calibre 0.6 betas now available - Page 18

Sanderfox · 06-19-2009, 07:01 PM

Way to go Kovid! Calibre is getting better and better

One question though. I read that image extraction when converting from pdf files is removed for now in 0.6. Is that true? Can you explain why? Most pdf files I would want to convert have some images in them, so this wouldn't be very helpful :P
Anyway... good job so far and keep the improvements coming!

kovidgoyal · 06-19-2009, 07:14 PM

No it's not. Image extraction in 0.6 is the same as in 0.5

Stingo · 06-19-2009, 10:42 PM

Ph.D.!!! Congratulations Kovid. Its great to see good things happen to good people.

user_none · 06-20-2009, 08:05 AM

Quote:

Originally Posted by Sanderfox

Way to go Kovid! Calibre is getting better and better

One question though. I read that image extraction when converting from pdf files is removed for now in 0.6. Is that true? Can you explain why? Most pdf files I would want to convert have some images in them, so this wouldn't be very helpful :P
Anyway... good job so far and keep the improvements coming!

I removed image extraction for PDF input in 0.6. I did this because image extraction supported by pdftohtml (used for the conversion process) is very limited. Image extraction failed for the majority of my test books. Instead of fielding bug reports about issues with it, I removed it completely. Now if it is something you really want and something that you're okay with not working all that well, it's a 4 character change to have it supported again.

sirbruce · 06-20-2009, 08:49 AM

I'd rather have image extraction work on some PDF books rather than none. I haven't noticed Kovid being reluctant to close bug reports he doesn't think will be fixed.

user_none · 06-20-2009, 08:51 AM

Quote:

Originally Posted by sirbruce

I'd rather have image extraction work on some PDF books rather than none. I haven't noticed Kovid being reluctant to close bug reports he doesn't think will be fixed.

The main issue is the bugs are in pdftohtml which makes them very hard to fix. I'll add it back and I'll add a no-image option for cases where it produces horrible results.

sirbruce · 06-20-2009, 10:35 AM

Quote:

Originally Posted by user_none

The main issue is the bugs are in pdftohtml which makes them very hard to fix. I'll add it back and I'll add a no-image option for cases where it produces horrible results.

It seems like Calibre has gone through several different PDF packages trying to find one that works well or is reliably maintained.

pars_andy · 06-20-2009, 10:43 AM

Yeah i have to say i tend to use mobipocket reader to convert to prc and then use calibre from that point on. It's no big deal and it gives me better results.

user_none · 06-20-2009, 10:55 AM

Quote:

Originally Posted by sirbruce

It seems like Calibre has gone through several different PDF packages trying to find one that works well or is reliably maintained.

Yeah... PDF really is a pain. Though, pdftohtml has and is still used for conversion from a PDF file to HTML. The various libraries are used for things like metadata.

If memory serves correctly, pdftothml is used for conversion of a pdf file to html. PyPDF is used for pdfmanipulate. Podofo is used for the metadata and cover extraction. Qt's PDF printing support is used for PDF output. I think that's it, but we use currently 3 different PDF packages.

kovidgoyal · 06-20-2009, 12:27 PM

Actually, one of my medium term projects is to look into using pdftoxml (the open source package that Mobipocket reader uses to replace pdftohtml.

pars_andy · 06-20-2009, 12:48 PM

Quote:

Originally Posted by kovidgoyal

Actually, one of my medium term projects is to look into using pdftoxml (the open source package that Mobipocket reader uses to replace pdftohtml.

Nice one! It doesn't get everything right but it does a very decent job. I notice this thread has slowed down considerably since the release of the last beta. It might just be the weekend effect but hopefully it's a sign that the bugs are on their last legs.

sirbruce · 06-20-2009, 01:48 PM

Quote:

Originally Posted by pars_andy

I notice this thread has slowed down considerably since the release of the last beta. It might just be the weekend effect but hopefully it's a sign that the bugs are on their last legs.

