[Libreoffice-ux-advise] [Bug 118455] Definition List (or Definition Index or Glossary) at the end of the page - EDITING

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=118455

--- Comment #15 from csongor at halmai.hu ---
I am not too familiar with indexes so I just made a couple of trials to see how
it works. Here is what I did:

I create the document and select one occurrence of "file". (Let's stay with
this example for the sake of simplicity.). On the Insert -> Table if Contents
and Index -> Index Entry dialog, I move "file" from the Entry field to the "1st
key" field and type the definition into the Entry field: "A file is a seqeence
of bytes.". I check the "Apply to all similar texts", click the Insert and then
on the Close button. After this, I go to the end of the document and Insert ->
Table of Contents and Index -> Table of Contents, Index or Biography. In the
Type field on the Type tab I select "Alphabetical Index" and voila, the index
is generated.

And here are my concerns.

1,
There is no indication around the occurrences of "file" showing that there is
an index entry for them. 

2,
I can modify the definition in the index (if enabled) but the change gets lost
once the index is re-generated. Which is needed if a new item and its
definition is added to the glossary.

3.
Finding the similar words can be problematic in several languages. In English
it is not so hard, "file" is always "file", perhaps "files" in worst case. With
prepositions it will become "with file", "in the file", etc. But the word
"file" is always "file". 

In other languages, however, the words can get prefixes and/or suffixes. Thus,
finding the other occurrences is almost impossible. For example, in Hungarian:

in file -> fájlban
to file -> fájlhoz
into file -> fájlba
from file -> fájlból
with file -> fájllal

BUT

feels pain in his leg -> fájlalja a lábát 

If you compare the last two translations (fájlal vs fájllal, single L versus
double LL), the substring "fájl" occurs in both but the second one has nothing
to do with computer files. It is about the pain. 

Thus, automatically finding the occurrences doesn't really work. On the other
hand, there is no way for manual assignment either. For example, I cannot
select the word "fájllal" and tell that this is an occurrence of "fájl" and
similarly, I cannot tell LO that the word "fájlal" is not related to the
computer files.

AFAIK, similar problems can occur in German as well, for example the singular
and plural version of "brother" is "Bruder" versus "Brüder", the plural and
other forms are hard to find automatically (please, correct me if I am wrong).

With my original proposed solution this would be possible easily. 

I think this work-around with indexes would bring a very poor user experience,
compared to the original proposal.

-- 
You are receiving this mail because:
You are on the CC list for the bug.