Issue 11215

I found some information on Lotus Wordpro Filters:
From: http://www.xpertss.com/cgi-bin/WebX?23@101.nptzachifeG.0@13%
40.ee735e2/0
"Here is what little I know about the LWP file format: It's based on 
something called Bento, developed originally by Apple in the early 
90's and used later for ScriptX and OpenDoc. The latter was a joint 
effort by Lotus, IBM, and others to develop an alternative to 
Microsoft's OLE format. OpenDoc didn't prevail, but Bento lives on 
in Word Pro and Notes (I don't know about the other SS apps). Here 
is a PDF with the first Bento specification: Bento Specification 
Revision 1.0d5 http://info.wgbh.org/upf/pdfs/BentoSpec1_0d5.pdf"

jw: reassigned to bh

*** This issue has been confirmed by popular vote. ***

Once we have some good Lotus SmartSuite (WordPro/123) import filters, 
our entire organization will completely migrate to OpenOffice.org. 
Import filters for Lotus WordPro and 123 Versions 97 through the 
current 9.8 should all be very similar, if not the same.

This was revisited under issue number 16369.  I've put a note in both places in
order to help keep them together.
see: http://www.openoffice.org/issues/show_bug.cgi?id=16369

Proposed (temporary) workaround found in discuss@openoffice.org.  It applies to
getting LWP documents into OOo under Linux/wine but the basics still apply.  

Briefly it means installing the Keyview viewers for lotus and copy>pasting the
data into OOo.  
The original post is at:
http://www.openoffice.org/servlets/ReadMsg?listName=discuss&msgNo=32597

I'm posting it here to keep the information together.

This has already been filed under
http://www.openoffice.org/issues/show_bug.cgi?id=11215

Wordpro filters are definitely needed.

*** Issue 27156 has been marked as a duplicate of this issue. ***

There is a package on the Lotus website that explains how to get XML to work
inside of Smartsuite, I'm not certain if this will help someone that knows what
it means to understand the file format or not, but here's the quote and the link:

"Use XML data in SmartSuite
Learn how to use XML data in SmartSuite. Details and samples are now available
for download."
http://www.lotus.com/products/product2.nsf/wdocs/cb09dd011bed60ba85256d5e006d8549

*** Issue 14426 has been marked as a duplicate of this issue. ***

I have gleaned this from a lead developer of Bento at Lotus/IBM:

Word Pro's document format is super-terrible format deluxe-- it is apparently
just a bunch of C pointer calls to the Word Pro functions that create objects.

Bento is a container which is tacked onto the end of the Word Pro document. You
may safely read Word Pro documents without having to parse the Bento container--
just read up until the Word Pro end-of-file marker. The purpose of Bento is to
put documents inside documents, like in Microsoft's DocFile, so simple Word Pro
documents do not require you to make sense of Bento. (Although, of course, you
will need to know it to export to Word Pro.)

I have the Bento specs, but apparently they've been scanned and uploaded
already. I also have the specifications for LMBCS, a very old multi-byte
protocol which might be used in Word Pro (I'm not sure).

I have a customer whose only major disappointment with the OOo/SO suite is its 
inability to import the default format of his existing office suite (indeed, 
anything released by IBM this century). If a viable Word Pro import gets added 
to OOo this year, you will have a second powerful OOo advocate in Perth, 
Western Australia (the first one being, of course, Jacqueline :-) as this man 
has wide networks of contacts in several fields. He's also interested in a 
this-century 1-2-3 spreadsheet importer (WK1 is seriously short of the mark) . 
 
I'd sit down and pitch in myself only I've already got enough work for at least 
three people, and it would take a substantial amount of time just to pick up 
enough OOo/SO context to write a useful import filter. Too busy mopping the 
floor to turn off the tap. )-: 
 
KOffice knows how to read Lotus AmiPro files, but AFAIK the file format is 
quite different to dotLWP. LWP can't be too hard, these guys - 
http://www.acii.com/online.htm - these guys - 
http://www.avantstar.com/stellent/groups/public/documents/word_files/quickviewplus8formats.pdf 
- and these guys - http://www.efax.com/en/efax/twa/page/supportedFileTypes and 
others can do it. 
 
