Apache OpenOffice (AOO) Bugzilla – Issue 88509
Numbering/bullets without text not included in word count
Last modified: 2013-08-07 14:44:35 UTC
According to the changes made for resolving issue 80815 the bullet and numbering characters should be counted as words, but if they have no text after them, they aren't counted as words. See the attached testcase.
Created attachment 53059 [details] Testcase
Reassigned to SBA
fme->troodon: Please check agains MS Word. There, numbering characters are only counted if there's some text following the numbering.
->fme I've checked before reporting this :-). With a tescase like the one I attached, both Word 97 and 2000 say there are 12 words (5 numberings, 5 bullets and 2 letters). AFAIK the way Word does the word count hasn't changed in later releases.
Created attachment 53113 [details] Same testcase for Word
Created attachment 53114 [details] Screenshot showing OOo DEV300m9 and Word 200 side by side
fme->troodon: This is too weird. Have a look at the new attachment. One empty, numbered paragraph - word count is 0. khong also made this observation as mentioned in issue 80815: "[...] To reproduce, in MS Word, click on bullet, the word count is 0, type a word, word count is 2. [...]". That's the way I implemented it in OOo Writer: The bullet/numbering string is not counted if the paragraph is empty. Looks like this needs some more investigation.
Created attachment 53122 [details] new test document
troodon->fme It's weird because someone agreed to implement a the same word count method as in Word, and Word has a very weird method :-). *Only* the new bullets/numbers inserted at the end of a list are (not) counted as you described in desc8. Try it yourself :-). In a list with n items, delete the text of any item, but not the item itself. That item won't have any text but the number/bullet will be counted as a word. Now try another thing. Insert a new item after any position < n (any except the last one) but don't enter any text. That item will be a new empty item but the number/bullet will be counted as a word.
troodon->fme CC'ing you to attract your attention over desc 10 :-)
I'll try to adjust the word count algorithm regarding numbering/bullets for OOo 3.0.
troodon->fme In case it helps, I've tried with Word 2007 and it exhibits the same behaviour as I described in desc 10 for previous versions.
fme->troodon/foobard: Right, I suggest to fix this the easy way: Count numbering/bullets no matter if the paragraph contains text or not. If you agree, I'll fix this for 3.0, otherwise I'll have to shift this to 3.x. What do you think?
->fme Getting it right is more important than getting it done. If I understand your suggestion, you want to cut some corners in a way that won't exactly duplicate how MS Word does it. I think this would be a mistake. If we can't hit 3.0, that's a shame, but it would be a bigger shame to close this issue without an adequate fix. my $0.02.
fme->foobard: [...] If we can't hit 3.0, that's a shame [...] Please choose your words more carefully. You should adjust your expectations a bit. This is not the only issue I have to fix and this definitely is not the most important one. From my point of view, a step in the right direction is better than no step at all. Just my 2 cents.
->fme You asked me what I thought. I'm sorry you didn't like the answer. It was not my intention to be rude. Do what you can with what you have. If you feel a halfway solution is the only thing doable by 3.0, given available resources and priorities, I won't complain. However, please don't close this bug, as the issue will only be partially solved.