Draft > Formatting? If so, this would override the text direction. If this is the case, Amber’s suggestion is a good one, but you would need to set up the formatting of the separate Hebrew documents and turn “Preserve Formatting” on. However, as you say that ODT etc exports this part fine (but has other problems) I suspect you don’t have “override text formatting” turned on.
Try creating some test text with different writing directions in TextEdit, then try exporting the different formats from there. Does that have the same problem? Scrivener uses the same OS X text engine, so if the problems are present there, then it is something to do with the OS X exporters.
I’m trying a demo of Scrivener and have high hopes it will serve my needs.
However, I also will need to create documents that utilize both L-to-R and R-to-L languages. Are you suggesting that every time I need to include a word or two in Hebrew that I need to make a separate document just for those words? That seems very cumbersome.
I was planning to export the document to Mellel.
Thanks for any help in understanding how to best incorporate Hebrew.