I was referring to metadata without having a technical view on this. I understand from a tika-developer point of view it's not possible to offer metadata that might be stored in proprietary/system dependent stores. Information like last write time or creator or even creation time are mostly stored by the filesystem, not the document.
Java 7's BasicFileAttributes would be a very nice improvement for tika! Kai-Uwe -----Ursprüngliche Nachricht----- Von: Allison, Timothy B. [mailto:[email protected]] Gesendet: Donnerstag, 10. Juli 2014 12:58 An: [email protected] Betreff: RE: Metadata at e.g. textfiles Ditto what Nick said on "internal" metadata. Are you referring to "external" metadata that we could get in Java 7 via BasicFileAttributes on OS's that support those? -----Original Message----- From: Nick Burch [mailto:[email protected]] Sent: Thursday, July 10, 2014 6:52 AM To: [email protected] Subject: Re: Metadata at e.g. textfiles On Thu, 10 Jul 2014, Kai-Uwe Schmidt wrote: > is there a way to get e.g. creator and or creation date into the > metadata dictionary? Only for file formats which store this information > When I extract from a file I get the following: > > Content-Encoding: ISO-8859-1 > Content-Length: 9 > Content-Type: text/plain; charset=ISO-8859-1 > resourceName: TikaTestText.txt Plain text files don't include creator or creation dates in them. If you try with a format that does, eg PDF or Word, you should see them reported Nick
