Failed to extract text from gutenberg books

[xxyzz]

I get the following error on some gutenberg books when I call "MobiReader.extract_text()". For example, both Kindle ebooks of Alice's Adventures in Wonderland at https://www.gutenberg.org/ebooks/11 will cause this error.

 mobiReader.extract_text()
  File "calibre/ebooks/mobi/reader/mobi6.py", line 802, in extract_text
  File "calibre/ebooks/mobi/reader/mobi6.py", line 802, in <listcomp>
  File "calibre/ebooks/mobi/reader/mobi6.py", line 797, in text_section
  File "calibre/ebooks/mobi/reader/mobi6.py", line 787, in sizeof_trailing_entries
TypeError: ord() expected a character, but string of length 0 found

KindleUnpack doesn't have this issue, I find the similar code at https://github.com/kevinhendricks/Ki...r.py#L816-L830 but don't know how to fix the bug.

[kovidgoyal]

converting pg11.mobi works fine for me, which uses that same code.

[xxyzz]

I forget to mention I also run the following code before calling "extract_text()"

 with open(book_path, 'r+b') as f:
    mu = MetadataUpdater(f)
    mu.update(mi, asin="BBJH94AM2L")

[kovidgoyal]

Doing

ebook-meta -t XXX pg11.mobi && ebook-convert pg11.mobi .epub

also works fine for me.

[xxyzz]

I not sure which part of the plugin code causes the error but here are the steps to reproduce it:

1. add the mobi book to calibre
2. install WordDumb plugin
3. use this plugin on the book

Only `MetadataUpdater.update(mi, asin=asin)` changes the book file, I also add the "mobi-asin" identifier to the book metadata.

The plugin code:

• update ASIN: https://github.com/xxyzz/WordDumb/bl...ata.py#L33-L44
• use extract_text(): https://github.com/xxyzz/WordDumb/bl...job.py#L56-L63

[xxyzz]

I move the code to a file:

#!/usr/bin/env python3

from calibre.library import db
from calibre.utils.logging import default_log
from calibre.ebooks.mobi.reader.mobi6 import MobiReader

lib_db = db('~/Calibre Library').new_api
alice_id = 0
for book_id in lib_db.all_book_ids():
    mi = lib_db.get_metadata(book_id)
    if mi.get('title') == "Alice's Adventures in Wonderland":
        alice_id = book_id
        break

book_path = lib_db.format_abspath(alice_id, 'MOBI')
mobiReader = MobiReader(book_path, default_log)
mobiReader.extract_text()

Use calibre-debug to run this code should reproduce the error.

[kovidgoyal]

That's a joint mobi file, you cant extract it like that, see plugins/mobi_input.py for how to do it.

[xxyzz]

I wasn't aware of this type of book, I should read the code more carefully. Thanks!