calibre 0.6.0 released - Page 6

kovidgoyal · 07-28-2009, 03:03 PM

Only LRF supports font embedding currently and you can only do that via the conversion options, not via CSS.

If you want to mess with CSS features like line-height, I suggest using the EPUB format.

leebok21 · 07-28-2009, 04:10 PM

So, there is no way to change the line height in LRF conversion, right?

For PDF, I can make it with ADOBE, so it doesn't hurt, anyway.

Thanks again,

Quote:

Originally Posted by kovidgoyal

Only LRF supports font embedding currently and you can only do that via the conversion options, not via CSS.

If you want to mess with CSS features like line-height, I suggest using the EPUB format.

GonzoGonzalez · 07-28-2009, 05:27 PM

Great program, but with a fatal bug:

I have version 0.6.2 on Vista-64. When I try to import 136 books (most in PDF format) from hard drive, system slows down to a crawl (I have Quad-Core, 8 gig of ram, 20 Gig free space on HD), runs out of memory in about 5 minutes and becomes unresponsive. SEEMS LIKE SOME MEMORY IS NOT BEING PROPERLY RELEASED.

When computer stopped responding completely (about 10 minutes, 30% of books imported), I managed to terminate it. Then I checked hard-drive- 20 Gig of space was GONE! Just like that. The Calibre book directory was only 20 MB, so I went to check windows temp directory, and, voila- it was overfilled with thousands of files named "magick-something", each one about 4 meg long. So, also- temporary files are not properly released.

What is going on?

kovidgoyal · 07-28-2009, 05:32 PM

You have discovered the joys of PDF. Go to preferences->plugins and disable the PDF metadata reader. And as a general note, when you terminate a program, temporary files will not be cleaned up.

johnwhelan · 07-28-2009, 06:35 PM

I note that I can now add in paper books to the database.

OK I have a few thousand so how can I automate the process?

I assume there should be a way to input the ISBN into something which grabs the details such as author title etc? and outputs something Calibre can accept as input?

Or even a bar code reader which plugs into the USB port etc etc.

Thoughts?

Thanks John

kovidgoyal · 07-28-2009, 10:53 PM

Adding an option that accepts ISBN and downloads metadata is on the TODO list. For the moment, you have to use the edit metadata dialog to manually input ISBNs and download metadata.

GonzoGonzalez · 07-29-2009, 09:42 AM

when you terminate a program, temporary files will not be cleaned up

The problem is- the "magick-" temporary files seem to be NEVER removed, whether the program terminates abnormally or normally. Also, the files should really be removed and memory released AFTER EACH BOOK. Do you think it is normal that the system uses up 20GB of space to just import about 40 books?

Are you using regular expressions to read meta-data from PDF's? If yes- that's the worst way to go. Searching within huge binary files with reg ex is HORRIBLY slow and resource consuming. On my reasonably fast quad core with 8 GB ram it takes sometimes 5 (!) minutes per PDF to search for meta-data. And if the data is absent- in all likeliness regular expression algorithm will run our of resources, just like it happens in my case. Just a thought from reg ex specialist

There are other much faster and less resource consuming possibilities to read pdf meta-data. For example, here is a very fast program that you can run from command line to get PDF meta data. Tested and works great- it even detects if PDF is password-protected:

http://www.becyhome.de/download_ger.htm#becypdfmetaedit

Let me know what you think. Thanks for a great program and keep up the good work!

Kostas · 07-29-2009, 10:05 AM

Wellcome Gonzo,

Don't forget that you are criticizing a FREE program...

GonzoGonzalez · 07-29-2009, 10:41 AM

I'm NOT criticizing it. I'm suggesting a way to IMPROVE it. If it's free it doesn't mean we have to ignore some important flaws. It is a great program, but can be made even better!

This program, becypdfmetaedit, retrieves PDF data in fraction of a second, as opposed to current Calibre plugin that sometimes takes a dozen minutes and runs out of resources. So why not use it? If can be executed from command line, it's output can be parsed and used with Calibre.

http://www.becyhome.de/download_ger.htm#becypdfmetaedit

Kostas · 07-29-2009, 10:50 AM

I mean that becypadfmetaedit is a well known program to the members of this forum.
I thinks that Calibre team would have used its features if it could.

kovidgoyal · 07-29-2009, 12:07 PM

calibre has over the months tried three different libraries to read PDF metadata, none of them work well. PDF is just a pain. As for your disk usage issues, that comes from rendering the PDF to try to extract the first page as an image.

slantybard · 07-29-2009, 12:09 PM

Quote:

Originally Posted by GonzoGonzalez

I'm NOT criticizing it. I'm suggesting a way to IMPROVE it. If it's free it doesn't mean we have to ignore some important flaws. It is a great program, but can be made even better!

