Suggestion: Spell Check Tool Enhancement

[Tex2002ans]

Originally Posted by BobC

ang - Old English and enm - Middle English are probably what is needed for the Chaucer/Shakespeare stuff and do the job.

Fantastic, I didn't know about undefined, or Olde English, or Middle English either.

This is great info.

I just converted a book with a MASSIVE Middle English quote, and it was painful.

Originally Posted by BetterRed

Because I'll only use it when the olde worlde text is a nuisance, I think I'll stick with 'und' and delete it when I'm done spell checking.

[...]

If/when I ever use it in the HTML it'll be transient. I was straightening out a 19th legal text recently that had a boatload of lengthy M.E. quotations, that's the sort of occasion when I might use a 'fake language'.

That is sort of why I like the "calibre_ignore_spellcheck" idea, is that if you accidentally forget and leave it in, it doesn't hurt anything (sort of like the "sigil_" classes).

But accidentally leaving in the wrong lang code would actually do harm.

I do this for a living, and even I sometimes forget some little "temporary" code in there.

And if you are going to be wasting your time spending your time very wisely marking the text anyway, why not just stick the ACTUAL language code then?

Originally Posted by BetterRed

I'd been wondering if the Tex2002ans handle was a 'nod' to TeX, seemingly not. Maybe TPTB would allow you to upgrade the x to an X, then it could be so :lol:

I wish I started using LaTeX when I first heard about it (I remember my physics professor telling me about how amazing it is). I didn't start messing around with it seriously until this year... and it is GLORIOUS.

Did you see my comparison images here?

https://www.mobileread.com/forums/sho...6&postcount=58
https://www.mobileread.com/forums/sho...6&postcount=60

If I knew about it, I would have used it in all my papers for school! Do you know how much it would have demolished that Microsoft Word/LibreOffice JUNK?

I mean, once you join the LaTeX Master Race, you just don't go back!

[BetterRed]

Originally Posted by Tex2002ans

That is sort of why I like the "calibre_ignore_spellcheck" idea, is that if you accidentally forget and leave it in, it doesn't hurt anything (sort of like the "sigil_" classes).

A class based solution would be better, pity one can't put language code in CSS

Originally Posted by Tex2002ans

I do this for a living, and even I sometimes forget some little "temporary" code in there.

I do it for free for me, family and friends, I try to avoid changing the code as much as possible - often all I do is spell check, as long as it readable on a couple of kindles & android tablets that's all I need. I don't care much about whether the code is pretty, elegant, efficient, whatever.

Originally Posted by Tex2002ans

And if you are going to be wasting your time spending your time very wisely marking the text anyway, why not just stick the ACTUAL language code then?

Mainly because where does one stop. If I'm going to mark up Middle English quotes in a book written by a Pommie, then I should mark up Alyawarr quotes in a book written by an Aussie

Originally Posted by Tex2002ans

I wish I started using LaTeX when I first heard about it (I remember my physics professor telling me about how amazing it is). I didn't start messing around with it seriously until this year... and it is GLORIOUS.

I used Plain TeX a bit at DEC, then I went to Xerox and fell into the habit of Star-gazing

We're still left with what to do about numbers, link text that is a url, and excluding files within a session.

BR

[Tex2002ans]

Originally Posted by BetterRed

A class based solution would be better, pity one can't put language code in CSS

Yeah, that is part of the reason why I created my own class called "greek' which I also apply. So I can just easily find it (and treat it differently) if needed.

As I said, those are easy as pie since they might need a different font (and you can easily tell just from the characters that are used).

Originally Posted by BetterRed

I do it for free for me, family and friends, I try to avoid changing the code as much as possible - often all I do is spell check, as long as it readable on a couple of kindles & android tablets that's all I need. I don't care much about whether the code is pretty, elegant, efficient, whatever.

Even if you will be the only one seeing it, still good practice to follow some standards. Same with coding a little program for yourself, sure, you might be the only one looking at it, but you never know what might happen in the future (or writing comments/clean code for your future "self" (many times you look back at old code and think "who wrote this crap?" or "What the hell was I thinking?")).

Originally Posted by BetterRed

Mainly because where does one stop. If I'm going to mark up Middle English quotes in a book written by a Pommie, then I should mark up Alyawarr quotes in a book written by an Aussie

One span at a time... One span at a time... 1 correctly marked language span is better than 0. 2 is better than 1, 3 is better than 2...

[Sabardeyn]

If you ignore "words" of several digits entirely that precludes finding errors in things like phone numbers, addresses and other numerical data in a more casual manner (I expect professional docs would have additional oversight to make sure numerical data is correct). In this case I'm thinking more of a mask comparison: there are 27 uses of 12345 and 01 uses of 12346, is this correct? In this way the potential outlier is obvious.

As for single in-word digits, while not particularly common, you may experience difficulty with sci-fi stories including leetspeak. This may also affect things like program names (ex: Win8).

Granted, these are more likely to be edge cases than mainstream edits, but I thought I would mention them for consideration.

[BobC]

Originally Posted by BetterRed

We're still left with what to do about numbers, link text that is a url, and excluding files within a session.

BR

The excluding files actually highlights the point that the spelling checker only works on the entire book (unless someone knows different). I wonder if it would be possible to apply it one file at a time, handy for big books where it would reduce the "clutter" that needs inspection.

BobC

[BetterRed]

Originally Posted by BobC

The excluding files actually highlights the point that the spelling checker only works on the entire book (unless someone knows different). I wonder if it would be possible to apply it one file at a time, handy for big books where it would reduce the "clutter" that needs inspection.

BobC

In principle I think it makes sense to spell check the whole book, because the same errors often occur in multiple files, IMO better to knock 'em off once.

I also use spell check to checkout commonly misspelt words. I was cleaning up a paper from a law school (the one that begins with Y that had a student called Dubbya). Over 30 occurrences of breech that should have been breach and none that remained as breech

If the exclusion was done with exclude/include options on the file context menu then that should cater for my wish to exclude front/back matter files or yours to spell check chapter by chapter.

Originally Posted by Tex2002ans

Same with coding a little program for yourself, sure, you might be the only one looking at it, but you never know what might happen in the future (or writing comments/clean code for your future "self" (many times you look back at old code and think "who wrote this crap?" or "What the hell was I thinking?")).

No matter who wrote it and no matter how well, if its declarative code I always ask who 'who rote this crap'. What was wrong with Courier New on Gestetner stencils or the smell of hot metal anyway :lol: Just had a flashback to paper tape and Friden Flexowriters

BR