Bookit plugin for Firefox - Page 8

beowulf573 · 06-13-2008, 09:52 AM

Hmm, that's odd, it was working. What platform are you running on?

I've got to take my wife to a Dr's appointment, I'll take a look at it later today.

beowulf573 · 06-13-2008, 09:57 AM

Quote:

Originally Posted by jotheman

And not all thext that's bold on the webpage will end up bold in the LRF.

I meant to add, can you give me an example? I'm guessing, but probably they're setting the bold attribute in a style sheet. Right now I'm just grabbing the selected html. I'll look and see if it's possible to grab style sheet info also.

jotheman · 06-13-2008, 11:30 AM

Hi,

I'm on Mac OS X 10.5.3

About the bold parts not being bold: those were the sites I tested it on: http://www.dw-world.de/dw/article/0,...410157,00.html
and http://www.heise.de/tp/r4/artikel/28/28121/1.html

The second one worked, so you're probably right and the other one's using stylesheets.

But the second one contains some German umlauts in the title and wouldn't work until I removed them, so there is definitely a problem with the charset in titles and/or resulting filenames.

No images were included in both cases (OK, there are no images in the second article, but generally images from that site don't seem to work, at least not with "selection to LRF": http://www.heise.de/tp/r4/artikel/28/28035/1.html . The images are there now when I convert the whole page, BTW, which is new - that's the same page I mentioned in an earlier post!)

jo.

beowulf573 · 06-13-2008, 08:52 PM

There's problem getting the encoding correct on content and filenames between XP, Linux and Mac and between Firefox 2.0 and Firefox 3.0. It's going to take a bit of digging to figure out what works where and implement it correctly.

So for now you'll see issues with texts and titles that have unicode characters such as Chinese.

jotheman · 06-15-2008, 06:40 PM

Hi,

I find myself using BookIt quite often, even though I'm not planning to buy a Sony Reader and don't knwo when Calibre will support conversion to ePub. It is just nice to "collect" stuff this way.

While doing so, I noticed that BookIt has a "comments" and a "category" field for eBooks. Text entered here doesn't seem to be carried over to Calibre. Calibre uses "tags" and "series" and also has a "comments" field. All of them stay empty, no matter what I fill in the BookIt fields...

Supporting the "tags" fild of Calibre would be especially nice!

Cheers,

jo.

Surfergirl · 06-16-2008, 09:08 AM

Beowulf

I have to say that, since the "convert selected" option was added, BookIt has become one of my most used utilities (second only to Calibre

.

Well done, and keep up the good work

Irene

beowulf573 · 06-16-2008, 12:59 PM

Quote:

Originally Posted by jotheman

Hi,
While doing so, I noticed that BookIt has a "comments" and a "category" field for eBooks. Text entered here doesn't seem to be carried over to Calibre. Calibre uses "tags" and "series" and also has a "comments" field. All of them stay empty, no matter what I fill in the BookIt fields...

Supporting the "tags" fild of Calibre would be especially nice!

Thanks for the feedback (everyone). I don't have time for extensive testing, so this helps.

I think there are two type of comments; those embedded in the file as meta-data and those stored in the Calibre database. Right now I can only modify the embedded metadata and then add the file to Calibre, I have no way of modifying the attributes in the database that are external to the file. Perhaps Kovid will see this and explain it better.

Running lrf-meta on a file shows the comment created with Bookit, but that comment doesn't appear in Calibre.

I've not had the time to research the unicode encoding issue, I hope to get to that tonight or tomorrow.

kovidgoyal · 06-16-2008, 02:42 PM

Unfortunately there's no tags field in the LRF metadata to extract into the calibre database. But I could add support for modifying metadata fields to the calibredb command

soilwork · 06-19-2008, 02:25 AM

I think this is related to invalid character problem reported earlier, but I found that sites with ':' in title will not work. For example, using Bookit on the following site will create empty file names 'Ctan'.
http://www.ctan.org/what_is_tex.html
If you can allow Bookit to automatically replace or remove such characters, it will be really helpful.