This program, becypdfmetaedit, retrieves PDF data in fraction of a second, as opposed to current Calibre plugin that sometimes takes a dozen minutes and runs out of resources. So why not use it? If can be executed from command line, it's output can be parsed and used with Calibre.

http://www.becyhome.de/download_ger.htm#becypdfmetaedit

The problem may be that becypdf appears to be win32, not cross platform, while calibre is OS independent.

GonzoGonzalez · 07-29-2009, 05:42 PM

calibre has over the months tried three different libraries to read PDF metadata, none of them work well

I've been using "becypdfmetaedit" for HUNDREDS of files- it works quickly and flawlessly. The only minus is that it is windows only.

itimpi · 07-29-2009, 06:37 PM

One of the big strengths of Calibre is its cross-platform support. Therefore Calibre will not include any components that do not run on the 3 platforms it supports (Mac, Windows and Linux).

I guess the answer is for someone to rewrite the PDF plugin to use the same algorithm's?

user_none · 07-30-2009, 07:15 AM

Quote:

Originally Posted by itimpi

One of the big strengths of Calibre is its cross-platform support. Therefore Calibre will not include any components that do not run on the 3 platforms it supports (Mac, Windows and Linux).

I guess the answer is for someone to rewrite the PDF plugin to use the same algorithm's?

Any code used by calibre must also be GPLv3 compatible.

07-28-2009, 05:27 PM	#78
GonzoGonzalez Member Posts: 13 Karma: 10 Join Date: Jul 2009 Device: prs-505	Bug on Vista 64 Great program, but with a fatal bug: I have version 0.6.2 on Vista-64. When I try to import 136 books (most in PDF format) from hard drive, system slows down to a crawl (I have Quad-Core, 8 gig of ram, 20 Gig free space on HD), runs out of memory in about 5 minutes and becomes unresponsive. SEEMS LIKE SOME MEMORY IS NOT BEING PROPERLY RELEASED. When computer stopped responding completely (about 10 minutes, 30% of books imported), I managed to terminate it. Then I checked hard-drive- 20 Gig of space was GONE! Just like that. The Calibre book directory was only 20 MB, so I went to check windows temp directory, and, voila- it was overfilled with thousands of files named "magick-something", each one about 4 meg long. So, also- temporary files are not properly released. What is going on? Last edited by GonzoGonzalez; 07-28-2009 at 05:29 PM.

07-28-2009, 06:35 PM	#80
johnwhelan Zealot Posts: 116 Karma: 475752 Join Date: Jan 2009 Device: Kobo Touch, Cybook Gen 3, Google Nexus7, LG V20, Kobo Clara HD	adding paper titles I note that I can now add in paper books to the database. OK I have a few thousand so how can I automate the process? I assume there should be a way to input the ISBN into something which grabs the details such as author title etc? and outputs something Calibre can accept as input? Or even a bar code reader which plugs into the USB port etc etc. Thoughts? Thanks John

07-29-2009, 10:41 AM	#84
GonzoGonzalez Member Posts: 13 Karma: 10 Join Date: Jul 2009 Device: prs-505	I'm NOT criticizing it. I'm suggesting a way to IMPROVE it. If it's free it doesn't mean we have to ignore some important flaws. It is a great program, but can be made even better! This program, becypdfmetaedit, retrieves PDF data in fraction of a second, as opposed to current Calibre plugin that sometimes takes a dozen minutes and runs out of resources. So why not use it? If can be executed from command line, it's output can be parsed and used with Calibre. http://www.becyhome.de/download_ger.htm#becypdfmetaedit Last edited by GonzoGonzalez; 07-29-2009 at 10:43 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
calibre 0.7.0 released!	kovidgoyal	Calibre	194	09-20-2010 02:04 PM
Calibre 0.7.0 released - see site details below	Paulc15	Kobo Reader	16	06-07-2010 12:27 PM
Calibre 0.6.55 released - see site for details	Paulc15	Kobo Reader	1	05-28-2010 08:54 PM
calibre 0.6.0 released!	kovidgoyal	News	86	08-12-2009 04:57 PM
calibre 0.5.0 released!	kovidgoyal	Calibre	99	03-12-2009 07:43 PM

