Check for missing Punctuation

[abraum]

Hi,

i have created an ebook from a scanned book. The OCR was done with FineReader and converted into ePub. After that i use Calibre Editor. I have noticed that a lot of Punctuation at the end of the sentences are missing. Is there something like a "grammer check" that can find missing Punctuation?

[Quoth]

You can use a regex to search for lines that end with spaces, digits or letters. Then you have to look at source to see if ” ’ ! . ? is missing. A : or — is also possible to be missing. I suppose missing … , or ; or even ] is possible at the end of a line too
Lines (paragraphs) should end with </p> or <br/> (or similar)

Headings and some kinds of paragraphs (preambles, marginalia, lists) may not end with punctuation. So you need to check.

Always use wordprocessor and convert docx to epub in calibre.

An OCRed text is best proof read on eink and annotations copied back to PC.
Then edit wordprocessor source (odt format for LO Writer). Extra Save As in docx and import to calibre and convert to epub2. Then convert epub2 to any format.

I do the regex searching in LO Writer as it's the definitive edit source. I only edit image CSS in Calibre unless the source is an ebook.

[Tex2002ans]

Originally Posted by abraum

Is there something like a "grammer check" that can find missing Punctuation?

No. But there are some ways to speed this along.

Originally Posted by abraum

i have created an ebook from a scanned book. The OCR was done with FineReader and converted into ePub. After that i use Calibre Editor. I have noticed that a lot of Punctuation at the end of the sentences are missing.

The best way is to go back to Finereader and visually see exactly what was in the original. It's the only way you'd be able to tell what the missing character actually is, instead of blindly guessing.

If you recognize some sort of basic pattern, like:

• vol 1 -> vol. 1
• pp 123–130 -> pp. 123–130
• From 19992002 -> From 1999–2002

then you can always create Regular Expressions to deal with that.

If you are talking punctuation always missing at the end of lines/paragraphs:

<p>This is an example paragraph</p>

<p>This is another sentence</p>

<p>And this one has a correct period.</p>

then again, you can use Regular Expressions for that. See my recent posts in:

• 2022: "False paragraph breaks & RegEx"

If you have missing punctuation in the middle of paragraphs:

• This is an example And a second sentence And a third sentence.

sometimes grammarcheckers can catch this.

Antidote is the best one I've run across to catch this type, but this will only catch a very small subset of all the missing punctuation.

For more info on that, see my posts:

• 2018: "Does Tool Exist to Spellcheck/Grammarcheck by Category?"
• 2021: "RegEx: Insert nbs between initials, etc. "

- - -

Nothing beats going back to the original and figuring it out.

And it's impossible to come up with a general solution, because every single book is going to have different patterns of "missing punctuation".