Edit book/Check book/Parsing failed: Inconsistencies in counting

[chaot]

Aim is an uniform (unbold) endnotes format.


⁹⁴ vs. ^<b>95</b>

My relevant regex skills are still basic, so simple F&R and Replace all.

Searching done_Replaced 75 occurrences

Subsequently, in a second step, the rest would then done.
Find: </b></sup>
Replace: </sup>

Previously, indeterminate motives let me run Check book.

Check book_Parsing failed: Opening and ending tag mismatch ... as it was to be expected. But not only in 3 cases, but in 75.

Why Check book shows only 3 errors?

Run: Try to correct all fixable errors automatically
Run Check book again: No problems found

Control:
Find: </b></sup> Result: No matches were found for </b></sup>

This means Check book had corrected all 75 occurrences, but showed only 3. Whether these were the 3 first ones (left to right procedure), I can not say for sure.

Does that have to do with smart settings !? This should be a joke!

Interna: Conze (DAudV)

[kovidgoyal]

Once again, because when a parsing error occurs, check book is aborted. It's a simple concept, when there is parsing error, no further checks can be performed, because the HTML cannot be parsed reliably.

[chaot]

Reference (presumably): Because when a parsing error occurrs, all further error checking is aborted.

No concept without rule/s. Otherwise it's arbitrary.

If I understand correctly, no other types of errors, mistakes etc. may be displayed upon the occurrence of parsing errors. In the present case, however, only exist parsing errors, altogether 75.

Why are only 3 displayed? Why these 3?

By no means I want to annoy with pedantic questions. But a satisfactory answer I need for my understanding of the matter.

[theducks]

Originally Posted by chaot

Reference (presumably): Because when a parsing error occurrs, all further error checking is aborted.

No concept without rule/s. Otherwise it's arbitrary.

If I understand correctly, no other types of errors, mistakes etc. may be displayed upon the occurrence of parsing errors. In the present case, however, only exist parsing errors, altogether 75.

Why are only 3 displayed? Why these 3?

By no means I want to annoy with pedantic questions. But a satisfactory answer I need for my understanding of the matter.

Because, that is the way it works
Many cleaning programs require multiple 'passes' to find and expose the 'next error' after a repair.

It keeps things simpler, just to stop when it hits a Parse error, rather than figure what it would take to MEANINGFULLY continue

I have been using computers from before the IBM PC. It has always been a standard debug technique to:
run, fix, run
until the compiler finishes without error (PS I have lots of BUG creation and fixing practice )

[chaot]

OK, that should suffice as an explanation.

Conclusion: It is necessary to know and take into account the special features regarding the treatment of parsing errors in calibre.