AttributeError: Custom field name '' must begin with '#'

[Peter Blaise]

AttributeError: Custom field name '' must begin with '#'

... after successful editing cleanup and saving an EPUB, then trying to convert it to MOBI, which failed, so re-opening the EPUB in Calibre Ebook 7.10 editor gets the above error, and Caliber Ebook now refuses to open the EPUB for editing.

Calibre Ebook v7.10 apparently can't auto fix it?

Why present an error and not a fix?

I tried other in-computer and on-line converters and 'fixers', zipping and re-opening, all failed.

Anyway, how to fix it, please.

And how to PREVENT it.

Oddly, Calibre Ebook 7.10 itself can open the EPUB for reading on screen, no problem, end to end, the entire book is there exactly as it left the editing process.

Post Script - Amazon converted it to KFX just fine ( via email send EPUB to device ).

So, if Calibre Ebook can't 'fix' an "AttributeError: Custom field name '' must begin with '#'" by inserting a '#' itself, how can I do it outside the Calibre Ebook program to move forward after the Caliber Ebook program munged the EPUB?

Thanks.

.

[Karellen]

Does the error point to where exactly the custom filed name is?
Maybe post a screenshot of the code you are using?

[theducks]

You are trying to embed / use data from a custom column when sending a book
Your template is missing the # (templates us LOOKUP names)

[Peter Blaise]

Thanks for considering this, let's play ...

After 18 successful conversions of other EPUB to MOBI with this Calibre Ebook installation software through versions 7.7, 7.9, and 7.10, from April to now, about a month, ( and I successfully test-converted an additional EPUB to MOBI after this error ) here are some error contents for this particular EPUB:

CONVERT EPUB TO MOBI ERROR:

Convert book 1 of 1 (... epub title ...)
Conversion options changed from defaults:
change_justification: 'left'
asciiize: True
linearize_tables: True
cover: 'C:\\Temp\\calibre_eta7esml\\jx1cst3l.jpeg'
verbose: 2
filter_css: 'margin-bottom,,background,padding-right,padding-bottom,color,background-color,margin-left,font-family,margin-right,margin,padding,padding-left,float,margin-top,padding-top'
unsmarten_punctuation: True
output_profile: 'kindle_voyage'
expand_css: True
read_metadata_from_opf: 'C:\\Temp\\calibre_eta7esml\\r2w04x3z.opf'
Resolved conversion options
calibre version: 7.10.0
{'asciiize': True,
'author_sort': None,
'authors': None,
'base_font_size': 0.0,
'book_producer': None,
'change_justification': 'left',
'chapter': "//*[((name()='h1' or name()='h2') and re:test(., "
"'\\s*((chapter|book|section|part)\\s+)|((prolog|p rologue|epilogue)(\\s+|$))', "
"'i')) or @class = 'chapter']",
'chapter_mark': 'pagebreak',
'comments': None,
'cover': 'C:\\Temp\\calibre_eta7esml\\jx1cst3l.jpeg',
'debug_pipeline': None,
'dehyphenate': True,
'delete_blank_paragraphs': True,
'disable_font_rescaling': False,
'dont_split_on_page_breaks': False,
'duplicate_links_in_toc': False,
'embed_all_fonts': False,
'embed_font_family': None,
'enable_heuristics': False,
'epub_flatten': False,
'epub_inline_toc': False,
'epub_max_image_size': 'none',
'epub_toc_at_end': False,
'epub_version': '2',
'expand_css': True,
'extra_css': None,
'extract_to': None,
'filter_css': 'margin-bottom,,background,padding-right,padding-bottom,color,background-color,margin-left,font-family,margin-right,margin,padding,padding-left,float,margin-top,padding-top',
'fix_indents': True,
'flow_size': 260,
'font_size_mapping': None,
'format_scene_breaks': True,
'html_unwrap_factor': 0.4,
'input_encoding': None,
'input_profile': <calibre.customize.profiles.InputProfile object at 0x000002079F4DF9D0>,
'insert_blank_line': False,
'insert_blank_line_size': 0.5,
'insert_metadata': False,
'isbn': None,
'italicize_common_cases': True,
'keep_ligatures': False,
'language': None,
'level1_toc': None,
'level2_toc': None,
'level3_toc': None,
'line_height': 0.0,
'linearize_tables': True,
'margin_bottom': 5.0,
'margin_left': 5.0,
'margin_right': 5.0,
'margin_top': 5.0,
'markup_chapter_headings': True,
'max_toc_links': 50,
'minimum_line_height': 120.0,
'no_chapters_in_toc': False,
'no_default_epub_cover': False,
'no_inline_navbars': False,
'no_svg_cover': False,
'output_profile': <calibre.customize.profiles.KindleVoyageOutput object at 0x000002079F4F3450>,
'page_breaks_before': '/',
'prefer_metadata_cover': False,
'preserve_cover_aspect_ratio': False,
'pretty_print': True,
'pubdate': None,
'publisher': None,
'rating': None,
'read_metadata_from_opf': 'C:\\Temp\\calibre_eta7esml\\r2w04x3z.opf',
'remove_fake_margins': True,
'remove_first_image': False,
'remove_paragraph_spacing': False,
'remove_paragraph_spacing_indent_size': 1.5,
'renumber_headings': True,
'replace_scene_breaks': '',
'search_replace': '[]',
'series': None,
'series_index': None,
'smarten_punctuation': False,
'sr1_replace': None,
'sr1_search': None,
'sr2_replace': None,
'sr2_search': None,
'sr3_replace': None,
'sr3_search': None,
'start_reading_at': None,
'subset_embedded_fonts': False,
'tags': None,
'timestamp': None,
'title': None,
'title_sort': None,
'toc_filter': None,
'toc_threshold': 6,
'toc_title': None,
'transform_css_rules': '[]',
'transform_html_rules': '[]',
'unsmarten_punctuation': True,
'unwrap_lines': True,
'use_auto_toc': False,
'verbose': 2}
InputFormatPlugin: EPUB Input running
on C:\Temp\calibre_eta7esml\wtir915l.epub
Found HTML cover cov.xhtml
Parsing all content...
Traceback (most recent call last):
File "runpy.py", line 198, in _run_module_as_main
File "runpy.py", line 88, in _run_code
File "site.py", line 83, in <module>
File "site.py", line 78, in main
File "site.py", line 50, in run_entry_point
File "calibre\utils\ipc\worker.py", line 215, in main
File "calibre\gui2\convert\gui_conversion.py", line 38, in gui_convert_override
File "calibre\gui2\convert\gui_conversion.py", line 25, in gui_convert
File "calibre\ebooks\conversion\plumber.py", line 1132, in run
File "calibre\ebooks\conversion\plumber.py", line 1336, in create_oebbook
File "calibre\ebooks\oeb\reader.py", line 97, in __call__
File "calibre\ebooks\oeb\reader.py", line 722, in _all_from_opf
File "calibre\ebooks\oeb\reader.py", line 164, in _metadata_from_opf
File "calibre\ebooks\metadata\opf2.py", line 694, in to_book_metadata
File "calibre\ebooks\metadata\opf3.py", line 1067, in read_metadata
File "calibre\ebooks\metadata\book\base.py", line 444, in set_user_metadata
AttributeError: Custom field name '' must begin with '#'

====================================

So let's try re-editing it:


ERROR OPENING EPUB FOR EDITING:

calibre, version 7.10.0
ERROR: Unhandled exception: <b>AttributeError</b>:Custom field name '' must begin with '#'

calibre 7.10 embedded-python: True
Windows-10-10.0.22621-SP0 Windows ('64bit', 'WindowsPE')
('Windows', '10', '10.0.22621')
Python 3.11.5
Windows: ('10', '10.0.22621', 'SP0', 'Multiprocessor Free')
Interface language: None
EXE path: C:\Program Files\Calibre7\calibre-parallel.exe
Traceback (most recent call last):
File "calibre\gui2\__init__.py", line 801, in dispatch
File "calibre\gui2\tweak_book\job.py", line 84, in job_done
File "calibre\gui2\tweak_book\boss.py", line 398, in book_opened
File "calibre\ebooks\oeb\polish\container.py", line 655, in mi
File "calibre\ebooks\metadata\opf2.py", line 694, in to_book_metadata
File "calibre\ebooks\metadata\opf3.py", line 1067, in read_metadata
File "calibre\ebooks\metadata\book\base.py", line 444, in set_user_metadata
AttributeError: Custom field name '' must begin with '#'

==========================================

Let's try embedding metadata:

calibre, version 7.10.0
WARNING: Failed for some files: Failed to embed metadata into some book files. Click "Show details" for details.

The EPUB format of (... epub title ...):

Traceback (most recent call last):
File "calibre\customize\ui.py", line 500, in set_file_type_metadata
File "calibre\customize\builtins.py", line 449, in set_metadata
File "calibre\ebooks\metadata\epub.py", line 259, in set_metadata
File "calibre\ebooks\metadata\opf.py", line 108, in set_metadata
File "calibre\ebooks\metadata\opf.py", line 95, in set_metadata_opf3
File "calibre\ebooks\metadata\opf3.py", line 1080, in apply_metadata
File "calibre\ebooks\metadata\opf3.py", line 1067, in read_metadata
File "calibre\ebooks\metadata\book\base.py", line 444, in set_user_metadata
AttributeError: Custom field name '' must begin with '#'

============================================

OK, let's polish the book:

calibre, version 7.10.0 (win32, embedded-python: True)
Conversion error: Failed: (...epub name...)

Polish (...epub name...)
Traceback (most recent call last):
File "runpy.py", line 198, in _run_module_as_main
File "runpy.py", line 88, in _run_code
File "site.py", line 83, in <module>
File "site.py", line 78, in main
File "site.py", line 50, in run_entry_point
File "calibre\utils\ipc\worker.py", line 215, in main
File "calibre\ebooks\oeb\polish\main.py", line 300, in gui_polish
File "calibre\ebooks\oeb\polish\main.py", line 278, in polish
File "calibre\ebooks\oeb\polish\main.py", line 181, in polish_one
File "calibre\ebooks\oeb\polish\main.py", line 158, in update_metadata
File "calibre\ebooks\metadata\opf.py", line 108, in set_metadata
File "calibre\ebooks\metadata\opf.py", line 95, in set_metadata_opf3
File "calibre\ebooks\metadata\opf3.py", line 1080, in apply_metadata
File "calibre\ebooks\metadata\opf3.py", line 1067, in read_metadata
File "calibre\ebooks\metadata\book\base.py", line 444, in set_user_metadata
AttributeError: Custom field name '' must begin with '#'

==============================================

Now, if I were to look for:

File "calibre\ebooks\metadata\book\base.py"

... where would I look?

Windows File Explorer found:

C:\Program Files\LibreOffice\program\python-core-3.8.19\lib\email\mime\base.py

# Copyright (C) 2001-2006 Python Software Foundation
# Author: Barry Warsaw
# Contact: email-sig@python.org

"""Base class for MIME specializations."""

__all__ = ['MIMEBase']

import email.policy

from email import message



class MIMEBase(message.Message):
"""Base class for MIME specializations."""

def __init__(self, _maintype, _subtype, *, policy=None, **_params):
"""This constructor adds a Content-Type: and a MIME-Version: header.

The Content-Type: header is taken from the _maintype and _subtype
arguments. Additional parameters for this header are taken from the
keyword arguments.
"""
if policy is None:
policy = email.policy.compat32
message.Message.__init__(self, policy=policy)
ctype = '%s/%s' % (_maintype, _subtype)
self.add_header('Content-Type', ctype, **_params)
self['MIME-Version'] = '1.0'

What would I do with it?