I have a few more books I can still push through Calibre 0,6, and I can try some conversions to other formats besided MOBI, but right now I'm limited by the fact that I keep getting crashes whenever I bulk convert more than 60 or 70 books. I don't think I'm running out of memory here (unlike the PDF issues) and this did not occur in Calibre 0.5. But I do think most of the conversion bugs have been squashed.

http://calibre.kovidgoyal.net/ticket/2598

kovidgoyal · 06-20-2009, 02:09 PM

Quote:

Originally Posted by sirbruce

I have a few more books I can still push through Calibre 0,6, and I can try some conversions to other formats besided MOBI, but right now I'm limited by the fact that I keep getting crashes whenever I bulk convert more than 60 or 70 books. I don't think I'm running out of memory here (unlike the PDF issues) and this did not occur in Calibre 0.5. But I do think most of the conversion bugs have been squashed.

http://calibre.kovidgoyal.net/ticket/2598

Those crashes likely have something to do with the fact that calibre now uses a new process for each conversion (0.5 reused processes) and the locking on windows is getting overwhelmed by each new process trying to lock access to the config files. I'll have to look into finding some other way to lock access to the config files in windows.

derrell · 06-20-2009, 07:52 PM

Has anyone gotten conversion to FictionBook2 to work. Don't know if this is a bug in the beta or I just don't know what I'm doing.

Using the windows ver. of 0.6.0b8

Tried different formats mobi, lit, html and ereader. Either get an error like the one below or the conversion finishes and when it displays in the cr3 all of the tags are visible.

Error message

Spoiler:

user_none · 06-20-2009, 10:43 PM

Quote:

Originally Posted by derrell

Has anyone gotten conversion to FictionBook2 to work. Don't know if this is a bug in the beta or I just don't know what I'm doing.

Bug. I've pushed up a fix.

06-20-2009, 07:52 PM	#269
derrell Jack O' Apes Posts: 227 Karma: 1939 Join Date: Dec 2007 Location: Oklahoma Device: Ebookwise 1150, Nokia N810, EZ-Reader, HTC Droid Incredible, Archos 70	Conversion to fb2 Has anyone gotten conversion to FictionBook2 to work. Don't know if this is a bug in the beta or I just don't know what I'm doing. Using the windows ver. of 0.6.0b8 Tried different formats mobi, lit, html and ereader. Either get an error like the one below or the conversion finishes and when it displays in the cr3 all of the tags are visible. Error message Spoiler: Convert book 1 of 1 (u'A Boy and His Tank') InputFormatPlugin: LIT Input running on C:\Documents and Settings\derrell\My Documents\My eBooks\Leo Frankowski\A Boy and His Tank (14)\A Boy and His Tank - Leo Frankowski.lit Parsing all content... Parsing 0671578502_top.htm ... Parsing 0671578502__p_.htm ... Reading TOC from HTML... Choosing other.ms-coverimage-standard:0671578502_Cover.jpg as the cover Merging user specified metadata... Detecting structure... Detected chapter: I'd like to dedicate this book to Owen Lock, who w Detected chapter: CHAPTER ONE Detected chapter: CHAPTER TWO Detected chapter: CHAPTER THREE Detected chapter: CHAPTER FOUR Detected chapter: CHAPTER FIVE Detected chapter: CHAPTER SIX Detected chapter: CHAPTER SEVEN Detected chapter: CHAPTER EIGHT Detected chapter: CHAPTER NINE Detected chapter: CHAPTER TEN Detected chapter: CHAPTER ELEVEN Detected chapter: CHAPTER TWELVE Detected chapter: CHAPTER THIRTEEN Detected chapter: CHAPTER FOURTEEN Detected chapter: CHAPTER FIFTEEN Detected chapter: CHAPTER SIXTEEN Detected chapter: CHAPTER SEVENTEEN Detected chapter: CHAPTER EIGHTEEN Detected chapter: CHAPTER NINETEEN Detected chapter: CHAPTER TWENTY Detected chapter: CHAPTER TWENTY-ONE Detected chapter: CHAPTER TWENTY-TWO Detected chapter: CHAPTER TWENTY-THREE Detected chapter: CHAPTER TWENTY-FOUR Detected chapter: CHAPTER TWENTY-FIVE Detected chapter: CHAPTER TWENTY-SIX Detected chapter: CHAPTER TWENTY-SEVEN Flattening CSS and remapping font sizes... Source base font size is 12.00000pt Cleaning up manifest... Trimming unused files from manifest... Parsing stylesheet.css ... Trimming '0671578502_PCLibrary.jpg' from manifest Trimming '0671578502_top.htm' from manifest Trimming '0671578502_Title.jpg' from manifest Trimming '0671578502_Library.jpg' from manifest Trimming '0671578502_PCCover.jpg' from manifest Creating FB2 Output... Converting XHTML to FB2 markup... Traceback (most recent call last): File "worker.py", line 103, in <module> File "worker.py", line 90, in main File "calibre\gui2\convert\gui_conversion.pyo", line 17, in gui_convert File "calibre\ebooks\conversion\plumber.pyo", line 685, in run File "calibre\ebooks\fb2\output.pyo", line 20, in convert File "calibre\ebooks\fb2\fb2ml.pyo", line 43, in extract_content File "calibre\ebooks\fb2\fb2ml.pyo", line 59, in fb2mlize_spine File "calibre\ebooks\fb2\fb2ml.pyo", line 86, in clean_text File "re.pyo", line 142, in search File "re.pyo", line 243, in _compile sre_constants.error: unbalanced parenthesis