Thanks for your great extension and it works great in FF3 without any problem.

beowulf573 · 06-19-2008, 09:03 AM

Great, thanks. I'll update the list of invalid characters today and shove out a new release. I'm hoping to spend some time tonight looking at the unicode issue.

kovidgoyal · 06-19-2008, 10:54 AM

you can find the list of invalid characters in the sanitize_filename function in the calibre sources.

beowulf573 · 06-19-2008, 03:05 PM

Cool, I'll check it out. Turns out I had ':' in the list but my regex isn't working correctly.

dsuden · 06-24-2008, 08:44 AM

Quote:

Originally Posted by beowulf573

This would work well for me, since I usually switch to a printable, single page version of the document if it's available and generate the ebook from there. This won't work so well if folks are generally creating ebooks from multipage sites.

So it's a trade off, better control over the content vs spaning multiple pages. It thought about spidering the page from inside the plugin, but that may be more work than I have time to do right now.

Thoughts?

I don't want to talk for others, so I hope everybody will chime in on this, but generally I don't try to spider a site with the plugin...I'm using it for long, single-page articles. I *do* really like when the illustrations in the article appear in the lrf.

beowulf573 · 06-24-2008, 12:31 PM

I've pushed 0.3.8 out, sorry for the delay I'm a bit busy with personal and work stuff right now.

In this release:

I’ve changed the code to use the calibredb command to add lrf files to the database. There’s a new preference that stores the path to calibredb, after updating please verify in the options dialog that the path is correct.

The code to check for a colon in the output file has been fixed.

Two existing bugs that I’ve not yet found a solution for both involving unicode characters. If the title of the page has unicode characters the output file generated by web2lrf under Windows NT may not be correct. For now just edit the title when creating the page to not include exotic characters. Also, under Firefox 2.x when creating a lrf book from the current selection and the selection has unicode characters, the lrf file does not always contain the correct text. It’s not a consistent bug, and I’ve not been able to reproduce it using Firefox 3.x.

beowulf573 · 06-24-2008, 12:40 PM

Quote:

Originally Posted by dsuden

I don't want to talk for others, so I hope everybody will chime in on this, but generally I don't try to spider a site with the plugin...I'm using it for long, single-page articles. I *do* really like when the illustrations in the article appear in the lrf.

Thanks. For a while at least I'll leave the way it works as it is. There's no good solution short of writing my own spidering code. I may still end up doing that, but that's a bigger project than I have spare time for right now.

06-15-2008, 06:40 PM	#110
jotheman Sleeper. Posts: 109 Karma: 10 Join Date: Dec 2007 Device: Boox Max2, Kinde Voyage, reMarkable, Dasung Paperlike Pro	almost empty metadata Hi, I find myself using BookIt quite often, even though I'm not planning to buy a Sony Reader and don't knwo when Calibre will support conversion to ePub. It is just nice to "collect" stuff this way. While doing so, I noticed that BookIt has a "comments" and a "category" field for eBooks. Text entered here doesn't seem to be carried over to Calibre. Calibre uses "tags" and "series" and also has a "comments" field. All of them stay empty, no matter what I fill in the BookIt fields... Supporting the "tags" fild of Calibre would be especially nice! Cheers, jo.

06-16-2008, 09:08 AM	#111
Surfergirl Enthusiast Posts: 49 Karma: 299 Join Date: Oct 2007 Location: South Wales, UK Device: PRS-505 (Blue)/PRS-505 (Red)/iPhone 3GS	Fantastic little tool! Beowulf I have to say that, since the "convert selected" option was added, BookIt has become one of my most used utilities (second only to Calibre . Well done, and keep up the good work Irene Last edited by Surfergirl; 06-16-2008 at 09:09 AM. Reason: Can't spell!!

06-19-2008, 02:25 AM	#114
soilwork useR! Posts: 299 Karma: 651 Join Date: Nov 2007 Location: NY Device: Onyx Boox Max 2, Kobo Libra H2O, iRiver Story HD	Comma ':' in title will create empty file I think this is related to invalid character problem reported earlier, but I found that sites with ':' in title will not work. For example, using Bookit on the following site will create empty file names 'Ctan'. http://www.ctan.org/what_is_tex.html If you can allow Bookit to automatically replace or remove such characters, it will be really helpful. Thanks for your great extension and it works great in FF3 without any problem.

