Calibre stops converting RTF files

[Picander]

Hi,

Last week I got a Kindle and I converted some RTF files into MOBI with Calibre and no problem.

Suddenly Calibre sotpped converting RTF files. I have tried with new files and also with the same files converted previously always getting the same error.

I have reinstalled Calibre 0.7.50 several times. I also have tried installing previous versions, but the problem persists.

My OS is Windows 7 Home Premiun 64 bits service pack 1.

These are the job details. I think that the problem begins in the red marked lines.


Convert book 1 of 1 (Sharpe y el aguila del imperio)
Resolved conversion options
calibre version: 0.7.36
{'asciiize': False,
'author_sort': None,
'authors': None,
'base_font_size': 0.0,
'book_producer': None,
'change_justification': u'original',
'chapter': u"//*[((name()='h1' or name()='h2') and re:test(., 'chapter|book|section|part|prologue|epilogue\\s+', 'i')) or @class = 'chapter']",
'chapter_mark': u'pagebreak',
'comments': None,
'cover': 'c:\\users\\luis\\appdata\\local\\temp\\calibre_0. 7.36_tmp_qrpb32\\calibre_0.7.36_stu_o4.jpeg',
'debug_pipeline': None,
'disable_font_rescaling': False,
'dont_compress': False,
'extra_css': None,
'font_size_mapping': None,
'footer_regex': u'(?i)(?<=<hr>)((\\s*<a name=\\d+></a>((<img.+?>)*<br>\\s*)?\\d+<br>\\s*.*?\\s*)|(\\s* <a name=\\d+></a>((<img.+?>)*<br>\\s*)?.*?<br>\\s*\\d+))(?=<br>)' ,
'header_regex': u'(?i)(?<=<hr>)((\\s*<a name=\\d+></a>((<img.+?>)*<br>\\s*)?\\d+<br>\\s*.*?\\s*)|(\\s* <a name=\\d+></a>((<img.+?>)*<br>\\s*)?.*?<br>\\s*\\d+))(?=<br>)' ,
'html_unwrap_factor': 0.4,
'input_encoding': None,
'input_profile': <calibre.customize.profiles.SonyReaderInput object at 0x05423730>,
'insert_blank_line': False,
'insert_metadata': False,
'isbn': None,
'keep_ligatures': False,
'language': None,
'level1_toc': None,
'level2_toc': None,
'level3_toc': None,
'line_height': 0.0,
'linearize_tables': False,
'margin_bottom': 5.0,
'margin_left': 5.0,
'margin_right': 5.0,
'margin_top': 5.0,
'max_toc_links': 50,
'minimum_line_height': 120.0,
'mobi_ignore_margins': False,
'no_chapters_in_toc': False,
'no_inline_navbars': True,
'no_inline_toc': False,
'output_profile': <calibre.customize.profiles.KindleOutput object at 0x054238D0>,
'page_breaks_before': u"//*[name()='h1' or name()='h2']",
'personal_doc': u'[PDOC]',
'prefer_author_sort': False,
'prefer_metadata_cover': False,
'preprocess_html': False,
'pretty_print': False,
'pubdate': None,
'publisher': None,
'rating': None,
'read_metadata_from_opf': 'c:\\users\\luis\\appdata\\local\\temp\\calibre_0. 7.36_tmp_qrpb32\\calibre_0.7.36_fklm_n.opf',
'remove_first_image': False,
'remove_footer': False,
'remove_header': False,
'remove_paragraph_spacing': True,
'remove_paragraph_spacing_indent_size': 1.0,
'rescale_images': False,
'series': None,
'series_index': None,
'smarten_punctuation': False,
'tags': None,
'timestamp': None,
'title': None,
'title_sort': None,
'toc_filter': None,
'toc_threshold': 6,
'toc_title': None,
'use_auto_toc': False,
'verbose': 2}
InputFormatPlugin: RTF Input running
on C:\JL\Libros\Calibre\Bernard Cornwell\Sharpe y el aguila del imperio (67)\Sharpe y el aguila del imperio - Bernard Cornwell.rtf
Converting RTF to XML...
Preprocessing to convert unicode characters
Python function terminated unexpectedly
[Error 5] Acceso denegado: 'c:\\users\\luis\\appdata\\local\\temp\\tmpvtrhgq' (Error Code: 1)
Traceback (most recent call last):
File "site.py", line 103, in main
File "site.py", line 85, in run_entry_point
File "site-packages\calibre\utils\ipc\worker.py", line 107, in main
File "site-packages\calibre\gui2\convert\gui_conversion.py", line 24, in gui_convert
File "site-packages\calibre\ebooks\conversion\plumber.py", line 853, in run
File "site-packages\calibre\customize\conversion.py", line 216, in __call__
File "site-packages\calibre\ebooks\rtf\input.py", line 254, in convert
File "site-packages\calibre\ebooks\rtf\input.py", line 119, in generate_xml
File "site-packages\calibre\ebooks\rtf2xml\ParseRtf.py", line 331, in parse_rtf
File "site-packages\calibre\ebooks\rtf2xml\colors.py", line 247, in convert_colors
WindowsError: [Error 5] Acceso denegado: 'c:\\users\\luis\\appdata\\local\\temp\\tmpvtrhgq'