If you want me to feed some files to LWP and see how they come out in the 
native format, that I _can_ do. I'll see if I can lay hand on a trial version 
or something and whack it onto my MS-lovin' friend's pet Windows 2003 box and 
RDP server for easy access. 

So - we are working on this problem.

verrry important: I need it to migrate to Openoffice (and we have thousands of
stored .lwp files). 

So - Bentoo is really not that useful wrt. the file format - at least for simple
files; the stream is just raw at the beginning.

Either way - we have a _very_ simple text importer - that does the moral
equivalent of 'strings' - just like (I suppose) an export->Ascii & re-import.

The code is available here: http://go-oo.org/ooo-build/patches/test/, 'lwp-*' we
can try to get it up-stream, but it really doesn't do all that much yet.

Basic style information is indexed via a table; unfortunately working out where
that table is is rather tricky; particularly because so much random stuff
changes between saves of almost identical files.

The lwp stuff will prolly ship in ooo-build builds for 2.0.0; I'm not optimistic
that it'll make it up-stream, with a cws, spec, iteam, issue, etc.

Created attachment 25771 [details]
The code that parses the lwp input stream and renders it in openoffice.

Created attachment 25772 [details]
Patch to detect the filter for lwp.

Created attachment 25773 [details]
patch to create a new directory sw/source/filter/lwp and to build the files in there.

*** Issue 16369 has been marked as a duplicate of this issue. ***

Anyone going to review and check in the changes to trunk?

converted existing patch to pluggable xml filter thingy, next attachment could
replace the above patches.

Created attachment 35066 [details]
patch for component import filter

Created attachment 35067 [details]
update ( removed unnessary stuff from makefile )

updated milestone & type to appear on the radar.

I'm available (and very excited) to test this when it becomes part of the
snapshot builds... I'm already subscribed to the QA Dev list... 

Michael: looks like we need someone who will take care about this bug. Who is it?

Can you please set target accordingly?

maybe Oliver is the guy, seems like this will be progressed via
http://wiki.services.openoffice.org/wiki/SummerOfCode2006#Writer:_Improved_Lotus_WordPro_Import_Filter

Oliver?

Sorry, the student who applied for this SoC project applied also for others. He
is now working on a bibliography project. 

New filters should not be located in the sw module. 
They should not use core functions of the Writer. Instead they should use the
UNO API and they should be implemented as an independent component. 

>Sorry, the student who applied for this SoC project applied also for others. He
>is now working on a bibliography project. 
Thats a pity

>New filters should not be located in the sw module. 
? this filter is not located in the sw module
>They should not use core functions of the Writer. Instead they should use the
>UNO API and they should be implemented as an independent component.
??? I presume you must be looking at the old patches, please look at the most
recent one(s), the filter although very very simple ( pure text, no formatting
etc.) is a component and doesn't use any writer functions. Still this is imo
useful, at least even in its present state someone can extract useful content
from a lwp doc ( better than no import ) 

You're right. I've seen the patch regarding sw/source/filter/lwp and that made
me think it's still planned to be put into sw.

One question remains: Why is sw depending on OOo:lotuswordpro?

>why is sw depending on OOo:lotuswordpro
it not ( well not really ), its only in the dependency list to force it to build
 which is presumably why sw also supposedly depends on the writerperfect filter.
I would imagine that there is a better way to do that :-) and both lotuswordpro
and writerperfect should be put in that ( unknown to me ) place

os: Ping? ;-)  Is it OK for integration?

mba: it's nearly 6 months since we re-wrote this for you to turn it into a
component so there was a chance of it going up-stream.
Shall we just commit it to a cws targetting 2.0.5 ?
Would you find some QA to look at it ? [ the feature set is extremely limited -
plain text extraction, so it shouldn't be hard ;-].

Sorry for the delay, I will check what was the reason for it. I assume that
Oliver was just buried under a lot of other work.

I will discuss with Oliver how we can make sure that this gets into the next
version (except if we find some big problems of course ;-))

Target changed to OOo 2.2
The patch will be integrated into cws os88

Created attachment 40154 [details]
WordPro file example with html and odt conversion

->npower: I attached an example lwp file and its conversion to html (all from
Google) and the resulting odt file. 

I don't think this filter should be integrated as it is now.

reassign

Can we expect that something happens to this issue at anytime soon? Means: will
someone improve the patch in the next few weeks? Otherwise I would prefer to
change the issue type from "Patch" to "Feature".

