Index Quirks in Writer

users
#1

I know the table of contents feature seems to work well, as I've used it
quite a bit in the past.
But creating an alphabetical index (a first for me, at least using Writer)
seems to have a number of rather annoying features. I've provided an
experiment you can do at the end, but first let me discuss the issues.

1) When creating an alphabetical index, the Columns feature works as
expected, BUT ONLY FOR THE STANDARD (DEFAULT) PAPER SIZE! If I attempt to
use columns on a 6" x 9" page, Writer still seems to assume a "standard"
paper size and doesn't seem to know that I'm using a 6" x 9" page. Thus no
matter how I try to tweak the setup, the right column is almost completely
off the page on the right. So I can't use columns for the Index - Bummer,
but not the end of the world.

2) If I go through and highlight each entry that I want indexed, everything
works great (so far as I can tell), and I have the options for "match case"
and "whole words only." But in a long document, using a concordance file
certainly seems to make more sense, and it SEEMS TO WORK, but it actually
doesn't.

  Issue a) No matter what boxes are checked, Writer goes and marks ALL
instances of whatever the concordance entry has: regardless of whether or
not it is a whole word or whether the capitalization matches.

  Issue b) If you add another entry to the concordance file and "Update
Index/Table," all instances (even incorrect ones) of the new entry are
indeed marked within the file and added to the index, but every earlier
entry GETS AN ADDITIONAL marker. As I modify the concordance file to add new
items and update the index, I find that the earliest entries have as many
markers as the number of updates I've done.

Here's an experiment you can do:

Open a new document using whatever standard size is in effect.
At the top of the document, type dt and press F3. This generates the dummy
text.
In a separate text editor create a concordance file with the following
entries:

Breeze;Breeze;;;0;0
Long;Long;;;0;0
Self;Self;;;0;0
Wrist;Wrist;;;0;0

Back in the document, create a new index at the bottom of the page, and mark
"Case Sensitive." Then choose the concordance file created above.

You'll see that the words "himself" and "along" are marked in the document
and included in the index as "Self" and "Long", even though these words
never appear in the text. Now add the following entry to the concordance
file:

Eat;Eat;;;0;0

Find it? It actually marked the last three letters of the word "sweat" which
doesn't match the case.

In a three hundred plus page document with a surname index this tends to
make the index pretty useless, since it's cluttered with erroneous entries
which will drive you nuts looking on a referenced page to find something
that just isn't there.

Sadly, I've obviously done that, which brings up another issue:

How do I get rid of the index markers? Going through one by one is far too
tedious. I recall that the alternate search and replace had a selection for
"index" under properties, but the I downloaded that and the latest version
seems to have a serious bug which doesn't let you get to a specific entry on
the drop down list for "properties" so that's out. And I don't know if that
would be what I need anyway.

Any help would be appreciated.

#2

CVAalkin,
The last time I worked with LO indexes, I was trying to make an alphabetic sort of the table of contents (heading styles), which did not work. So, I am not all that experienced with indexes. I find them a bit of a pain to create. However, the paper size problem you mentioned may be just a matter of changing the Index entry paragraph style(s). You might try that to bring your margins in.
HTH.
Girvin Herr

#3

CVAlkan wrote

Here's an experiment you can do:

Open a new document using whatever standard size is in effect.
At the top of the document, type dt and press F3. This generates the dummy
text.
In a separate text editor create a concordance file with the following
entries:

Breeze;Breeze;;;0;0
Long;Long;;;0;0
Self;Self;;;0;0
Wrist;Wrist;;;0;0

Back in the document, create a new index at the bottom of the page, and
mark "Case Sensitive." Then choose the concordance file created above.

You'll see that the words "himself" and "along" are marked in the document
and included in the index as "Self" and "Long", even though these words
never appear in the text. Now add the following entry to the concordance
file:

Eat;Eat;;;0;0

Find it? It actually marked the last three letters of the word "sweat"
which doesn't match the case.