06-24-2008, 12:31 PM	#119
beowulf573 Addict Posts: 208 Karma: 1523 Join Date: Jul 2007 Location: Houston,TX Device: PRS-T1	Released 0.3.8 I've pushed 0.3.8 out, sorry for the delay I'm a bit busy with personal and work stuff right now. In this release: I’ve changed the code to use the calibredb command to add lrf files to the database. There’s a new preference that stores the path to calibredb, after updating please verify in the options dialog that the path is correct. The code to check for a colon in the output file has been fixed. Two existing bugs that I’ve not yet found a solution for both involving unicode characters. If the title of the page has unicode characters the output file generated by web2lrf under Windows NT may not be correct. For now just edit the title when creating the page to not include exotic characters. Also, under Firefox 2.x when creating a lrf book from the current selection and the selection has unicode characters, the lrf file does not always contain the correct text. It’s not a consistent bug, and I’ve not been able to reproduce it using Firefox 3.x.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
New Plugin Type Idea: Library Plugin	cgranade	Plugins	3	09-15-2010 12:11 PM
BookIt and 64 bit	jlbfoot	LRF	0	03-09-2009 03:24 PM
Idea for a "Bookit" Plugin -- Maybe Kovid?	dsuden	Sony Reader	55	01-03-2009 11:22 AM
Great new Idea! Bookit button	=X=	Feedback	0	10-27-2008 01:49 PM
Making MobiRead Threads BookIt Friendly	=X=	Feedback	3	08-11-2008 11:24 PM

06-13-2008, 09:52 AM	#106
beowulf573 Addict Posts: 208 Karma: 1523 Join Date: Jul 2007 Location: Houston,TX Device: PRS-T1	Hmm, that's odd, it was working. What platform are you running on? I've got to take my wife to a Dr's appointment, I'll take a look at it later today.

06-13-2008, 11:30 AM	#108
jotheman Sleeper. Posts: 109 Karma: 10 Join Date: Dec 2007 Device: Boox Max2, Kinde Voyage, reMarkable, Dasung Paperlike Pro	Hi, I'm on Mac OS X 10.5.3 About the bold parts not being bold: those were the sites I tested it on: http://www.dw-world.de/dw/article/0,...410157,00.html and http://www.heise.de/tp/r4/artikel/28/28121/1.html The second one worked, so you're probably right and the other one's using stylesheets. But the second one contains some German umlauts in the title and wouldn't work until I removed them, so there is definitely a problem with the charset in titles and/or resulting filenames. No images were included in both cases (OK, there are no images in the second article, but generally images from that site don't seem to work, at least not with "selection to LRF": http://www.heise.de/tp/r4/artikel/28/28035/1.html . The images are there now when I convert the whole page, BTW, which is new - that's the same page I mentioned in an earlier post!) jo.

06-13-2008, 08:52 PM	#109
beowulf573 Addict Posts: 208 Karma: 1523 Join Date: Jul 2007 Location: Houston,TX Device: PRS-T1	There's problem getting the encoding correct on content and filenames between XP, Linux and Mac and between Firefox 2.0 and Firefox 3.0. It's going to take a bit of digging to figure out what works where and implement it correctly. So for now you'll see issues with texts and titles that have unicode characters such as Chinese.

06-16-2008, 02:42 PM	#113
kovidgoyal creator of calibre Posts: 44,036 Karma: 22669822 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Unfortunately there's no tags field in the LRF metadata to extract into the calibre database. But I could add support for modifying metadata fields to the calibredb command

06-19-2008, 09:03 AM	#115
beowulf573 Addict Posts: 208 Karma: 1523 Join Date: Jul 2007 Location: Houston,TX Device: PRS-T1	Great, thanks. I'll update the list of invalid characters today and shove out a new release. I'm hoping to spend some time tonight looking at the unicode issue.

06-19-2008, 10:54 AM	#116
kovidgoyal creator of calibre Posts: 44,036 Karma: 22669822 Join Date: Oct 2006 Location: Mumbai, India Device: Various	you can find the list of invalid characters in the sanitize_filename function in the calibre sources.

06-19-2008, 03:05 PM	#117
beowulf573 Addict Posts: 208 Karma: 1523 Join Date: Jul 2007 Location: Houston,TX Device: PRS-T1	Cool, I'll check it out. Turns out I had ':' in the list but my regex isn't working correctly.

Advert

Advert