I would be very gratefully for your help.

Thanks in advance.

[kovidgoyal]

Fixed in next release

[Picander]

Thank you very much for your quick response.

I wait for the next release. Meanwhile I am achieving very good results converting in Word from DOC to HTML instead RTF, and from HTML to MOBI in Calibre.

Regards.

[Manichean]

A hint when using that workflow: Save as filtered HTML in Word, if you aren't already. Also, it may interest you that there is a plugin available for OpenOffice that exports to ePub.

[Ortep]

Originally Posted by Manichean

Also, it may interest you that there is a plugin available for OpenOffice that exports to ePub.

Why would he want that? He has a Kindle.

[Manichean]

Originally Posted by Ortep

Why would he want that? He has a Kindle.

Yes, but ePub may be a better conversion source than HTML output through Word. I've been on this forum long enough to be vary of that particular brand of HTML code even though I never used it myself (due to not having MS Office, mostly). Of course, that depends on the conversion results and whether or not he is happy with the results from converting via HTML.

[Ortep]

Originally Posted by Manichean

Yes, but ePub may be a better conversion source than HTML output through Word.

Hmmm funny...I try to get rid of ePub when workin on eBooks. I have much more confidence in HTML (Even if it comes from Word)

Calibre can convert almost everything to everything. But it cant convert ePub to a single HTML file. There is a reason for that. You never know what you are going to find inside a ePub. Everybody always says that ePub is HTML in a container...well I never seen one were this was true.
There is a kludge were Calibre generates an zip file that contains something that is remotely related to HTML but most of the time it is a big bowl of spaghetti from dozens of html, xhtml, css and other junk with no apparent structure.

Try to find a ePub to HTML converter and you will find out that it is more easy to find a PDF to HTML converter

Not one ePub is build in the same way, they all look different on the inside and they even look different on different devices because nobody really knows how this 'standard' is implemented.

[DeadMan]

Originally Posted by Ortep

Try to find a ePub to HTML converter and you will find out that it is more easy to find a PDF to HTML converter

Yep, just went through this mess myself. Was going to try one of those HTML joiner programs to see how that works.

[Manichean]

I honestly fail to see the need for a single monolithic HTML file, but your mileage may vary. EPub is, as far as ebook formats go, the most versatile format as far as formatting goes, as well as being the easiest to manage with Calibre, since most of the times, no conversion is needed in order for the metadata to be correct when I export a file. The first reason is why I suggested it here- if you have a base format preserving much, if not all, of the layout, you can expect pretty good conversion results.

Oh, and did I mention there's a WYSIWYG editor available, as well?

[Ortep]

Originally Posted by Manichean

Oh, and did I mention there's a WYSIWYG editor available, as well?

Oh and did I mention that this is as WYSIWYG as VI is a full screen editor?

You'll end up with dozens of different files that you have to edit separately.
In Sigil I still have to edit the CSS file by hand, what is WYSIWYG about that? That is, if there is a CSS

I don't need Sigil for that. Notepad++ is more easy to use. Simply from Calibre use the Tweak option.

This morning I ran into an ePub that existed of 27 HTML/XHTML files. No CSS but all the fortmatting was done in the separate parts. Hurray!!!! Versatility above all

Open File one, correct the 2cm wide margins en 3 cm bottom margins and save it. Open file two, correct the 2.5 cm margins...can leave the bottom margins intact because they are not set here, save file two....

etc etc etc

...save file 27, rebuild ePub. And hope I did not make an error on the way because I won't be able to see it in one go if it is somewere in part 17

In Sigil this would not be any easier, you will have to open 27 tabs and switch between them because all the different parts are treated separately

That is the reason why I want one single file and not 27 or 30 or 4

If Calibre was able to produce one big HTML file I simply could open one file and use a real WYSIWYG editor and make alle changes in one go.

