Format error in content.xml -- but unzip does not work

[meythland]

Hallo zusammen!

Euch, und allen, die Euch am Herzen liegen, ein gutes, gesundes und erfolgreiches 2011!

Ich habe mein problem bereits in einer anderen Community veröffentlicht, aber es kann mir offenbar niemand helfen. Ich nutze seit 2 Jahren OpenOffice, derzeit Version 3.2 auf Windows Vista, und habe nun mein erstes (aber massives) Problem mit Calc:

Beim Versuch, eine ods-Datei zu öffnen, bekomme ich die Fehlermeldung "Formatfehler im Teildokument content.xml an Position 1,0 (Zeile, Spale) in der Datei entdeckt". Wie ich beim Suchen gerade eben herausgefunden habe, ist dies ein gängiges Problem, welches man lösen kann durch

1. Umbenennen von ods in zip
2. Entpacken
3. Aufräumen und bearbeiten der content.xml an der angegebenen Stelle (meist Atrribute, die mehrmals vorkommen, entfernen)

Soweit, so gut, nur läßt sich meine in ".zip" umbenannte Datei weder mit 7-Zip noch mit TUGZip entpacken. Ersteres erklärt, die Datei sei kein Archiv, zweiteres zeigt beim Doppelklick auf die zip-Datei eine Datei namens "mimetype" und ein Verzeichnis namens "Configurations2", in welchem die beiden Verzeichnisse "statusbar" und "accelerator" zu finden sind. Ersteres dieser beiden ist leer, im zweiten findet sich die Datei "current.xml". Das war's! TGUZip zeigt mir keine weiteren Dateien an. Auspacken kann es die angezeigten Dateien allerdings auch nicht. Es sieht aus, als würde es entpacken, aber es passiert nichts -- ganz ohne Meldung.

Auffällig:
1. Ich habe unter den Dutzenden gefundenen Meldungen im Netz niemanden gefunden, bei dem der Fehler direkt am Anfang (also 1,0) war.
2. Beim letzten Speichern der Datei kam eine Fehlermeldung, bei der ich nur OK klicken konnte. Das habe ich geistesabwesend getan. Die kleine Diskette oben erlosch, sprich, es wurde gespeichert bzw. seit dem letzten Speichern wurden keine Änderungen durchgeführt, damit gab ich mich (abgelenkt) zufrieden und schloß das Programm -- bis heute.

Diese Datei ist eine Übersicht über meine privaten Finanzen. Sie ist mir sehr wichtig, enthält 14 Blätter, große Tabellen, viele Kalkulationen und Grafiken und ist zusätzlich noch liebevoll gestaltet. Über Weihnachten habe ich im Urlaub richtig viel Neues hinzugemacht und wollte die Datei heute, wie immer wöchentlich, sichern. Sprich, mir würden mindestens 16 Stunden an aufwändigen Änderungen verloren gehen!

Ich kann nur auf Eure Tipps hoffen. Bereits im voraus tausend Dank!

meythland

[Villeroy]

German forum: OOo: http://de.openoffice.info/

[meythland]

Thank you for the quick answer. I'll try in English then:

First of all, have a good and healthy 2011!

I am using OpenOffice since two years, currently version 3.2 on Windows Vista. While trying to open an ods file, I got an error message telling me there's a format error in content.xml at position 1,0. I found out this is a typical problem which can be solved easily by

1. renaming ods to zip
2. unzip
3. edit content.xml at the position mentioned (usually by erasing atrributes that show more than once in one line)

However, my renamed zip file does not unzip. I tried 7-Zip and TGUZip. The first one tells me the file is not an archive, the latter shows me the zip file contains a file named "mimetype" and a folder named "Configurations2", in which the two folders "statusbar" and "accelerator" can be found. The first one is empty, the letter contains a file named "current.xml". That's it! TGUZip does not show any other file. However, it also does not unzip the file. It looks like it does but it does not -- without any message at all.

Suspicious:
1. I found many usuers having the same problem but no one got the format error in content.xml at position 1,0
2. There was an error message when I last saved the file. I can not remember it since it was 2 weeks ago. I could only klick OK. It did not bother me since after saving the little disk icon greyed out, usually telling me there were no changes after last save (and it saved my changes).

This is a huge file which is immensly important for me. I am hoping someone can help me. Thank you very much in advance.
meythland

[Villeroy]

Is it password protected?

