Bug 58214

Summary:	Error with outlook while opening an msg files extracted by POI from another msg file
Product:	POI	Reporter:	Alexandre <alexandre.brun>
Component:	HSMF	Assignee:	POI Developers List <dev>
Status:	NEEDINFO ---
Severity:	normal
Priority:	P2
Version:	3.13-dev
Target Milestone:	---
Hardware:	PC
OS:	All
Attachments:	The master mail with attachment Embedded outlook msg extracted by POI Embedded outlook msg saved by Outlook My simple java code Attachment from outlook, POIFSLister result Attachment from POI, POIFSLister result Attachment from outlook, HSMFDump result Attachment from POI, HSMFDump result

Description Alexandre 2015-08-05 12:56:22 UTC

Created attachment 32960 [details]
The master mail with attachment

As describe in this bug : https://bz.apache.org/bugzilla/show_bug.cgi?id=58211 

I have an Outlook message (Test mail attachment) in attachment of another Outlook message (Master mail.msg). 
The msg file saved using POI (Test mail attachment_from POI.msg) generate an error when I try to open it in Outlook. Outlook does not recognize it.

I join the attachment msg file saved by outlook (Test mail attachment_from outlook.msg). 

I also join my code in case. (ExtractMsg.java)

Comment 1 Alexandre 2015-08-05 12:57:14 UTC

Created attachment 32961 [details]
Embedded outlook msg extracted by POI

Comment 2 Alexandre 2015-08-05 12:58:33 UTC

Created attachment 32962 [details]
Embedded outlook msg saved by Outlook

Comment 3 Alexandre 2015-08-05 12:58:58 UTC

Created attachment 32963 [details]
My simple java code

Comment 4 Nick Burch 2015-08-05 14:35:57 UTC

Thanks for those

Any chance you could try running dev tools like org.apache.poi.poifs.dev.POIFSLister and org.apache.poi.hsmf.dev.HSMFDump against the two extracted files, and see if there are any obvious differences between them? Sections in one not the other, different IDs, that sort of thing. That should help us narrow in on what to change

Comment 5 Alexandre 2015-08-05 14:53:46 UTC

Created attachment 32966 [details]
Attachment from outlook, POIFSLister result

Comment 6 Alexandre 2015-08-05 14:54:15 UTC

Created attachment 32967 [details]
Attachment from POI, POIFSLister result

Comment 7 Alexandre 2015-08-05 14:54:53 UTC

Created attachment 32968 [details]
Attachment from outlook, HSMFDump  result

Comment 8 Alexandre 2015-08-05 14:55:13 UTC

Created attachment 32969 [details]
Attachment from POI, HSMFDump  result

Comment 9 Alexandre 2015-08-05 14:57:45 UTC

There are differences between two extracted files using both tools. 

Hard for me to interpret differences.

Comment 10 Alexandre 2015-08-31 14:39:53 UTC

Hi, 

do you have any update about this issue ?

Comment 11 Nick Burch 2015-09-01 10:54:03 UTC

If you're able to identify the differences, we might be able to make a quick fix, or otherwise guide you on making the fix yourself

Otherwise, this issue will remain until someone volunteers to spend their free time looking at it, be they a committer or just someone else interested from within the community

Comment 12 Alexandre 2015-09-02 07:44:26 UTC

I see. Unfortunately  I'm not able to identify the differences yet. I will hope for a interested volunteers.
Thanks.

Comment 13 Tim Allison 2015-09-02 12:19:02 UTC

Not that this is any consolation/help, but it looks like POI (at least via Tika) is able to read the contents of the embedded document, both from the outer container document and from the version that you attached as extracted by POI.