[GUI Plugin] Modify ePub

[Kaetrin]

excellent, thx!

[LadyKate]

I just changed the language on an epub and it will not update. The error is

Modify ePubs
Logfile for book ID 111824 (Bookname / Author Name)
111824
Modifying: C:\Users\Kate\AppData\Local\Temp\calibre_hyfut4\pl _lz0_modify_epub\111824.epub
Updating metadata and cover
Parsing xml file: content.opf
Parsing xml file: toc.ncx
Looking for files to remove: [u'META-INF/calibre_bookmarks.txt']
BookName - ERROR: Traceback (most recent call last):
File "calibre_plugins.modify_epub.modify", line 79, in process_book
File "calibre_plugins.modify_epub.modify", line 134, in _process_book
KeyError: u'strip_spans'

ePub not changed after 0.19 seconds


This epub did not have any bookmarks but the plugin is configured to remove bookmarks, update metadata, smarten punctuation, remove metadata covers

This is the first time I've tried to use it since updating to 2.x of Calibre.

[davidfor]

@LadyKate: Can you make sure you have the latest version of the plugin installed. This is 1.3.12.

If that doesn't help, delete the saved settings for the plugin. To do this, open the calibre preferences, go to the "Miscellaneous" section and press the "Open calibre configuration Then go into the directory "plugins" and find the file "Modify ePub.json" and delete it. The restart calibre and try the plugin again.

[LadyKate]

Thanks for responding davidfor. I must have needed a few more restarts of calibre. I had updated the plugins... all of them, and restarted BUT after making a couple of other changes and restarting about six times the next time I tried (which I did after almost every restart) it worked. Must have been in limbo or something. Sorry for bothering all.

Must be because it was a major update, I hadn't updated from 1.48 until tonight.

[zephyrot]

Hello.

Thanks for such a great plugin!.

I do know nothing at all about python language (so sorry if my changes are horrible), but anyway I did an small modification to your plugin for my own personal use.

This is because I wanted your plugin to remove certain files that bother me a lot within epub files. They are mostly covers with ads from webpages, or book resumes that had spanish file names and thus they were not deleted with "delete all jackets" option.

I changed file "jacket.py" lines 61 ~70 to the following:

def remove_all_jackets(container, log):
    log('\tLooking for all jackets')
    dirtied = False
    for name in list(container.name_path_map.keys()):
        if 'jacket' in name and name.endswith('.xhtml'):
            data = container.get_parsed_etree(name)
            if is_current_jacket(data) or is_legacy_jacket(data):
                log('\t Jacket removed: ', name)
                dirtied = True
                container.delete_from_manifest(name)
	for name in list(container.name_path_map.keys()):
		if 'sinopsis' in name and name.endswith('.xhtml'):
			data = container.get_parsed_etree(name)
			log('\t Sinopsis jacket found: ', name)
			dirtied = True
			container.delete_from_manifest(name)
	for name in list(container.name_path_map.keys()):
		if 'Cubierta' in name and name.endswith('.xhtml'):
                       data = container.get_parsed_etree(name)
                       log('\t Cubierta jacket found: ', name)
                       dirtied = True
                       container.delete_from_manifest(name)
	for name in list(container.name_path_map.keys()):		
		if 'titulo' in name and name.endswith('.xhtml'):
                       data = container.get_parsed_etree(name)
                       log('\t Titulo jacket found: ', name)
                       dirtied = True
                       container.delete_from_manifest(name)
    return dirtied

This will delete files "sinopsis.xhtml", "cubierta.xhtml" and "titulo.xhtml". It has worked very well for my large collection of downloaded books where most of them had those files.

So, would you be so kind to consider in next version (or if you feel bored enough) to add one option to automatically delete certain files that match a given name?.
This can be useful for many countries where english is not the main language.

Thanks in advance and greetings from Spain.

[davidfor]

Originally Posted by zephyrot

I changed file "jacket.py" lines 61 ~70 to the following:

def remove_all_jackets(container, log):
    log('\tLooking for all jackets')
    dirtied = False
    for name in list(container.name_path_map.keys()):
        if 'jacket' in name and name.endswith('.xhtml'):
            data = container.get_parsed_etree(name)
            if is_current_jacket(data) or is_legacy_jacket(data):
                log('\t Jacket removed: ', name)
                dirtied = True
                container.delete_from_manifest(name)
	if 'sinopsis' in name and name.endswith('.xhtml'):
		log('\t Sinopsis jacket found: ', name)
		dirtied = True
		container.delete_from_manifest(name)
	if 'Cubierta' in name and name.endswith('.xhtml'):
		log('\t Cubierta jacket found: ', name)
		dirtied = True
		container.delete_from_manifest(name)
	if 'titulo' in name and name.endswith('.xhtml'):
		log('\t Titulo jacket found: ', name)
		dirtied = True
		container.delete_from_manifest(name)
    return dirtied

I have made some changes. You didn't need to do a loop for each filename. Plus, the extra loops were within another and you were calling get_parsed_etree for each without using the data. I haven't tested this, and you will need to check the indentation, but it should work.

