Document Repository

Hi, I'm interested in making part of our TWiki into a document repository. I want a user to be able to upload a PDF file and enter some attributes of the document into an associated form (eg: author, title, etc) for searching later. Then, of course, they should also be able to search for documents based on author, title, etc. Before I start trying to write a plugin to do this.. am I missing something that's already in the TWiki or a plugin that someone has already written? The Know web is close but I don't see a way to attach a file when creating a topic. Also don't see a way to create/edit a topic without seeing the big textarea.. i just want the form and an attach option. So should I start writing my plugin now?

-- RobDuarte - 30 Jul 2002

I can't answer your question myself, and I don't want to dissuade anybody from writing a potentially useful plugin for Twiki, but I do see a potential usbility problem here. It would be much better to use a search engine[0] or some other app which indexes and grabs the author,etc. from PDF files themselves. People don't like filling in this kind of metadata and have a tendency to skip whatever they can.

0. Google Appliance, Namazu, Perlfect, and probably others. Related Twiki pages: SearchEngineVsGrepSearch, SearchAttachments

-- MattWilkie - 31 Jul 2002

you probably want to use the SearchEnginePluceneAddOn

-- WillNorris - 16 Jul 2006

(after 4 years!?)

No way that a search engine will create structured data from unstructured PDFs. I would have a look at a couple of TWiki applications in the Sandbox web that offer simple fill-in forms:

• BookClub
• SearchBookTitles
• AnnotatedBibliography

-- ArthurClemens - 17 Jul 2006

why do you assume PDF files are unstructured? they can have tagged metadata

-- WillNorris - 17 Jul 2006

Given TWiki does not mail interested parties when someone replies someone ought update the question-asker manually.

-- MartinCleaver - 17 Jul 2006

TWiki could also run a metadata extractor when an attachment is uploaded. This would copy the meta data from the attachment (e.g. PDF or word Properties fields) and put them into the META field.

-- MartinCleaver - 17 Jul 2006