RTL languages and finding patterns (simple or regex)

Question

I’m working for the first time in PT with a right-to-left language project (I’ve only used RTL resources up to now). I’m finding I can’t match verse range markers with a regex, and on examining the SFM, I’m finding this:

• \v 7U+200F-8

Where U+200F is the <RTL> marker. It’s clear that this is what is stopping my regex matching. Why is PT putting that in the middle of the USFM code? Just before I checked it with a hex viewer, I deleted and retyped the marker, so PT seems to have inserted the <RTL> afresh.

I was trying to correct verse ranges with stray spaces in them. For the moment I’ve solved the <RTL> problem with a regex like this:

(?<=\\v\W)(\d+)\W*-\W*(\d+)\W*(?=\w) → \1-\2␣

(where “ → ” divides the Find string and the Replace string, and “␣” is a space.)

Answer 1

Paratext inserts the RLT marker to get the text to display as most users
have expected it to. The - will appear to be in the wrong location without
it. This has to do with the complications of having LTR strings of numbers
within RTL text.

There have been some who wanted verse references to appear differently than
they do in RTL text, but that would be a feature request.

John+Wickberg
Paratext Support