It seems that the patch will not be used anymore. So I change the type to
"feature". Target 2.2 is unrealistic -> 2.x.

I've checked-in the related files in the projects
officecfg, scp2, filter, postprocess
the new project /cvs/sw/lotuswordpro has been created _without_ cws tagging

target 3.x

as IBM officially joined the openoffice development team a few months ago, and
IBM owned the company which wrote lotus smartsuite, how about asking them from
the format?

I'm not sure how to go about it but IBM's Symphony
(http://symphony.lotus.com/software/lotus/symphony/home.jspa)is using the OOo
engine and has already provided LWP import functionality.  I wouldn't be opposed
to finding out the right way to beg them to provide it for OOo as it stands now.

I've created a request in the forum on the Symphony site:
http://symphony.lotus.com/software/lotus/symphony/supportThread.jspa?threadID=4714

It's the only place I could find to do anything like this.

In case you are not aware there is a working Lotus Wordpro filter available for
OpenOffice.org. Lwpfilter is an extension which you can install on OOo 3.1. I
have Win XP and it works fine, you should not have any problems with Vista
either. There is no version available for Mac. Go and download the extension
from the lwpfilter project page: 

http://code.google.com/p/lwpfilter/

The previous version of lwpfilter worked on OOo 3.0 but the current version
supports version 3.1. Lwpfilter is not available for OOo 2.x.

Please could someone interested in this filter check if the extension is good
enough? I gave it a try and it was able to load at least some simple documents.

http://code.google.com/p/lwpfilter/

As has been generally requested I've tried this filter with several WordPro 
document, some short and simple other long and complex. It's loaded the text, 
headers, footers correctly. The bullets and fonts are sometine different but 
OK. No formulas or images are included although their approx placement is shown.

The first time I used it - it crashed but this hasn't proved repeatable.

Thanks for using this extension in http://code.google.com/p/lwpfilter/ .

As mentioned in http://code.google.com/p/lwpfilter/, this filter is from cws
ooo11lotusfilter which is based on OOo 1.1. Because there are much different
between OOo 1.1 and OOo 3.0 and I have not much time. I had to disable some
features including image, OLE when I ported it to OOo 3.0.

IIRC, Lotus wordpro filter is included in IBM's Symphony. and IBM announced they
were migrating IBM's Symphony to OOo 3.0 in OOoCon2008. 

It would much better if asking IBM to upstream their new filter based on OOo3.0.
I don't want to duplicate the effort.

I tried the use the filter found at Google in OOo 3.1.0 on Windows XP and it
causes OOo to crash on every attempt to load an LWP file on my system. 
Unfortunately the crashes are repeatable on my system.

Pleading with IBM symphony to release their functioning filters makes excellent
sense.  They have the full Bento spec and have apparently made it work.  

pflin -> tcblack 

->I tried the use the filter found at Google in OOo 3.1.0 on Windows XP and it
->causes OOo to crash on every attempt to load an LWP file on my system. 
->Unfortunately the crashes are repeatable on my system.

would you please upload your file or send it to me so that I could have a look?

tcblack > pflin
Sadly (?) after a system crash I reinstalled my OS and OOo+ the filter and it
works!  So there was a problem in there somewhere from OS level all the way to
filter level, but I'll never be able to figure out where.  It works now, at
least on the several documents I've tested.  

I have noticed that the import does not "remember" the sections that were in the
LWP file and turn them into OOo sections.  

I have also noticed that it will not import some/all graphics.  Especially if
the graphics were Lotus draw object the import only results in a blue
translucent square OOo draw object.  So a bunch of my charts are effectively
lost unless I go ahead and re-install my old wordpro disk just long enough to
export them as JPG.

Unfortunately after migrating to OOo 3.2.1  The LWP filter from the Google code
site no longer installs.    (see
http://code.google.com/p/lwpfilter/issues/detail?id=3 and related links.)

The mentioned thread indicates that Go-oo (http://go-oo.org) does work with LWP
files.  Any chance we can get filters from there?

Here's a link to a search through the commit logs for LWP support: 
http://cgit.freedesktop.org/ooo-build/ooo-build/log/?qt=grep&q=lwp

Reset assignee on issues not touched by assignee in more than 1000 days.

Status set to patch
Maybe devteam can have a look?