[meythland]

Heck, the calc file is, yes. Thank you.
What does that mean?
meythland

[Villeroy]

Heck, the calc file is, yes. Thank you.
What does that mean?
meythland

It means that all files within the archive are encrypted (apart from the mimetype file). Password protect any other document and watch the results in your zip tool. It should list all the encrypted xml and text files, but when you try to open the files the content will be scrambled due to the encryption.
I'm afraid your file is severely damaged since the encrypted files are not even listed which may explain the strange 1,0-location of the format error. There are repair tools for zip archives out there.
Whatever you try: always use backup copies and keep a copy of the defect original file.

[meythland]

Man!

DiskInternals' ZipRepair recoverd my zip file! Thank you so far. It now includes a content.xml. I downloaded Notepad++ and opened it -- but here my efforts stopped again due to the reason you described: its' content is scrambled.

So that's it? No chance to repair a password protected fiel this way? I will search for this since it is a completely new question now. I might ask it in a new thread since it's differnt than the current headline.

Anyway, thank you for your effort!
meythland

[Hagar Delest]

Well, keep using this topic, it's the same problem.

You need first to decipher the content of the file. Of course it cannot be accessed like that since it's password protected. Once it has been recovered, can OOo open it? If it can't then you've to find a way to decipher it from the zip manager. I don't know if it's feasible. Else, some googling should tell how it can be done.

Good luck.

[RoryOF]

It is worth trying to open the repaired zip file with OOo. OOo's password protection is effectively unbreakable, so if OOo can't open the file you may have a very difficult task.

[meythland]

Thank you guys!

I tried opening the recovered zip file with OOo. It asks me about the file type and offer 160 filters to open/convert. Unfortunately they all sound old to me (OpenOffice.org 1.0 document, ODF drawing, ODF formula, ODF diagramm, StarCalc, Star Writer, MathML and roughly 150 other). I tried a few (like OOo global document, various ODF types and rich text). Any suggestion of which filter to use?

Renaming it to ods before trying to open it gives the same error message (format error in content.xml at position 1,0). Trying to open content.xml gives scrambled chaos (in Writer, surprisingly).

Thanks again.
meythland

[Hagar Delest]

No filter will ever work. The file is completely unreadable since it's scrambled, it's just strings of characters. You have to find out what is the process applied by OOo to decipher the file.

Note that the repair tool may have thought it had repaired the file by changing some bits so that checksums are fine but it may have wrongly modified the file, hence still a problem for OOo. If the file had been really repaired, OOo should have detected the password protection and asked you for a password.

[meythland]

OK, tried something else and found out something new:

I created a completely new and empty Calc docment, renamed it to zip, unpacked it, replaced the "new" (and empty) content.xml of that file with my old, scrambled one, zipped it, renamed it to ods and tried to open it. Result is the same format error message.

I noticed that the unzipped new and empty file containes additional files like meta.xml, settings.xml and styles.xml which my f..ked-up one doesn't. Also, there are the folders META-INF and Thumbnails my old file does not have. On the opposite, the not.working one has a folder named ObjectReplacements which the new and empty one does not hava. Just in case this is useful.

Well, seems to be a looong evenening...
meythland

[meythland]

And another try: I password protected the new and empty ods file with the same password, followed the procedure above to exchange the new with the old scrambled content.xml, zipped and renamed it --- and OOo now asks me for a password! I cannot tell if this is a progress but at least it's something different. The downside is it tells me the PW is incorrect which is not true.

I tried the same thing again with not only exchanging the content.xml but also removing the meta.xml. Same result.

Just to let you know.
meythland

[Villeroy]

This is what I would read next: http://ringlord.com/dl/Decrypting%20ODF%20Files.pdf

[meythland]

OK. Wow. I did read this. Thank you!

Can it be that since my recoverd file does not contain a folder named META-INF, I can stop here and put my efforts into opening the last back-up version and invest the time to remember all my recent changes? It seems to me a password protected ods file needs this folder. And since mine doesn't, it only looks like the recovery worked. True?

meythland

[Villeroy]

No /META-INF/manifest.xml with all the seeds, salt and hashes
Bad luck, indeed.
Yes, I think you can give up at this point.

[meythland]

Yep, no META-INF, no manifest.xml, no seeds, no salt, no hashes.
In case nobody else tells me otherwise, I will stop.

In any case, thank you guys for your tremendous support.
Have a good 2011!

meythland

P.S. I will back-up my personal documents folder daily now. I will start looking for a tool that automates this tomorrow.