Bug: Spell Check Language

[Tex2002ans]

The Problem

If you double click on a word in the Spell Check window, it jumps to all instances of that word, completely ignoring Language.

Example Code:

<p>This is an example Spanish sentence: <span xml:lang="es">No señora, no estoy borracho</span>.</p>

<p>This is an example English sentence: <span xml:lang="en">No Ma’am, I’m not drunk</span>.</p>

When you double click on the Spanish "No", it leads you to all instances of "No" throughout the book, no matter the language. This becomes a problem when you are working on a large book with differing languages with overlapping words.

Steps to Reproduce

1. Open the Spell Check dialog window.



2. Double click on the Spanish "No". The cursor will highlight "No señora".

Double click again, the cursor will highlight "No Ma’am".

Expected Functionality

The Spanish "No" should only jump to "No señora", and the English "No" should only jump to "No Ma’am".

[kovidgoyal]

It's not really a bug, more a performance optimization, making jumping to words language aware means the the entire document has to be parsed to jump to the word -- which can be slow for large documents and is a price everyone has to pay all the time, for a relatively rare use case.

If you simply click replace, you will find that only the Spanish version is replaced, not the English one, so this only affect jumping to words not actual replacing.

[kovidgoyal]

On second thoughts, this can actually be made performant by using the syntax highlighting cache that the editor already maintains. https://github.com/kovidgoyal/calibr...acfda92306003f

[Tex2002ans]

Originally Posted by kovidgoyal

If you simply click replace, you will find that only the Spanish version is replaced, not the English one, so this only affect jumping to words not actual replacing.

I thought of this "potentially related bug" and tested it as well. I was surprised to see the Replace worked perfectly, even if this double clicking didn't!

Originally Posted by kovidgoyal

On second thoughts, this can actually be made performant by using the syntax highlighting cache that the editor already maintains. https://github.com/kovidgoyal/calibr...acfda92306003f

This is one of those things I didn't know I needed until I worked on a large book with multiple languages properly marked using XML (English/Latin/French/German/Greek). This should make it much more helpful when words are correct in one language, but wrong in another.

Now, to just get similar functionality added to Sigil's Spell Checker!

[kovidgoyal]

IIRC Sigil does not support multi-langauge spell checking, at all.