But somehow I suspect that's not the problem, it's just the complainer, the problem is somewhere else, somewhere earlier.

More important to me as an end user . . . . why present me with an error I cannot fix INSTEAD of just going ahead and fixing the error and moving on?

Saying "Note, fixed an error by adding # in front of custom field name, if this auto-repaired EPUB misbehaves, please reload the original and try again".

Considering that Amazon had no problem converting this supposedly now uneditable by Calibre Ebook EPUB, I'd think that, yes, an auto fix is quite possible.

What to do now?

Thanks for exploring what might be going wrong here.

Post script: I tried deleting all user medatada, no difference, same errors for this EPUB. I have ONE custom column that works for everything else, even for this EPUB, but toggled off or on, empty or with something in the column, same errors:





.



.

[theducks]

#sent is not valid THERE. That should just say sent THERE

OTOH your Template MUST use {#sent}

In essence, you doubled up the #, which confused the parser

[kovidgoyal]

Somebody has tampered with that epub file leaving behind invalid metadata in its OPF. https://github.com/kovidgoyal/calibr...2b080895ad476f

[Peter Blaise]

Fascinating insights.

The original EPUB file edits fine, and converts fine, and of course, Amazon convers the 'tampered metadata' EPUB file fine, and the 'tempered metadata' EPUB reads fine even in the Calibre Ebook native on-screen reader.

Calibre Ebook can read and display the EPUB just fine, but cannot open the 'tempered metadata' EPUB to edit, save metadata, polish, or covert the EPUB to another format.

Calibre Ebook program was the only thing that touched the EPUB, and I used the various auto-fix features copiously without error before saving and exiting out before then trying to convert:

• Remove unused CSS rules

• Fix HTML - all files

• Beautify all files

• Check book

And Calibre Ebook puts it's own # in the custom column control panel, I did not put a # in the column name, see here:



The ODF file is 17 KB and has no # marks in it, according to Windows Notepad character search.

===========================================

I CAN duplicate the error, starting with a fresh copy of the original EPUB before editing.

When simplifying and flattening a table of contents ( all entries are h2 at the top of their own separate page file, subordinate 'headings' inside a 'chapter' are converted to <p><b>...</b></p> and are no longer in the table of contents), and then removing the 'noise' after #, such as #toc_1", #toc_2" in table of contents and reference files, replacing:

• #(.*?)"

with

• "

. . . "check book" says all is OK.

But save and exit, then I can't re-open the book for editing or converting.

Ouch!

So, on the one hand, don't clean up the ...#toc"-style table of contents references, even though "check book" says those are errors.

On the other hand, why does "check book" think a file is then OK with those errors removed if then the file can never be opened again after saving?

And, most importantly, why can't Calibre Ebook 'fix' the error on reopening the file to allow the user to get to the EPUB contents regardless of any metadata errors ( metadata errors that are apparently meaningless to Caliber Ebook's own reader and meaningless to Amazon's converter )?

In other words, thanks for the report on metadata, but as an end user, I'm using Calibre Ebook as an ebook editor ( primarily ), not a metadata editor ( a hopefully invisible yet discardable thing to us end users ).

That Calibre Ebook editor can manipulate metadata is terrific, but subordinate to the actual authored ( 'precious' ) contents of an ebook.

My goal is to strip an ebook file down to essentially only what can be heard when reading the contents out loud - that does not include metadata in anyway ( though I understand there may be metadata codes for read-out-loud programs to control pauses, inflection, and such, but such audio metadata is not on my event horizon at the moment ).

All that aside:

. . . could there be a way for Calibre Ebook to offer to open the contents of an ebook stripped of it's metadata, especially when Calibre Ebook finds errors in the metadata?

. . . instead of just saying "I can't open this ebook's contents, it's too hard to figure out what's what".

Or separately, is there a way to 'import' just the ebook's authored contents on demand?

