conversion_options not used

[mauropiccolo]

Hi,
sorry for my english.

May be a silly problem, but in my recipe I have
conversion_options = {
'dehyphenate': False,
'delete_blank_paragraphs': False,
'format_scene_breaks': False,
'margin_bottom': 1.0,
'margin_left': 1.0,
'margin_right': 1.0,
'margin_top': 1.0,
'mobi_file_type': 'old',
'unwrap_lines': False,
}

but after
$ ebook-convert django.recipe .mobi --output-profile kindle --test -vv --debug-pipeline debug
I have:

Resolved conversion options
calibre version: 1.5.0
{'asciiize': False,
......
'dehyphenate': True,
'delete_blank_paragraphs': True,
......
'margin_bottom': 5.0,
'margin_left': 5.0,
'margin_right': 5.0,
......

Seems that conversion_options is not used.

FYI I'm trying to convert django documentation and my problem is
that code indentation ( between <pre> tags) is lost, so is not so useful in python ;-)

Regards, Mauro

[kovidgoyal]

Why dont you just pass the conversion options to the command line directly?

[mauropiccolo]

Originally Posted by kovidgoyal

Why dont you just pass the conversion options to the command line directly?

because are many changes and it is easier write options in recipes than in command line.
Is not the purpose for conversion_option ? ;-)

I like to know if I'm wrong or if it is a bug.

By the way, i tried in cli, but
--disable-unwrap-lines --disable-delete-blank-paragraphs --disable-remove-fake-margins
do not work for my purposes,
Still html in /debug/input/ is different from original.
The spaces between <pre> tags are not preserved.

[kovidgoyal]

No conversion_options is not intended to replace command line options. Generally speaking, overriding conversion options in a recipe is a bad idea, it should be avoided in all but extreme cases. And the conversion_options are being used in your case, the resolved conversion options in the debug output are resolved before conversion_options is applied. This is one of the reasons it is bad idea to use conversion_options.

If you find it difficult to edit command lines, create a script and edit that.

None of the options you are using have anything to do with processing of <pre>. In fact, other than fake margins all the other options are heuristics options which are turned off anyway unless you pass --enable-heuristics.

And the html will always be different since it is parsed, converted to XHTML cleaned up and re-written.

IIRC old MOBI does not support <pre> though I may be misremembering, convert to azw3 or EPUB.

[mauropiccolo]

thanks for response.
Sadly, my problem with <pre> is still here and is not related with mobi.

( debug/input is before conversion in mobi and I tryed conversion in epub too)

I will try downloading html and then convert in epub/mobi.

For django there is the option for downloading documentation in epub format, but there are other sites that I like convert in mobi for offline reading with my kindle.

Regards, Mauro