[GUI Plugin] Wiki Reader

[chaot]

Originally Posted by anonymust

Has this been updated, it doesnt seem to work for me.

What? Where? How?

May be here you find an answer!?

[eschwartz]

@anonymust,

"doesn't seem to work" is rather vague. If you could provide details about how it is failing, that might help someone fix this.

Are you getting error messages? Are you getting mangled output? Are you getting NO output?

Anything odd showing up in the job logs?

[compurandom]

This plugin seems to work very well and replaces the missing epub generator (and broken pdf generator) that wikipedia use to have.

Now if it could just accept the url of a book made with book creator instead of manually entering a bunch of urls... Even if all it did was ignore all the extra metadata and just grab the list of articles listed in a book, I'd be happy.

https://en.wikipedia.org/wiki/Help:Books

[kovidgoyal]

Sure, that is pretty easy to do.

[compurandom]

That was quick! Cool!

However, it seems to read the first section and then give up.
For example, https://en.wikipedia.org/wiki/Book:3D_printing
only gets the first two articles, quitting before the second heading. Comparing it with the next book, my guess is that it got tripped up on a blank line between the sections and stopped there.

Trying a second book
https://en.wikipedia.org/wiki/Book:2...rricane_season
It looks like it got the whole thing except that it missed the last article (Otto) , replacing it with a second copy of the first article.


Books I tried that are in user space don't work at all. Those would have a URL format like
https ://en.wikipedia.org/wiki/User:USERNAME/Books/BOOKNAME
Looking at the debug output and comparing with one that actually generates an ebook, it gets the correct list of articles, and then dies with this:

Traceback (most recent call last):
File "site.py", line 63, in main
File "site-packages/calibre/utils/ipc/worker.py", line 193, in main
File "site-packages/calibre/gui2/convert/gui_conversion.py", line 25, in gui_convert
File "site-packages/calibre/ebooks/conversion/plumber.py", line 1042, in run
File "site-packages/calibre/customize/conversion.py", line 241, in __call__
File "site-packages/calibre/ebooks/conversion/plugins/recipe_input.py", line 117, in convert
File "site-packages/calibre/web/feeds/news.py", line 1029, in download
File "site-packages/calibre/web/feeds/news.py", line 1202, in build_index
ValueError: No articles found, aborting

For example, I tested on
https://en.wikipedia.org/wiki/User:B.../_Black_Friday
which is a random book from https://en.wikipedia.org/wiki/Catego...8user_books%29

This article (not in user space) fails in the same pattern:
https://en.wikipedia.org/wiki/Book:Black_Hole_Physics

[kovidgoyal]

Sigh, wikipedia and its inconsistent markup. Should be fine now. Note that if a book contains separate "articles" that point to different locations on the same URL, they will be duplicated in the download.

[compurandom]

Originally Posted by kovidgoyal

Note that if a book contains separate "articles" that point to different locations on the same URL, they will be duplicated in the download.

I think that's reasonable -- it would be a bug in the book itself on the wikipedia side, not a bug in this tool.

This is great, it restores functionality I've missed badly since wikipedia removed it.

[Mixx]

This new feature is incredibly useful, thank you!

Cheers, Mixx

[mgrmgrde]

Hi,

I tried the plugin today with URL https://en.wikipedia.org/wiki/List_o...er-view_events

But as a result I only got a 2 page book with a title "ARTICLES"

???

[kovidgoyal]

That is not a wikibook page. You need to use a wikibook page. For ordinary wikipedia articles, you paste the url into the main dialog and click OK

[ekblom]

Hi kovidgoyal,

It would be awesome if you could add this to wikipedia.recipe to get rid of those anying edit-links:

dict(name='span', attrs={'class': 'mw-editsection'})

Actually tested it myself and it worked fine.

Also, sometimes the enginge split the document right after a headline, resulting in an almost blank page, dont know if its easy to fix, but couldnt hurt to notify you. Attached an example. This was generated from https://sv.wikipedia.org/wiki/Allmän...tivitetsteorin

Regards,
Ekblom

[kovidgoyal]

Sure. As for page breaks, the conversion engine defaults to adding them before h1 and h2 tags. You can turn that off by adding

'page_breaks_before':'/'

to conversion_options in the recipe.

[Popaul31]

Hello Krittika Goyal,

This application is a very good idea .
ebooks created with the application open very well in class but I have a blank page when I open my reading light ( PocketBook 626) .
could you help me please? Thanks in advance.

Is it possible to create the ebook without the links in the Wikipedia article ?

sorry for my bad English language.

[tiandijie]

looks like this terrific tool has not been updated for quite a long time......but still I want to know if there is any possibility for further tweak. The current version did everything well expcept the low-sesolution image quality. I love reading history wikis and I had to download the original image by myself to replace the grabbed image in html files

[mcdummy]

Hi,

Wiki articles use incomplete references to other Wiki articles, which are copied as they are by the plugin.

For instance, references within wiki articles do not include "https:" and start with href="//..." or leave out the "https://en.wikipedia.org" and start with href="/wiki/" or href="/W/.

Since most readers are able to access the internet, it would be great if these references were correct to, e.g.,

href="https://

or

href="http://en.wikipedia.org

during the generation of the ebook.

In addition, in case an article makes reference to another article (or a section thereof) of the resulting ebook, it would be great if the original reference to the web page were replaced by the reference to the corresponding article in the ebook.

Regards,

McDummy