[louabill]

I've been searching for some way to take some of my old powerpoint files and get the text portions into OmniOutliner as true outlines.

Advice I've seen is
. Use Keynote as the translator by importing into Keynote and then importing into OmniOutliner. My version of Keynote (1.1.1) cannot import the ppt files. Unfortunately, I don't have a newer version of Keynote.
. Use the translator at OPMLWorkstation.com. Unfortunately, it requires an account, and neither Safari nor Firefox nor Camino (under Mac OS X), nor Firefox nor IE (under MS Windows on Parallels) render anything on the page for new accounts.

I tried poking around with some of the Save as... formats in NeoOffice, but there is nothing of use (like OPML) there.

I have tried carting the ppt files to a machine that has MS Powerpoint installed on it, exporting the presentations as rtf, and then importing them into OmniOutliner. OmniOutliner does not import the outline properly, however. All items are at the top level, and though children have an indented format, they are not true children of their parents.

The only method I have found is to export to rtf, open in TextEdit, get rid of the bad rtf formatting (since it is easier in TextEdit than in OmniOutliner), and then import the result into OmniOutliner. After it is in OO, I can go through and reconstruct the outline structure. This is slow and tedious, however.

I have the feeling I'm missing something really simple and easy. Does anyone have any solution for this? I don't care about graphics---I would simply like the text as an outline.

Any tips would be much appreciated.

[louabill]

OK, there is a short-term solution: get an evaluation copy of iWork08.

Still... is there a solution that will work 1 month from now when my evaluation period has run out?

[louabill]

For them that's interested:

Keynote can be used as an intermediary to bring MS Powerpoint files into OmniOutliner with the following caveats:

• The file must be saved as an iWork'05 file. iWork'06 and iWork '08 get read in as a long series of empty items by OmniOutliner.
  
• All lines in the presentation which have multiple font faces will be blank. It appears that OmniOutliner's importing routines cannot read anything within <sf:span></sf:span> tag pairs in the xml generated by Keynote. (In reality, this can be more than just the lines in the presentation which have multiple faces... sometimes entire sub-lists are affected.)
  
  If you use emacs, here is a chunk of code which can be used to remove the offending span pairs (not that intersection of OO users and Emacs users is has many members, perhaps it is 1). Simply open index.apxl.gz file inside the .key bundle, run the following, and save. (Note that the large gaps in the [ ] at starts of lines are a space and a tab; it works better to literally add a tab character than to use the escaped character.)
  

  (defun oo-delete-span ()
    "Delete <sf:span></sf:span> tag pairs from iwork 05 files... allowing
  OmniOutliner to read in all the lines of an exported presentataion, sans
  formatting."
    (interactive)
    (save-excursion
      ;; most spans are on a single line
  	(goto-char (point-min))
  	(while 
  		(re-search-forward "
  [ 	]*<sf:span.*>\\(.*\\)</sf:span>" nil t)
  	  (replace-match "\\1"))
      ;; for some reason, spans around <sf:tab/> span multiple lines
  	(goto-char (point-min))
  	(while 
  		(re-search-forward "
  [ 	]*<sf:span.*>.*
  [ 	]*<sf:tab/>
  [ 	]*</sf:span>" nil t)
  	  (replace-match ""))
      ;; now take out spurious space left behind the spans
  	(goto-char (point-min))
  	(while 
  		(re-search-forward "[ 	]*
  ?[ 	]+\\(</sf:p>\\)" nil t)
  	  (replace-match "\\1"))
  	))

  I'm sure this isn't the best of lisp code, but it works (so far).
  
• All text boxes which are in the outline itself (such as boxes used to annotate a graphic) will be omitted from the outline. This is to be expected.

Well, there you have it. Not a pretty way to work, but it'll do for the things I have to translate right now.

Once again, if anyone has a simpler method, I would love to know it.