Re: [office] OFFICE-2102 test files

dear fans of the white-space soap opera,

so i tried to write this proposal for the issue and it sounded so easy,
but then i noticed that i can't just replace "descendant elements" with
"elements that allow <text:s> etc. children" and be done with it.

then i've looked in detail at what our implementation actually does, and
was surprised that there were things i hadn't noticed yet...

for lack of a better idea, i've added some comments to OFFICE-2102
containing several competing proposals, all of which describe what the
LO implementation does :)

unfortunately it appears that consumers differ a bit in how they
interpret the cases that the algorithm in ODF 1.2 neglects to even
mention, as i found with some test documents:

- whitespace.odt:
  produced by LO current master build (5.4), contains examples with
  whitespace both inside and following all the interesting elements,
  one paragraph per element
  (naturally this round-trips to itself with LO)

- whitespace-test.odt:
  the same, manually edited to replace every <text:s/> with a U+0020
  SPACE, to see if these are ignored

- whitespace-test-rt.odt: whitespace-test.odt roundtripped with LO

- whitespace-test-word2010.odt: whitespace-test.odt roundtripped
  with Word 2010 (only version i have)

- whitespace-word2010.odt: whitespace.odt roundtripped with Word 2010

please see comments in OFFICE-2102.

i suggest we monitor the situation closely before it escalates further...

-- 
Michael Stahl | Software Engineer
Platform Engineering - Desktop Team
Red Hat

Better technology. Faster innovation. Powered by community collaboration.
See how it works at redhat.com

Red Hat GmbH, http://www.de.redhat.com/, Sitz: Grasbrunn,
Handelsregister: Amtsgericht München, HRB 153243,
Geschäftsführer: Charles Cachera, Michael Cunningham, Michael O'Neill,
Eric Shander