Split on Every Incidence of the Word Chapter

[DoctorOhh]

When I convert a book, I can go to Structure detection (see image) and insert the following expression in the "Detect chapters at (XPath expression" and insert this expression

//*[re:test(., "Chapter", "i")]

then go to Chapter mark: and choose pagebreak and the conversion will split the HTML at every instant of the word chapter. Making the creation of a TOC easy.

My question is when I'm in the book editor and I choose to split at multiple locations (see image). Which expression can I insert here to have the same result? Currently, after merging the files, if I see the need I will save the book and back out then convert the book to accomplish my desired results. I would much prefer to be able to insert an expression here and do the same thing.

Edit: when I cut and paste the expression above into the edit book split area I get an error. Here are the 2 most common errors.

ERROR: Unhandled exception: <b>AbortError</b>:Cannot split on the &lt;body&gt; tag

ERROR: Unhandled exception: <b>AbortError</b>:The expression //*[re:test(., "Chapter", "i")] did not match any nodes

[theducks]

Did you try the wizard (part of split at multiple locations)?

Usually I just split as is (the way I got the book) and then go back
and check those filenames that do not have the extras from a split., those I merge or delete if empty of content

[BetterRed]

Originally Posted by DoctorOhh

When I convert a book, I can go to Structure detection (see image) and insert the following expression in the "Detect chapters at (XPath expression" and insert this expression

//*[re:test(., "Chapter", "i")]

then go to Chapter mark: and choose pagebreak and the conversion will split the HTML at every instant of the word chapter. Making the creation of a TOC easy.

My question is when I'm in the book editor and I choose to split at multiple locations (see image). Which expression can I insert here to have the same result? Currently, after merging the files, if I see the need I will save the book and back out then convert the book to accomplish my desired results. I would much prefer to be able to insert an expression here and do the same thing.

Same one I use in Convert (cut and paste) just worked for me. But it didn't persist, maybe it will if you create the expression with the Wizard.

BR

[DoctorOhh]

Originally Posted by theducks

Did you try the wizard (part of split at multiple locations)?

I can use the wizard for underlying html code: p, h1, h2, b, etc... but haven't figured how to use the wizard to split on actual text content.

Originally Posted by theducks

Usually I just split as is (the way I got the book) and then go back and check those filenames that do not have the extras from a split., those I merge or delete if empty of content

Six of one half a dozen of the other. I've done it both ways, but I usually merge the content files first then split.

[DoctorOhh]

Originally Posted by BetterRed

Same one I use in Convert (cut and paste) just worked for me. But it didn't persist, maybe it will if you create the expression with the Wizard.

I haven't been able to use the wizard to split on actual book content. Split on html such as h1, h2, p, b, i, etc... easy, but not on words in the content.

[kovidgoyal]

They are both xpath expressions you should be able to use the same in both places.

[BetterRed]

Originally Posted by DoctorOhh

I can use the wizard for underlying html code: p, h1, h2, b, etc... but haven't figured how to use the wizard to split on actual text content.

I can't recall having any consistent token to which I would apply a rule. I do rare one offs with the splitter in Preview. And, given I mainly read on a Surface Go and I don't often share books, I don't know why I bother with splitting.

BR

[theducks]

Originally Posted by DoctorOhh

I haven't been able to use the wizard to split on actual book content. Split on html such as h1, h2, p, b, i, etc... easy, but not on words in the content.

Split multiple MUST split between block levels or you have invalid fragmented HTML

[DoctorOhh]

Originally Posted by kovidgoyal

They are both xpath expressions you should be able to use the same in both places.

That would make sense, but I have never had it run successfully in the edit books split location, yet it always works during the conversion process.
here is an example error.

Spoiler:

calibre, version 7.22.0
ERROR: Unhandled exception: <b>AbortError</b>:Cannot split on the &lt;body&gt; tag

calibre 7.22 embedded-python: True
Windows-10-10.0.26100-SP0 Windows ('64bit', 'WindowsPE')
('Windows', '10', '10.0.26100')
Python 3.11.5
Windows: ('10', '10.0.26100', 'SP0', 'Multiprocessor Free')
Interface language: None
EXE path: C:\Program Files\Calibre2\calibre-parallel.exe
Successfully initialized third party plugins: DeACSM (0, 0, 16) && Gather KFX-ZIP (from KFX Input) (2, 18, 0) && DeDRM (10, 0, 9) && Package KFX (from KFX Input) (2, 18, 0) && Audit Log (1, 0, 19) && Barnes & Noble (1, 5, 6) && Count Pages (1, 13, 6) && Diaps Editing Toolbag (0, 5, 0) && EpubCheck (0, 2, 5) && EpubMerge (3, 2, 0) && EpubSplit (3, 9, 0) && Extract ISBN (1, 6, 5) && FanFicFare (4, 40, 0) && Fantastic Fiction (1, 7, 1) && Fantastic Fiction Adults (1, 3, 1) && Favourites Menu (1, 3, 1) && Find Duplicates (1, 10, 9) && Generate Cover (2, 3, 4) && GetFileName (0, 2, 0) && Goodreads (1, 8, 2) && Goodreads Sync (1, 16, 11) && Job Spy (1, 0, 239) && KFX metadata reader (from KFX Input) (2, 18, 0) && From KFX (2, 18, 0) && KFX Input (2, 18, 0) && Set KFX metadata (from KFX Output) (2, 13, 0) && KFX Output (2, 13, 0) && Kindle Collections (1, 7, 29) && Kindle hi-res covers (0, 5, 1) && KindleUnpack - The Plugin (0, 83, 1) && Kobo Books (1, 9, 2) && Manage Series (1, 5, 2) && Modify ePub (1, 8, 5) && Open With (1, 8, 3) && Overdrive Link (2, 57, 0) && Quality Check (1, 13, 16) && Quick Preferences (1, 7, 1) && Reading List (1, 15, 4) && Resize Cover (1, 2, 1) && Save Virtual Libraries To Column GUI (2, 3, 0) && ScrambleEbook (0, 6, 0) && Search The Internet (1, 11, 3) && SmartEject (2, 5, 0) && View Manager (1, 10, 3) && Virtual Library Switcher (1, 0, 1) && Walk Search History (1, 5, 4)
Traceback (most recent call last):
File "calibre\gui2\tweak_book\boss.py", line 1455, in multisplit
File "calibre\ebooks\oeb\polish\split.py", line 282, in multisplit
calibre.ebooks.oeb.polish.split.AbortError: Cannot split on the <body> tag

[DoctorOhh]

Originally Posted by theducks

Split multiple MUST split between block levels or you have invalid fragmented HTML

That what it seems, but doesn't have that requirement during a conversion. However, Kovid indicates above that the expression should work in both places.

[DoctorOhh]

Originally Posted by BetterRed

I don't know why I bother with splitting.

I split only when creating TOCs based on files. Cleaning things up and splitting each chapter into its own file feels better when I'm reading.

[kovidgoyal]

Ah that's because splitting is a different algorithm. In the conversion what happens is the immediate tag is marked and the splitting happens on marked tags later on. Splitting directly is more complex. Instead use the transform html tool to makr the tags with a class or other attribute and then you use splitting wizard in the next step.

[DoctorOhh]

Originally Posted by kovidgoyal

Instead use the transform html tool to makr the tags with a class or other attribute and then you use splitting wizard in the next step.

Thanks for the insight.

I had decided simply saving the epub after I merge the files I desire to split, then switching to conversion is easier for me, than marking as you suggested. I'll give it a second look, maybe, once learned, marking will be best.

Thanks again.