This is due to the "0;0" at the end of each entry. These two options
represent Match case;Word only.

CVAlkan wrote

How do I get rid of the index markers? Going through one by one is far too
tedious.

Right-click on the index > Edit Index/Table... > Index/Table tab > Options
section > pull down the concordance list and selection Edit... and then you
can check case and word only entries as required. Clicking OK/OK, should
immediately update the document.

#4

Hi :slight_smile:
Apparently it's well worth getting
Zotero
and apparently that works with both LO and MS Office. I think it's
not an Extension for LO but is a fully-fledged 3rd party tool
Regards from
Tom :slight_smile:

#5

Owen:

Re: the 0;0 at the end. Changing this to 1;1 does indeed set the 'match
case' and 'word only' parameters. I got the format for the concordance file
from another LO user and stupidly didn't bother asking what the two zeroes
were for (I was in a rush and figured defaults are usually satisfactory). In
my defense, however, I just went back and looked at the LO Help, and even
the LO Writer user guide, and couldn't locate this information.

BUT, this doesn't solve the issue I'm having, which is that the errant index
marks are all in the document already and remain there after updates. The
only way to remove them that I can see is to scroll through manually and
delete them, since there isn't any way (at least that I can see) to search
for them much less search and replace them.

Thanks, however, for pointing out the use of the zeroes in the concordance
file.

Frank

#6

Hi Girvin:

Thanks for the response, but I tried that and couldn't find anything in the
style that would cause that.

I even created a brand new style from scratch and applied that to the index
entries.

What I did observe is that, although page numbers for the index are set to
be flush right, neither the default style nor the one I created included any
tabs at all. It sure appears that the right hand starting point seems to be
"fixed" somehow.

Frank

#7

CVAlkan wrote

In my defense, however, I just went back and looked at the LO Help, and
even the LO Writer user guide, and couldn't locate this information.

I understand Frank. We all make mistakes in our haste. I make plenty. The
Writer Guide v4.0 does not list this detail but this help page does:
https://help.libreoffice.org/Writer/Edit_Concordance_File

CVAlkan wrote

BUT, this doesn't solve the issue I'm having, which is that the errant
index marks are all in the document already and remain there after
updates. The only way to remove them that I can see is to scroll through
manually and delete them, since there isn't any way (at least that I can
see) to search for them much less search and replace them.

I did a test using v4.1.4.2 and found that all marks inserted by the
concordance file could be removed by updating the concordance file. You can
probably make a global change in a text editor of the zeroes to ones, but
the key is to then re-edit the concordance file while in the Index/Table
dialog (via the Edit option I mentioned). Let me know if it works, because
my test was just using the basic example you indicated. It did however
appear to update according to the selections I made when editing the
concordance file in this manner.
Best wishes, Owen.

#8

If you go to Edit Index/Table | Entries | Structure and Formatting | Structure, you will see that each entry comprises "F T #" - or Entry, Tab stop, Page number - with the possibility of interspersed text. (I think it really means Tab character, not Tab stop.) If you click the "T", you will see that the default fill character is a dot and that the Tab stop position is adjustable but by default is overridden by "Align right".

I trust this helps.

Brian Barker

#9

Brian:

Thanks as usual for your explanation - someone needs to make sure all your
posts are forwarded to the documentation team so they can see what they've
inadvertently overlooked.

Part of the confusion (at least to me) is the relationship between the index
structure settings and the paragraph definitions in the normal styles used
for the index table entries. I began experimenting with how these interact
(not a good idea as I'm actually trying to accomplish something, but I was
curious). Luckily it was very cold outside so the windows were tightly shut
and I couldn't jump. It all seems to make perfect sense when setting up a
table of contents, so why this should be so confusing I don't know - maybe
I'm just too old.

BUT - the main point is that you got the formatting squared away. Still
other issues with the alphabetical indexing however, but I'm about to make
another post on that.

Frank