07-28-2009, 03:03 PM	#76
kovidgoyal creator of calibre Posts: 45,359 Karma: 27182818 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Only LRF supports font embedding currently and you can only do that via the conversion options, not via CSS. If you want to mess with CSS features like line-height, I suggest using the EPUB format.

07-28-2009, 05:32 PM	#79
kovidgoyal creator of calibre Posts: 45,359 Karma: 27182818 Join Date: Oct 2006 Location: Mumbai, India Device: Various	You have discovered the joys of PDF. Go to preferences->plugins and disable the PDF metadata reader. And as a general note, when you terminate a program, temporary files will not be cleaned up.

07-28-2009, 10:53 PM	#81
kovidgoyal creator of calibre Posts: 45,359 Karma: 27182818 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Adding an option that accepts ISBN and downloads metadata is on the TODO list. For the moment, you have to use the edit metadata dialog to manually input ISBNs and download metadata.

07-29-2009, 09:42 AM	#82
GonzoGonzalez Member Posts: 13 Karma: 10 Join Date: Jul 2009 Device: prs-505	when you terminate a program, temporary files will not be cleaned up The problem is- the "magick-" temporary files seem to be NEVER removed, whether the program terminates abnormally or normally. Also, the files should really be removed and memory released AFTER EACH BOOK. Do you think it is normal that the system uses up 20GB of space to just import about 40 books? Are you using regular expressions to read meta-data from PDF's? If yes- that's the worst way to go. Searching within huge binary files with reg ex is HORRIBLY slow and resource consuming. On my reasonably fast quad core with 8 GB ram it takes sometimes 5 (!) minutes per PDF to search for meta-data. And if the data is absent- in all likeliness regular expression algorithm will run our of resources, just like it happens in my case. Just a thought from reg ex specialist There are other much faster and less resource consuming possibilities to read pdf meta-data. For example, here is a very fast program that you can run from command line to get PDF meta data. Tested and works great- it even detects if PDF is password-protected: http://www.becyhome.de/download_ger.htm#becypdfmetaedit Let me know what you think. Thanks for a great program and keep up the good work!

07-29-2009, 10:05 AM	#83
Kostas Still wondering why Posts: 253 Karma: 800 Join Date: Jun 2009 Location: Athens, Greece Device: PRS 505, (BlackBerry Bold ?)	Wellcome Gonzo, Don't forget that you are criticizing a FREE program...

07-29-2009, 10:50 AM	#85
Kostas Still wondering why Posts: 253 Karma: 800 Join Date: Jun 2009 Location: Athens, Greece Device: PRS 505, (BlackBerry Bold ?)	I mean that becypadfmetaedit is a well known program to the members of this forum. I thinks that Calibre team would have used its features if it could.

07-29-2009, 12:07 PM	#86
kovidgoyal creator of calibre Posts: 45,359 Karma: 27182818 Join Date: Oct 2006 Location: Mumbai, India Device: Various	calibre has over the months tried three different libraries to read PDF metadata, none of them work well. PDF is just a pain. As for your disk usage issues, that comes from rendering the PDF to try to extract the first page as an image.

07-29-2009, 05:42 PM	#88
GonzoGonzalez Member Posts: 13 Karma: 10 Join Date: Jul 2009 Device: prs-505	calibre has over the months tried three different libraries to read PDF metadata, none of them work well I've been using "becypdfmetaedit" for HUNDREDS of files- it works quickly and flawlessly. The only minus is that it is windows only.

07-29-2009, 06:37 PM	#89
itimpi Wizard Posts: 4,553 Karma: 950151 Join Date: Nov 2008 Device: Sony PRS-950, iphone/ipad (Marvin/iBooks/QuickReader)	One of the big strengths of Calibre is its cross-platform support. Therefore Calibre will not include any components that do not run on the 3 platforms it supports (Mac, Windows and Linux). I guess the answer is for someone to rewrite the PDF plugin to use the same algorithm's?