Ideas for New or Updated Plugins

[KevinH]

Hi All,

I am personally interested in new ideas people might have for new Sigil plugins.

I am also interested in identifying useful but "abandoned" Sigil plugins that need help to work with current Sigil (assuming they have an opensource license that allows for changes and redistribution). For example, things like moving to PySide6 from PyQt5 or from tk, epub3 support, bugs fixed, etc.

So please let me know what you think.

I can not promise to create or update these but will consider any ideas seriously when I have free time.

Thanks,

KevinH

[RbnJrg]

It would be great a plugin to translate an epub from one language to another (for example, from english to spanish or whatever). Calibre has a great plugin to do this task but it would be fantastic to have it also in Sigil.

[KevinH]

Originally Posted by RbnJrg

It would be great a plugin to translate an epub from one language to another (for example, from english to spanish or whatever). Calibre has a great plugin to do this task but it would be fantastic to have it also in Sigil.

Did you look at the plugin's insides? What tools or web-based translation engines are they actually using?

Translation using web-based or AI translations tools are still pretty poor in my testing (ie English to German and back for round-trip testing), forcing a full proof read and much restructuring of text anyway. In documents with more technical terms, the output was actually laughable.

Have the engines improved that much in the last 6 months?

[RbnJrg]

Originally Posted by KevinH

Did you look at the plugin's insides? What tools or web-based translation engines are they actually using?

Translation using web-based or AI translations tools are still pretty poor in my testing (ie English to German and back for round-trip testing), forcing a full proof read and much restructuring of text anyway. In documents with more technical terms, the output was actually laughable.

Have the engines improved that much in the last 6 months?

The plugin allows you to use practically all engines (for free, like Google, or other paid services; DeepL is great). I have translated many novels from english to spanish (by selecting the Google engine) and the results were very good. And translations by AI are practically perfect (from english to spanish; I can't opine about other languages).The main issue is not the translated text but to change the format of dialogs in the novel (as you know, in english a dialog is enclosed in quotes but in spanish dialogs start with the em dash) and of course, the plugin doesn't do that.

[DNSB]

I tried translating several tech manuals and the results were uniformly humorous.

[RbnJrg]

Originally Posted by DNSB

I tried translating several tech manuals and the results were uniformly humorous.

Did you do it with the Calibre plugin? I don't know about tech manuals but with novels, translations work fine (with the plugin).

[DNSB]

It was with the calibre plugin. I tried a couple of novels and they weren't too bad. The tech manuals were laden with industry jargon and the translations were more than slightly amusing. One manual was for a Cisco device and comparing the machine translated manual to the French language version downloaded from Cisco made it obvious that quite a bit of jargon was either not translated or was horribly mistranslated.

[mrprobert]

Hi,

I have two ideas for plugins that would be useful to add to my automate list that I use to "clean" epubs.

1. Rename all the text file extensions to .xhtml. Of course, duplicate filenames would be an issue. I've seen more a few epubs that have a contents file named toc.html along with a Nav file named toc.xml.

2. A sort plugin for css files.

Thank you for your consideration.

[KevinH]

But if you sort css entries in a css file you can and will change its meaning as order matters in css. Also the order that link refers to them matters.

As for changing file extensions and mime types, that sounds like a job for an input plugin.
What generates txt extensions when it is an xhtml file? I have never seen that.

You do know you can use Sigil's BookBrowser to mass rename file extensions.

[KevinH]

For those interested in the RegexFunctionReplace plugin it has now been updated to work with newer Sigil:

see this post:

https://www.mobileread.com/forums/sh...5&postcount=23

[mrprobert]

Originally Posted by KevinH

But if you sort css entries in a css file you can and will change its meaning as order matters in css. Also the order that link refers to them matters.

As for changing file extensions and mime types, that sounds like a job for an input plugin.
What generates txt extensions when it is an xhtml file? I have never seen that.

You do know you can use Sigil's BookBrowser to mass rename file extensions.

Points 1 and 3 are absolutely true. I am not clear on what you mean by the second point. No matter. I figure you are half way there to bulk renaming in an automate list because of the SetBookBrowserToAllHTML function.

[KevinH]

The second point was actually asking what tools are you using that generates an xhtml file with the file extension of .txt?

I have never seen that or come across it before. In the Opf manifest do these files have the proper application/xhtml+xml media type?

If these are just pure text files and not .xhtml files, you can use Sigil's text importer to do a simple conversion.

[mrprobert]

I'm sorry if I wasn't clear about how I would use this function. I just want to regularize text filenames to account for .xml, .html, .htm and .xhtml file extensions.

I called them text files because the default location for xhtml files in a Sigil-normed epub is relative to ../Text/. I figure if Sigil puts a file in that section it is fair game to change the extension to .xhtml.

I think there was a post from Diap a while back in which he said, if I recall correctly, that the thing he does when first editing an epub is to rename files to .xhtml to reflect that they are in actuality, xhtml files.

[KevinH]

Yes and that can be done in one easy step using Sigil's BookBrowser. Select all and right click on rename. When it asks for a pattern, just replace what is there with .xhtml

All selected files would be properly renamed to have the .xhtml extension. Then of course behind the scenes all of the links and references to the old names all throughout the epub are properly updated using concurrent threads.

It is this last step that makes this more than a bit of a pain to handle in a plugin. The actual renaming of the files is trivial. It is what comes next (ie finding and performing all the reference updates) that is hard. Simply duplicating all that fast C++ and C code in python is just not worth it IMHO.

[mrprobert]

Originally Posted by KevinH

Yes and that can be done in one easy step using Sigil's BookBrowser. Select all and right click on rename. When it asks for a pattern, just replace what is there with .xhtml

Yes, that's the way I do it now. My goal is to automate the task. I see your point about the underlying adjustments. Maybe this could be easily implemented by an Automate List tool?