Saved Search/Regex Functions

[theducks]

Originally Posted by eschwartz

We have a sticky thread you can post this in.




@theducks, would you mind fixing the thread title for that sticky? I think it predated Function-Replace mode.

"Saved Search" ==> "Saved Search/Regex Functions"

Done

[roger64]

Hi

This approach is very promising and interesting as it allows to correct some recurrent OCR mistakes (after a dictionary check).

In French, we have a list of hundreds of words that deserve to be checked out of OCR. With such an approach, of course, you can't avoid false positives, which means we need not to correct them but just to highlight them, to be able later to speed up a manual checking.

I give one example, if you find "trame" which is a correct but fairly rare word, 98% of the time, it should be written "traine". It makes sense to highlight it.

However, I do not know how to write a single entry.

Could some kind soul write a code paragraph example of this Calibre function allowing me to highlight "trame"?

[kovidgoyal]

Simply modify the function to wrap the word in some markup instead of correcting it, so instead of replacing the word trame with its correction, replace it with <span class="mistake">trame</span>

Then you can use another search to jump to these words one by one and correct them or not, as you like.

[roger64]

@kovidgoyal

Thanks for your neat idea which is so simple that I understood it.

Now to try it.

[Arjayem]

smack: If that's aimed at me, I'd be happy to oblige, Regex Functions seems an apt tag. BUT I'm new to this, and the suggestion goes way over my head, can you step through it ?

Originally Posted by eschwartz

We have a sticky thread you can post this in.




@theducks, would you mind fixing the thread title for that sticky? I think it predated Function-Replace mode.

"Saved Search" ==> "Saved Search/Regex Functions"

[eschwartz]

YOU just repeat your post, in that sticky thread.

@theducks has already performed the moderator administration duties I asked for.

[BetterRed]

Originally Posted by eschwartz

YOU just repeat your post, in that sticky thread.

@theducks has already performed the moderator administration duties I asked for.

Rather than the OP doing doing that, and to avoid having 'duplicate' posts, I suggest theducks merges this thread into the sticky "Saved Search/Regex Functions" thread.

BR

[theducks]

Originally Posted by BetterRed

Rather than the OP doing doing that, and to avoid having 'duplicate' posts, I suggest theducks merges this thread into the sticky "Saved Search/Regex Functions" thread.

BR

I am worried about breaking things
The forum (Mod actions) prompts are a bit

[BetterRed]

Moderator Notice
Merge done

[nqk]

Dear you guys,

Please help me with a regex function to move found searches to "endnotes.html" file in the book. I picked a code from this forum, which looks like this.

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    endnotes = open('D:\endnotes.txt', 'a')
    notes = match.group()+'\n'
    endnotes.write(notes)
    return ''
replace.file_order = 'spine'

This works, yes. It moves all the texts found to the external endnotes.txt and I would manually copy them to endnotes.html that I create in the book.

What I would like to do is have them written directly to the endnotes.html (or even create the file if not there already).

Thanks.

[roger64]

Tagging selected foreign words

If we manage to wrap this

<span xml:lang="xx" lang="xx">foreign</span>

around selected foreign words, we may help improve TTS reading. Foreign words will be pronounced with their native accent instead of an haphazard and sometimes unintelligible way.

The requisite is to write first a list.txt of these selected words -which can be easily selected and copied from the spellchecker panel of unrecognized words (to whom they all belong). For the purpose of this thread we suppose that such a list is available.

Help requested

I am looking for a function that I could launch on an ePub with the Calibre editor and which would use sequentially this list to wrap the above spans around each occurrence of these foreign names.

[roger64]

Search and replace excluding the headers

I am trying to insert old nums in an ePub. For this I need to wrap a span around these characters like:

(\d+)
<span class="smcp" >\1</span>

However I need to exclude the headers, if not I have to correct all of them...
Could a function do it?

[davidfor]

Originally Posted by roger64

Search and replace excluding the headers

I am trying to insert old nums in an ePub. For this I need to wrap a span around these characters like:

(\d+)
<span class="smcp" >\1</span>

However I need to exclude the headers, if not I have to correct all of them...
Could a function do it?

If this is just numbers in paragraphs, you could do:

(<p>.*?)(\d+)(.*?</p>)

With the replace:

\1<span class="smcp" >\2</span>\3

That does have a problem if a paragraph has more than one number in the paragraph. The following might be better, but, it still has issues:

(<p>.*?[^\>\d])(\d+)(.*?</p>)

[roger64]

@davidfor

Thanks very much for your help. I'll try it.

In regex-function mode "create/edit" brings up a dialogue box with two sections: 1. function name and 2. Code.

Function name has a drop down list of about a dozen functions. I thought that choosing a built-in function would populate the code panel with the appropriate Python code so I could learn from it, but it does not change from the default:

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
return ''

There are a few examples in the User Manual, but not all of the built-in functions, which would be handy. Where might I find the code for those functions and others? Is there, perhaps a thread the saved searches thread?

If this is some secret that a novice shouldn't know, please let me know. I don't want to accidentally cross the beams and annihilate the universe.