[GUI Plugin] Quality Check

[kiwidude]

Fix -> Check and rename book paths

Note that if the author first name is followed by a period, calibre will use an _ as in Windows you cannot have a folder name ending in a period.

[pghaworth]

Originally Posted by kiwidude

Fix -> Check and rename book paths

Note that if the author first name is followed by a period, calibre will use an _ as in Windows you cannot have a folder name ending in a period.

That's what it is then - middle initials, etc. creating it and I can't fix it. Thank you for all your help.

[gers1978]

I have 9 books that always show up in a "Check no calibre SVG cover" search. I've Replaced Cover using Modify ePub, I've removed cover and re-added using Modify ePub, I've replace using Polish Books. I've examined the code of titlepage.xhtml against the same file for a book that DOESN'T show a missing SVG cover, and the codes are nearly identical - both book have a present and correct <svg> tag.

Any ideas?

UPDATE: I tell a lie, using Calbre's Polish Books on the book in question then causes it to no longer be found under a "Check no calibre SVG cover" search

[MelBr]

When checking epubs for missing images, I'm getting a lot of these errors in the log:

ERROR parsing book: /Users/melbr/Documents/Calibre Library/<book name here>
Traceback (most recent call last):
File "calibre_plugins.quality_check.check_epub", line 663, in evaluate_book
File "site-packages/calibre/utils/zipfile.py", line 751, in __init__
File "site-packages/calibre/utils/zipfile.py", line 786, in _GetContents
File "site-packages/calibre/utils/zipfile.py", line 797, in _RealGetContents
File "site-packages/calibre/utils/zipfile.py", line 263, in _EndRecData
error: unpack requires a string argument of length 22

Any ideas?

[kiwidude]

@MelBr - without seeing one of those ePubs myself there is not much I can suggest. It looks like the ePub is corrupted? What happens when you use Tweak ePub to explode the ePub, does that succeed? This is calibre code that is expanding the contents of the book not my plugin code, perhaps Kovid may have a suggestion.

[Merischino]

Just a quick note to say thank you for this plugin. I've been on a major calibre library cleanup program and this plugin has really, really speeded things up for me considerably.

[rifka]

Is it possible to use this plugin to search for "bad" conversions. A specific example is where paragraphs a split:

_____________________
Forty-six, forty-seven. No reason to turn back; she’d finish it out for sure now. Have a little rest at the top, find out

how and why the bastard does it, but above all stop him. Oh—very important—try to forget about the increasing

distance to the ground. What was it to the top? Seventy-five or eighty feet? That wasn’t much, no, not when one was

________________________________

If the plugin can't be used do you have any suggestions?

Thanks

[kiwidude]

@rifka - you can specify regular expressions in the Seach ePub feature of this plugin. So for instance something like this would find those cases above:

[A-Za-z,]</p>\s+<p[^>]*>[a-z]

Of course you may get a lot of false positives as well depending on the type of book (things like inlined poetry etc) - you really need to open the book in Sigil and apply similar expressions to step through each case and decide whether it is really a problem or not...

[rifka]

One more question. And I'm putting it here because I think using the Search epub function of Quality Check is the only way to do it. I have a pile of history books and I am looking for a several specific names in them. But I want it to find that exact match. So for example how can I get it to find Jane but not Janet or West but not Western.

Thanks for your help. Hopefully I can do this, it will save me hours.

[rifka]

Originally Posted by kiwidude

@rifka - you can specify regular expressions in the Seach ePub feature of this plugin. So for instance something like this would find those cases above:

[A-Za-z,]</p>\s+<p[^>]*>[a-z]

Of course you may get a lot of false positives as well depending on the type of book (things like inlined poetry etc) - you really need to open the book in Sigil and apply similar expressions to step through each case and decide whether it is really a problem or not...

Thank you for your help.

Kiwidude can I suggest this as an additional check. I can send you a bunch of examples. I think they are mostly the result of bad conversions from PDF. Some of them are not that bad they just have the page numbers and/or author and/or publisher every few pages. Others are completely unreadable. Let me know if you would like the samples.

[manawydan]

I am not sure if I have not noted it before or it changed with Calibre 1.1..
Whenever I have two authors (or more) and not all are with comma the switch LN -> FN does not work.

This happens:
Marking
Douglas Preston & Child, Lincoln
Fun Fix, swap author FN LN ..
Preston, Douglas Preston & Lincoln Child

Doing it again and we are back to
Douglas Preston & Child, Lincoln

Why does it swap where there is no comma?

Any way out here except manually doing it?

[theducks]

Originally Posted by manawydan

I am not sure if I have not noted it before or it changed with Calibre 1.1..
Whenever I have two authors (or more) and not all are with comma the switch LN -> FN does not work.

This happens:
Marking
Douglas Preston & Child, Lincoln
Fun Fix, swap author FN LN ..
Preston, Douglas Preston & Lincoln Child

Doing it again and we are back to
Douglas Preston & Child, Lincoln

Why does it swap where there is no comma?

Any way out here except manually doing it?

The PI is designed to flip the state of the name.
Your naming is severely not standard.

Use the Tag Browser's rename to get the author in a consistent state

[shyguy]

I get an Error everytime i try to use "Fix -> Check and rename book paths"

calibre, version 1.3.0
ERROR: Unhandled exception: <b>AttributeError</b>:'LibraryDatabase' object has no attribute 'construct_path_name'

calibre 1.3 Portable isfrozen: True is64bit: False
Windows-8-6.2.9200 Windows ('32bit', 'WindowsPE')
('Windows', '8', '6.2.9200')
Python 2.7.5
Windows: ('8', '6.2.9200', '', 'Multiprocessor Free')
Traceback (most recent call last):
  File "calibre_plugins.quality_check.dialogs", line 51, in do_book_action
  File "calibre_plugins.quality_check.check_fix", line 187, in evaluate_book
AttributeError: 'LibraryDatabase' object has no attribute 'construct_path_name'

Happens to me since I updatd my Portable Version to >= 1.0

Thanks

[theducks]

Originally Posted by shyguy

I get an Error everytime i try to use "Fix -> Check and rename book paths"

calibre, version 1.3.0
ERROR: Unhandled exception: <b>AttributeError</b>:'LibraryDatabase' object has no attribute 'construct_path_name'

calibre 1.3 Portable isfrozen: True is64bit: False
Windows-8-6.2.9200 Windows ('32bit', 'WindowsPE')
('Windows', '8', '6.2.9200')
Python 2.7.5
Windows: ('8', '6.2.9200', '', 'Multiprocessor Free')
Traceback (most recent call last):
  File "calibre_plugins.quality_check.dialogs", line 51, in do_book_action
  File "calibre_plugins.quality_check.check_fix", line 187, in evaluate_book
AttributeError: 'LibraryDatabase' object has no attribute 'construct_path_name'

Happens to me since I updatd my Portable Version to >= 1.0

Thanks

Not limited to 'Calibre Portable', Not something I use regularly so I had not noticed before.
Same message on XP wit regular Calibre

[manawydan]

Originally Posted by theducks

The PI is designed to flip the state of the name.
Your naming is severely not standard.

Beg your pardon?
Why is it severely not standard?

The above refers to authors, not authors_sort in case this is the reason you are surprised.
Authors_sort seems to be ok BTW. But I'd like to swap it in authors and if there is just one name or two and both with comma there is no problem.

Use the Tag Browser's rename to get the author in a consistent state

I almost never use the tag browser and I would not know what for? Renaming with the tag browser? *confused*