Up till now I use the RTF road. That works, most of the time. It only fails if someone made a mess of the ePub. For example screw up with the size of the different parts and Calibre is not able to generate a correct file. It can miss the last part of the book...or the last part of chapter 13.

ePub is so versatile that anybody is able to produce a big pile of junk in 30 seconds.

[DoctorOhh]

Originally Posted by Ortep

You'll end up with dozens of different files that you have to edit separately.

Sigil allows you to search and replace all html inside the epub at once.

Originally Posted by Ortep

This morning I ran into an ePub that existed of 27 HTML/XHTML files. No CSS but all the fortmatting was done in the separate parts. Hurray!!!! Versatility above all

If you don't have a css file do a epub to epub conversion in calibre and you will have a css file.

Originally Posted by Ortep

Open File one, correct the 2cm wide margins en 3 cm bottom margins and save it. Open file two, correct the 2.5 cm margins...can leave the bottom margins intact because they are not set here, save file two....

Doing a epub to epub conversion with the current version of calibre will also, in most cases, remove the 2cm right and left margins automatically.

Originally Posted by Ortep

...save file 27, rebuild ePub. And hope I did not make an error on the way because I won't be able to see it in one go if it is somewere in part 17

In Sigil this would not be any easier, you will have to open 27 tabs and switch between them because all the different parts are treated separately

Wrong, it would be easier. Once you have a css, Sigil allows you to adjust all pages at once via find and replace in code view. Even without having the css Sigil allows the find and replace of those repetitive edits in one move using code view and find and replace. Much simpler then dealing with 27 separate files.

Originally Posted by Ortep

That is the reason why I want one single file and not 27 or 30 or 4

If Calibre was able to produce one big HTML file I simply could open one file and use a real WYSIWYG editor and make alle changes in one go.

A single html file would have some pluses but for fine tuning a epub Sigil does allow you to treat all html as one.

Originally Posted by Ortep

ePub is so versatile that anybody is able to produce a big pile of junk in 30 seconds.

This is true but correcting the junk is easy in Sigil, since in code view Sigil allows the user to treat all files as on for regex use in find and replace.

Learn more in the Sigil forum.

[Ortep]

Originally Posted by dwanthny

Sigil allows you to search and replace all html inside the epub at once.

So it is not WYSIWYG, you will have to edit HTML code by hand.

If you don't have a css file do a epub to epub conversion in calibre and you will have a css file.

I will try that but I do not want an ePub...I want to stay as far as possible from them. Their 'versatillity' often gives me totaly different results on different readers. (We have three).
I used to use ePub but I ran back to Mobi as fast as I could. It gives me almost te same results on any reader. Fact is: Why would I need this in the first place? ePub is supposed to be standard, not a bowl of spaghetti

Doing a epub to epub conversion with the current version of calibre will also, in most cases, remove the 2cm right and left margins automatically.

Been there, done that, got the t-shirt. It doesn't work in 80% of the cases I get. And the fun part comes when it only works in chapters 1,2,3, 7, 9, 13 an the last one. Then you don't see it at first but it will show up during reading.
The problem is that everybody uses a different approach and they will find a way to make as problematic as possible.

Wrong, it would be easier. Once you have a css, Sigil allows you to adjust all pages at once via find and replace in code view. Even without having the css Sigil allows the find and replace of those repetitive edits in one move using code view and find and replace. Much simpler then dealing with 27 separate files.

There you go: once you have a CSS And then I still have to edit the (x)html code by hand. I find it much more easy to to that with Notepad++. It has macros and a great syntax finder.

Here a part from the user manual of Sigil

http://web.sigil.googlecode.com/hg/searching.html

The Look in combo-box offers various places where Sigil should search. Currently, it can only search in the currently open file or all the HTML files
In the future it will be able to search across all the CSS files and all the files
of any type in the EPUB book.

And only in the Code View, not in the Book View. This limitation exists because of technical restrictions, and will be removed in time.

In Sigil I do not have 27 different files but 27 different tabs. One for each chapter. That is almost the same as opening 27 files in Notepad++ That wil allso give me 27 tabs in one window...And I am still 'coding' in HTML and that is not simply editting a file.

A single html file would have some pluses but for fine tuning a epub Sigil does allow you to treat all html as one.

No it doesn't. I can not see the complete file in one go. Or I must 'combine' alle parts in the first run. And I cannot do something like Select all, Remove formatting. Or select this 'formattig' part and find out if there is more of it. Simply like you can do in Word or OOo

This is true but correcting the junk is easy in Sigil, since in code view Sigil allows the user to treat all files as on for regex use in find and replace.

