Book alphabetical index exclude title and roman numeral pages

[maxamilian]

In Writer 7.6.7.2 Fedora 39 linux, when I add an alphabetical index at the end using my .sdi file, I choose Entire Document option and it indexes words from title page un-numbered, the roman numeral pages, and then the rest. I want to exclude the front unnumbered and roman numeral pages. If I choose Chapter instead of Entire Document, there is no index. What did I miss?
Basically my Index at the end should have only arabic numbers 1-N, not the i, ii, iii, iv.
Here is an example of what I get: Robot 2-iv, vi, viii, 20-22, 24, 26-29,…
I want it to be: Robot 20-22, 24, 26-29,…

I’ve searched and can’t find the answer to this specific question.

[ajlittoz]

This is what should be expected with a concordance file: it is the lazy way to index all occurrences of words, no matter where they are located (perhaps except in headers and footers).

You can’t exclude parts of your document from concordance matching. So you’re left with a manual re-indexation:

1. use Edit>Find & Replace to search for a word with Find All
2. with all the occurrences selected, Insert>TOC & Index>Index Entry
   Enter the parameters for this word
3. remove manually the index entries in the “non-indexed” parts
4. erase the word in the concordance file

[maxamilian]

Thanks. I think that will work because in the front title/publishing and roman numeral i,ii,iii,iv… pages each indexed word is highlighted or flagged with a leading grey area. So I just need to select each word that was visibly indexed, and choose the remove option in the popup (I recall there was a popup), and what remains is the indexing for the remain arabic 1-N pages. I need to keep my .sdi, >800 words.

I remember this now because a few years ago I made an automation to remove all indexed words manually with gui and xdotool to click the gui remove/delete (I don’t do that anymore, whew!).

[maxamilian]

Well, there used to be a popup window that would traverse the alphabetical indexes in order through all pages, and it had a delete option. I can’t find that anymore. I was going to use that to delete the alpha indexes from “non-indexed” front.

Ah, I think I remember now; Before I switched to sdi I used the internal method and that method allowed the popup to traverse entries with a delete option. The delete option is what your #3 describes, I think. I’ll switch back to the internal method and not the sdi file.

[ajlittoz]

To get the “traversal dialog”, right-click on an index entry in text (if you enabled the various hints in View menu, the presence of an index entry is indicated by a gray background). You then have the “left” (previous) and “right” (next) navigation arrows.

[maxamilian]

Ahah! I think I misunderstood your reply. Did you mean that if I use the sdi method that I have to followup with a manual re-indexation steps 1-4? I can do 1 and 2, but at the end of 2 I click Insert. I don’t know how to do #3 remove manually the index entries in the “non-indexed” parts. The popup for #2 doesn’t have the remove option. Does “remove manually” mean to unzip the files and edit with vi, or is there a gui to do that? I remember a few years ago there was a gui popup to traverse alphabetically indexed words, but I can’t find it now.
Thanks!

[maxamilian]

I found it, right click the greyed area on index, popup has Edit Index Entry, select and the popup has the controls I remember including next and delete. So I can use this to remove the indexes from the front parts that shouldn’t be indexed, and using your steps to 1-4.
Thanks!

[maxamilian]

So it seems the two choices for alphabetical book index are non-concordance and concordance sdi file. I have 800+ entries so there’s no way I’m going to manually 1 by 1 find all occurrences and add to index. With the concordance sdi file I can wait to the end of project and create index w/sdi file, enable editable index, and in 1 minute have the index created 2 columns and formatted. Then ~20 minutes to manually remove all page entries <7 which is right after the TOC, that’s ok. If I had a wish list, it is to have the Index creation/edit page have a page range instead of entire document, then I could specify 7-298, leaving out the title/publish and Roman Numeral pages.

I did try to make to index words without concordance sdi file, robot and pair, and then after choosing the sdi way I had to delete all those with the Edit->IndexEntry->Delete. After a few clicks I realized this would be an hour, so I revived my old automation on Unix X11; xdotool click 1. This can be automated because when delete runs, the next entry is ready for deletion automatically, so a simple repetition of mouse clicks is needed. The shell has a while loop. After all deletions are done the clicks continue so make sure the popup isn’t in front of other buttons behind it, then move cursor to while terminal and ^C to stop it. The first sleep 5 gives you 5 seconds to move from the terminal window to the popup delete button.

user@host:~$ sleep 5; while /bin/true ; do

xdotool click 1
sleep .5
done