Sick and tired of acting as proofreader for lazy publishers

[doubleshuffle]

Originally Posted by gbm

This one works in my test.

( <a id="page_([0-9]+)"></a> )

In the Calibre Editor? Rejected it as invalid when I tried it.

[eschwartz]

Originally Posted by doubleshuffle

Indeed. I'm generally rather clueless about Regex myself, but RegexBuddy and Mobileread together are a great lifesaving outfit...

Or their (RegexBuddy's) free guide here: http://www.regular-expressions.info/

The most helpful regex tutorial I have come across, it is my go-to reference.

[gbm]

Originally Posted by doubleshuffle

In the Calibre Editor? Rejected it as invalid when I tried it.

Yes, see screenshot.

bernie

[DiapDealer]

Originally Posted by eschwartz

Or their (RegexBuddy's) free guide here: http://www.regular-expressions.info/

The most helpful regex tutorial I have come across, it is my go-to reference.

I'll second that. Got that puppy on speed-dial.

It would seem to me that none of the parentheses in the expressions being bandied about are actually necessary. Surely

 <a id="page_\d+"></a>

would suffice (a space preceding and following)? Replace it with nothing.

[doubleshuffle]

Originally Posted by gbm

Yes, see screenshot.

Ah, upon a closer look I saw that yours is a little different from HYPE's. Yours works for me too.

[Joques]

Thank you very much for all your input! I'll try it out tomorrow. Just one more question, even if it is a bit to the side of the main point of complaining about lackluster proofing and formatting on the part of certain publishers:

Is it possible, in Calibre, to search/replace globally in an entire book, as opposed to in individual chapter files like I'm doing now?

[doubleshuffle]

Originally Posted by Joques

Is it possible, in Calibre, to search/replace globally in an entire book, as opposed to in individual chapter files like I'm doing now?

Of course. You have to select "All text files" in one of the dropdown menus (see gbm's screenshot a few posts above.)

[Joques]

Thank you again. Who ever said that whining never accomplishes anything?

[doubleshuffle]

And isn't it great for the lazy publishers that they have a whole team here to correct their mistakes?

[j.p.s]

Doesn't even the most rabid command line hater have to admit that copy and pasting:

unzip file_name.epub
sed -ie 's; <a id="page_[0-9]*"/> ;;g' *.html
zip -u file_name.epub *.html

is the simplest and easiest way, provided unzip, zip, and sed are installed? (The unzipped files would then need to be deleted, but that should not be a problem.)

[JSWolf]

Randy has spent al <a id="page_540"/> most no time around military people.

Randy has spent <a id="page_540"/> almost no time around military people.

If you have any of the two situations in the same eBook, you cannot use regex to fix it as you'll be removing spaces you want to keep. So unless you are 100% sure the regex won't be getting rid of spaces, go for it. But even if .01% unsure, don't do it.

[j.p.s]

Originally Posted by JSWolf

Randy has spent al <a id="page_540"/> most no time around military people.

Randy has spent <a id="page_540"/> almost no time around military people.

If you have any of the two situations in the same eBook, you cannot use regex to fix it as you'll be removing spaces you want to keep. So unless you are 100% sure the regex won't be getting rid of spaces, go for it. But even if .01% unsure, don't do it.

That is a good point. It seems very unlikely that this page count thingy would always be in the middle of a word.

[DiapDealer]

I don't think anyone's recommending blindly doing a Replace All. I Know the Cryptonomicon is pretty dang large, but once you get in a rhythm, it wouldn't take that long to step through and verify every occurrence. Once through and you're done.

It's pretty safe (not to mention smart) to assume that no regex is "Replace All"-safe. That assumption is why many of these errors creep in to begin with ... too much confidence in one's own S&R skills.

[JSWolf]

But given that that's page 540 that's showing and the book is over 1000 pages, that's going to be a lot of manual search/replace.

[DiapDealer]

Hour and forty-five minutes. Tops.