I don't know about others, but I would feel uncomfortable about including something like this. There is to big a risk of hitting something that shouldn't be removed. Maybe if there was an option to specify a list of names or a regex for the files to be removed. I'm also not exactly sure how much of a need there is.

[Rev. Bob]

Originally Posted by davidfor

I don't know about others, but I would feel uncomfortable about including something like this. There is to big a risk of hitting something that shouldn't be removed. Maybe if there was an option to specify a list of names or a regex for the files to be removed. I'm also not exactly sure how much of a need there is.

I'm not exactly in charge of the plugin, but FWIW, I agree with this assessment. It seems like too much of a gamble for me; too many false negatives and false positives alike.

[Terisa de morgan]

I'm not in charge either but, knowing a bit of Spanish , I agree that it's not sure to make those changes. It's like saying you're going to remove "title.xhtml", "jacket.xhtml" (if it's not standard) and "summary.xhtml" by default.

[zephyrot]

Hi Davidfor, Rev. Bob and Terisa de Morgan.

Thanks for your kind concern, I do see and understand your points so let me give you more background on my particular taste.

As per my knowledge, Calibre don't detect, edit, update or remove jackets whose filename is written in another language rather than english. This means that I have to manually check each and every file with Sigil or Calibre "Edit epub" tool to remove such duplicated jackets. This can be really tiresome as my book collection is rather large, exactly 4179 books right now, all obtained from many different places with many different jacket schemes, font sizes, paragraph sizes........

As I said I understand your view, but my intention is explicitly this:
"Removing certain named files to avoid having duplicate covers and synopsis".
Spanish filenames are used only for certain files that calibre will generate anyway, so I'm not really concerned about potentially deleted files. It is very unlikely and "original_epub" can be used to bring the missing pages back. A list is needed to manually set which files to remove, though.

To be honest, I will probably die from age before finishing reading half of them, given that my little fiddling around with Kiwidude plugin with a limited set of books was successful I'm willing to apply it to all my books.
I just though that this could be useful for someone else but I can't know if there is much request about this feature.

Davidfor: Thanks for your kind update. I had no idea about how to code with Python. Since my first try worked perfectly I left it as is, but I will gladly update with your suggestion.

Oh, at last I regret to say that I forgot to use a real nice word that should be mandatory when requesting things: "Please".

Thank you all and have a nice weekend


Edit: I overlooked a particular point on Davidfor post. Yes, I do think that this needs to have an option to select what filenames to remove if found, we just can't leave this as is. I'm sorry I missed it. I have added the clarification on the post.

[Terisa de morgan]

@zephyrot: Yes, if something like this is implemented, the user should select the name: I would never have a file called "cubierta" but "sobrecubierta", neither "sinopsis" but "resumen", for example.

But the main point is: the removed file isn't "sobrecubierta" in English, but the standard name and extension (that's important: jacket.xhtml works, jacket.html doesn't) that calibre uses for including this jacket. You're not removing something which is a normal part of the book, but something calibre has added.

[JSWolf]

That would be a nice option to have Modidy ePub present us with a list of files in the ePub for us to select which ones we want removed and then they can be removed from the OPF and NCX ToC (if there).

[Rev. Bob]

Originally Posted by zephyrot

As I said I understand your view, but my intention is explicitly this:
"Removing certain named files to avoid having duplicate covers and synopsis".
Spanish filenames are used only for certain files that calibre will generate anyway, so I'm not really concerned about potentially deleted files.

Stupid question time:

If the goal is to avoid duplicates, because Calibre is making duplicate files, why not remove the Calibre-made files? That should be much easier to control.

[BetterRed]

Originally Posted by JSWolf

That would be a nice option to have Modidy ePub present us with a list of files in the ePub for us to select which ones we want removed and then they can be removed from the OPF and NCX ToC (if there).

I think you would need a page preview too, I just looked at an epub from PG, the file names give no clue as to what's in the file.

Also just because a file is named something doesn't guarantee that's what it is, I'm sure I've seen epubs where the Title.xhtml file contained the half title, title and colophon.

- I have a hacked version of Modify that puts the calibre jacket at the back of the book.

It would be nice if Modify had an option to put the jacket.xhtml after the titlepage.xhtml (as it does now) or as the last file in the book. I won't offer my hack - because it removes functionality that others undoubtedly use.

BR

[JSWolf]

I have no need for a jacket. So I remove them if they exist. So to me, it matters not where the jacket is placed. I'll remove it.

Another thing I remove the the page of reviews we sometimes get. I already have the book, I don't need to read review snippets.

[BetterRed]

Jon, as you are with jackets so am I with covers, the reason I first used Modify was to remove them. I like to have a jacket with all the metadata in the book, because the readers I have on my Tab are limited to displaying the DC elements only.