Quick simple way to remove <a> ids?

[Siavahda]

I have an EPUB that's scattered through with apparently random <a id>s, which wouldn't bother me except the marked text shows as underlined on my Kobo.

Is there some way to do a search and replace, or something similar, to get rid of them all at once? Obviously each ID is unique, so I can't think of any other way to get rid of them other than manually, one by one. Which is a literal pain, since my disabled hands can't handle that much clicking.

Each ID number/string of letters appears in the stylesheet like so

.calibre25 {
-aw-bookmark-end: _Hlk91081997;
}

I tried deleting those rules (ie removing calibre25 from the stylesheet), but that did nothing. I tried removing the id number itself from the rules (so calibre25 was just empty), but nope. Obviously if I do a search and replace for <a id=> what I'm left with is a bunch of things like this <_Hlk91081997>, which then appear like that in the html and would still need to be manually removed individually. Using the remove unused css rules or the beautify all files tools didn't help either.

Converting to DOCX and then back to EPUB worked, in that it got rid of the <a>s, but it messed up a lot of the other formatting.

If you have an idea, please explain like I'm 5, I'm not super tech-savvy!

[Karellen]

Find: <a id=".*?">
Replace: leave it blank

Set your search mode to Regex and All text files.
step through a couple one at a time to make sure it works as expected, then use Replace All if it is working correctly.
Don't know where the calibre25 comes from or how it fits into it.

[Siavahda]

Originally Posted by Karellen

Find: <a id=".*?">
Replace: leave it blank

Set your search mode to Regex and All text files.
step through a couple one at a time to make sure it works as expected, then use Replace All if it is working correctly.
Don't know where the calibre25 comes from or how it fits into it.

Thank you so much, that worked perfectly!

RE calibre25: I had no idea what might be relevant, so I was trying to provide as much info as possible, is all.

[JSWolf]

Even easier is to use Diaps Editing Toolbag in the calibre editor.

[Karellen]

Great! Glad it worked.

[theducks]

I clean the CSS (what you showed) entry of any -aw-bookmark (there are start and end types)

\.calibre\d+ \{\s+\-aw\-bookmark\-.+?: .+?\s+\}

those with:
any calibre number {
any that start with the phrase
}

[DNSB]

Originally Posted by theducks

I clean the CSS (what you showed) entry of any -aw-bookmark (there are start and end types)

\.calibre\d+ \{\s+\-aw\-bookmark\-.+?: .+?\s+\}

those with:
any calibre number {
any that start with the phrase
}

Would that not still leave the <a..>...</a> in the text portion? Remember that by default, an ePub underlines and uses blue for a link.

I'm using Sigil so I delete all the -aw-bookmark items from the CSS, remove unused from the text and then use the editing toolbox to remove any naked <a> tags.

[DNSB]

Originally Posted by Karellen

Don't know where the calibre25 comes from or how it fits into it.

When the book was converted to ePub, calibre put every bookmark into the CSS with calibrexxx.

[Karellen]

Originally Posted by DNSB

When the book was converted to ePub, calibre put every bookmark into the CSS with calibrexxx.

Yuk. The css must be immense.
Would running "remove unused css rules" clean it all out?

[slm]

It may be worth noting that you can set up Calibre's file conversion to remove all <a> tags: In "the "Look and Feel" part go to the "Transform HTML" tab (last tab.
Add a rule that sets "If the tag name is" to "a" and select "remove tag only" as the action.

Its always amazing how much is packed into Calibre!

[DNSB]

Originally Posted by Karellen

Yuk. The css must be immense.
Would running "remove unused css rules" clean it all out?

No when calibre creates those rules, they are being used. What I do is use a regex to remove all the rules from the stylesheet and then remove all unused rules from the text and lastly remove all <a> tags.

That is one reason I personally preferred to use KindleUnpack to convert azw3/KF8 to ePub.

[DNSB]

Originally Posted by slm

It may be worth noting that you can set up Calibre's file conversion to remove all <a> tags: In "the "Look and Feel" part go to the "Transform HTML" tab (last tab.
Add a rule that sets "If the tag name is" to "a" and select "remove tag only" as the action.

Its always amazing how much is packed into Calibre!

The problem is that you would need a rule for every ID for a bookmark in the book. Removing all the <a whatever> tags would be a baby/bathwater situation.

[JSWolf]

Originally Posted by slm

It may be worth noting that you can set up Calibre's file conversion to remove all <a> tags: In "the "Look and Feel" part go to the "Transform HTML" tab (last tab.
Add a rule that sets "If the tag name is" to "a" and select "remove tag only" as the action.

Its always amazing how much is packed into Calibre!

If the eBook has footnotes, they get blown away. If the book has links from an index, that gets blown away. Any other valid links l=such as the HTML Toc (if you want to keep it) gets blown away and you are left with just a text list.

So don't do this or you could ruin the eBook.