So still no WYSIWYG, it is almost the same as Notepad++. I agree regex it is extremely powerfull. But you will haver to agree that regex is not the most user friendly way of finding/changing things. I do know how to use it, but it often needs a lot of trial and error for you have the correct one. And try to explain it to my mother in law. She can work with Word, Exel and Powerpoint..but regex?!?!?!?!

Learn more in the Sigil forum.

Learn more at http://notepad-plus-plus.org/

And of course you can use any other programmers editor.

The main argument is: There is no good WYSIWYG edit tool for ePub. Only a lot of small tools to repair them...in a way.
But there are a LOT of HTML editors that are extremely easy to use. In any flavour you wnt. So I'm craving for a good ePub to HTML conversion tool. Also because the end goal for me (and all Kindle users) is not ePub. So why clean it up and make it usable in the first place?
And until that tool is there I will convert to RTF and edit that file with a tool you can find on any Windows computer: Wordpad. Easy and in one go. If you really want ePub? Simply convert it back.

Tip: The really easy and fast way to improve problematic ePub files is to open the ePub with something like Winrar or whatever archiver program you have. Find the css and simply delete it. You will lose all the formatting, but it was bad in the first place.

[DoctorOhh]

Originally Posted by Ortep

So it is not WYSIWYG, you will have to edit HTML code by hand.

Never said it was.

Originally Posted by Ortep

Been there, done that, got the t-shirt. It doesn't work in 80% of the cases I get. And the fun part comes when it only works in chapters 1,2,3, 7, 9, 13 an the last one. Then you don't see it at first but it will show up during reading.

This feature is new this week. I only had one chance to test it and in that one instance it worked. I'm surprised you already have so much experience with it.

Release: 0.7.51 [25 Mar, 2011]

New Features

• Conversion: Detect and remove fake page margins that are specified as a margin on (nearly) every paragraph. This can be turned off via an option under Structure Detection, in case it removes margins that should have been kept.

Originally Posted by Ortep

There you go: once you have a CSS And then I still have to edit the (x)html code by hand. I find it much more easy to to that with Notepad++. It has macros and a great syntax finder.

I too like Notepad++.

Originally Posted by Ortep

In Sigil I do not have 27 different files but 27 different tabs. One for each chapter. That is almost the same as opening 27 files in Notepad++ That wil allso give me 27 tabs in one window...And I am still 'coding' in HTML and that is not simply editting a file.

For doing regex cleanups of existing epub files you only have to open up one file in code view to apply your changes to all of the html files.

But I'm not here in a calibre forum to further defend or clarify Sigil usage.

Originally Posted by Ortep

The main argument is: There is no good WYSIWYG edit tool for ePub.

If that's your main argument then it isn't a argument you have with me.

I don't create ePubs, only clean them up and Sigil is a quick way to make the changes I need.

Originally Posted by Ortep

So I'm craving for a good ePub to HTML conversion tool. Also because the end goal for me (and all Kindle users) is not ePub. So why clean it up and make it usable in the first place?

It looks like our goals are different since I have a Sony.

Originally Posted by Ortep

Tip: The really easy and fast way to improve problematic ePub files is to open the ePub with something like Winrar or whatever archiver program you have. Find the css and simply delete it. You will lose all the formatting, but it was bad in the first place.

If the ePub is that problematic you should throw it out. It's rare I ever have to open a ePub to clean it up, but when I do a few quick edits and I'm out of there. I usually don't have to mess with the CSS file.

Tweak ePub built into calibre will expand the ePub, when I do mess with the css file (previously for margins) I open it up with this method and do quick edits using Notepad++.

[Manichean]

Originally Posted by Ortep

So it is not WYSIWYG, you will have to edit HTML code by hand.

Sure, there's a code view that allows you to edit the XHTML by hand. However, there's also a book view that allows for, dare I say it, WYSIWYG editing of the book. You don't have to do anything.

Also, there's no real need for you to get that worked up about this issue of using ePub and/or Sigil as a conversion step. I only stated what I've found to work best when converting, neither you nor the OP have to use that method.

[ldolse]

If you have a crappy epub and you want to get it to as clean a version of single file html as possible you could convert to mobi, and then convert back to epub.

When you convert back to epub, you have two options to make it a single HTML file:
- enable debug output, and grab the html from the 'input' directory
- change your epub output to not split on page breaks and change the max split size to a few megabytes.

Staying in the html world is generally a better idea than going to rtf. You could also try converting to Markdown/Textile, but formatting retention is highly dependent on how various bits of the html is structured. It can be a good method to clean up crappy html in a number of cases.