==========================================

OK, three things:

1 - the "check book" inside the editor seems out of sync with the book checking routines in the external main-window parts of the Calibre Ebook program, such that they do not agree with each other whether an EPUB is OK or not - it would be nice if the editor's "check book" had the same savvy and warning BEFORE exiting the ebook editor to then never be able to reopen it for editing.

2 - It would be terrific if the editor, upon opening an ebook, offered to fix broken metadata, even tossing metadata if necessary, to allow the user to at least get to the authored ebook contents, even if devoid of formatting or demarcations that metadata might have provided.

3 - is there a way to import a stripped ebook without it's metadata?

Thanks.

.

[BetterRed]

Moderator Notice
@Peter Blaise - Please read and comply with MR Guidelines, in this instance #9.

Thank you

BR

[Peter Blaise]

Metatopic: "...BetterRed wrote: @Peter Blaise - Please read and comply with https://www.mobileread.com/forums/fa...ing_guidelines MR Guidelines, in this instance #9. Thank you BR ..." From that link, #9 ... Images included in posts must be no more than 600x600 pixels ... By doing what? Are you saying that MobileRead does not itself present image references within a 600x600 pixel box? That would be a terrific solution! The links I provided are to images hosted elsewhere so MobileRead does not have to host images - I don't even see the ability to upload images to a MobilRead post. Is there a way to 'thumbnail' the image links to 600x600 or less when we 'attach' an image link? You seem to have re-attached a copy of my image as just a number as if the copy is now hosted inside MobilRead ( I didn't give permission, but I don't care, someone else might care who hosts their own images ). When we click on the "Insert image" it asks for a link, but apparently MobileRead does not resize or check the size? How do we tell MobilRead to resize in the post or just put a thumbnail of the reference image? Here are the image addresses I used:
https://i.ibb.co/FVqv2Nd/Calibre-Ebo...2024-05-08.png https://i.ibb.co/WKJSKDK/calibre-put...2-19-09-PM.png I guess it keeps the thread from having a BIG image in the middle of it. Is there a 'best practices' way of giving folks a full image available for them to see what we're writing about? Thanks.

[DNSB]

Take a look at the paperclip tool at the top of the message entry box in Advanced mode. Or the Manage Attachments button below the message entry box in Advanced mode. Simply click on the thumbnail to see the full size image.

[theducks]

Another reason is some folk are bandwidth limited or have tiny screens.

[Peter Blaise]

How about a META DISCISSIONS forum so we can have side chats, "discussions about our discussions", "discussions about the MobileRead forums operation itself", OFF THREAD? So interrupters can briefly write "over here ___ link to discuss ____ problem", OFF THREAD.

.

I'm just trying to let master chef Kovid Goyal know that there's a mismatch between

• what the "Edit book" screen's "Check book" this is OK, what the "E-book viewer" thinks is OK,

• versus what the main "Calibre" program's "Embed metadata", "Edit Book", and "Convert books" think is NOT OK.

And ask if there's a way to 'fix' or re-import an EPUB without the supposedly now malformatted metadata.

Thanks.

.

[DNSB]

Originally Posted by Peter Blaise

I'm just trying to let master chef Kovid Goyal know that there's a mismatch between

• what the "Edit book" screen's "Check book" this is OK, what the "E-book viewer" thinks is OK,

• versus what the main "Calibre" program's "Embed metadata", "Edit Book", and "Convert books" think is NOT OK.

And ask if there's a way to 'fix' or re-import an EPUB without the supposedly now malformatted metadata.

Instead of Check Book which is a rather simplistic check, try using epubcheck. As for the re-import? You do not save a backup copy before editing?

[JSWolf]

Originally Posted by DNSB

Instead of Check Book which is a rather simplistic check, try using epubcheck. As for the re-import? You do not save a backup copy before editing?

I use both as Check Book does a better job checking the CSS and it catches some things that epubcheck doesn't.

[theducks]

and all the checkers in the world won't catch all the wrong info INSIDE the box.
Title set check
date set check (was set to 2095 but check was happy)