Duplicate ID Error line numbers wrong

[phossler]

The reported error lines (16 and 20) don't agree with the displayed line numbers (22 and 28)

Is there a miscount, or am I reading it wrong?

Seems like in every book.

[theducks]

You are reading it just confused.
The check is following the NCX and found issues

That is saying the referenced document has those 2 lines of which 1 is the duplicate
I frequently get this when recombining book that were (undesired result during conversion) split on headers (eg h1 as a chapter number and h2 used as a chapter title)
Both ended up with the same "id="

[phossler]

The reported error lines (16 and 20) don't agree with the displayed line numbers (22 and 28)

I understand the error condition. I was saying that the reported and hyperlinked to lines do not appear to be the ncx lines with the actual duplicate id error

So it says "Dup id on 16" but the ncx line seems to be line 22

[kovidgoyal]

Open a bug report and attach a book to reproduce the issue.

[phossler]

I tried to reliably reproduce the condition, but couldn't

Originally, the reported error lines (16 and 20) don't agree with the actual line numbers (22 and 28) in the ncx (Post #1), but now they do

Used the original pre-edit version of the book, I did what I had done, and now the line numbers agree 22=22 and 28=28

Is it possible that somehow I made some changes that affected the NCX and didn't run the check again?

If it happens again (and I've seen it before), I'll try to make a good test case

[theducks]

Originally Posted by phossler

I tried to reliably reproduce the condition, but couldn't

Originally, the reported error lines (16 and 20) don't agree with the actual line numbers (22 and 28) in the ncx (Post #1), but now they do

Used the original pre-edit version of the book, I did what I had done, and now the line numbers agree 22=22 and 28=28

Is it possible that somehow I made some changes that affected the NCX and didn't run the check again?

If it happens again (and I've seen it before), I'll try to make a good test case

I found that If you hand edit (add/remove lines) the NCX, the lines are off until you reload the book. Been that way for so long, I don't even think on it.

[phossler]

Originally Posted by theducks

I found that If you hand edit (add/remove lines) the NCX, the lines are off until you reload the book. Been that way for so long, I don't even think on it.

Cleaning errors is a sequential process for me so maybe that was it. I don't fix errors and warnings in any particular oder

I did have some errors (Link to "Not there file") and deleted some lines in the NCX as well as other errors

But I usually just re-run Check, not re-load the book, so that could explain it

Obviously if I add/delete NCX lines, I'd expect the reported and the actual line numbers to be off.

Thanks for the tip

[theducks]

Originally Posted by phossler

Cleaning errors is a sequential process for me so maybe that was it. I don't fix errors and warnings in any particular oder

I did have some errors (Link to "Not there file") and deleted some lines in the NCX as well as other errors

But I usually just re-run Check, not re-load the book, so that could explain it

Obviously if I add/delete NCX lines, I'd expect the reported and the actual line numbers to be off.

Thanks for the tip

Actually, if I rescan for errors, I WOULD NOT expect them to NOW be wrong, but sometimes they are.