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Calibre metadata.calibre not allowing updates	Chuckels550	Calibre	10	08-09-2010 05:12 PM
Using Calibre as a client for another Calibre instance?	toddos	Calibre	27	06-30-2010 04:57 AM
Sigil 0.2.0 betas available	Valloric	Sigil	98	05-03-2010 04:07 PM
cannot open calibre on osx 10.6-- "Calibre is already running" message	jlip	Calibre	4	01-02-2010 11:05 PM
calibre command line utilities and calibre defaults	astrodad	Calibre	2	08-07-2008 03:27 PM

06-19-2009, 07:01 PM	#256
Sanderfox Member Posts: 11 Karma: 10 Join Date: May 2009 Device: bebook	Way to go Kovid! Calibre is getting better and better One question though. I read that image extraction when converting from pdf files is removed for now in 0.6. Is that true? Can you explain why? Most pdf files I would want to convert have some images in them, so this wouldn't be very helpful :P Anyway... good job so far and keep the improvements coming!

06-19-2009, 07:14 PM	#257
kovidgoyal creator of calibre Posts: 45,403 Karma: 27756918 Join Date: Oct 2006 Location: Mumbai, India Device: Various	No it's not. Image extraction in 0.6 is the same as in 0.5

06-19-2009, 10:42 PM	#258
Stingo Fanatic Posts: 582 Karma: 1334691 Join Date: Nov 2006 Location: Miami Device: KH2O, KPW2, KDXG, KPW1, K3, S505	Ph.D.!!! Congratulations Kovid. Its great to see good things happen to good people.

06-20-2009, 08:49 AM	#260
sirbruce Provocateur Posts: 1,859 Karma: 505847 Join Date: Feb 2009 Location: Columbus, OH Device: Kindle Touch, Kindle 2, Kindle DX, iPhone 3GS	I'd rather have image extraction work on some PDF books rather than none. I haven't noticed Kovid being reluctant to close bug reports he doesn't think will be fixed.

06-20-2009, 10:43 AM	#263
pars_andy Connoisseur Posts: 65 Karma: 10 Join Date: Apr 2009 Device: Sony PRS505	Yeah i have to say i tend to use mobipocket reader to convert to prc and then use calibre from that point on. It's no big deal and it gives me better results.

06-20-2009, 12:27 PM	#265
kovidgoyal creator of calibre Posts: 45,403 Karma: 27756918 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Actually, one of my medium term projects is to look into using pdftoxml (the open source package that Mobipocket reader uses to replace pdftohtml.

Advert

Advert