The TeXDocC Document Server Technical implementation and features

1
The TeXDocC Document Server-Technical
implementation andfeatures
2
What is the TeXDocC Document Server?

• Distinguish between
• TeX Document Center with its web resources
• TeX Document Center Document Server as a service
• Core component of the TeX Document Center project
• Inspired by the arXiv server (www.arxiv.org)
• slightly different design goals
• Main purpose e-print server
• validating documents
• archiving documents
• Tool for libraries to handle La(TeX) documents

3
Design Goals

• provide a validating and archiving service
• flexibility
• easy usage
• easy to integrate into a given library
  infrastructure
• open interfaces
• use of state-of-the-art technology
• be accessible by other services
• special access for librarians to check and store
  documents as a service for the users
• storing documents of different authors
• multi-delivery of documents
• All those requirements lead to

4
Fundamental decision

• Split the TeXDocC Document Server into two
  components
• Component 1 The archiving and validating service
• Checks, if the document can be compiled using
  some kind of standard environment
• Store the original compiling environment
• Discussion Store additionally some kind of XML
  conversion of the document
• Generate and store metadata
• Component 2 The Web Interface
• The standard web interface for the TeXDocC
  Document server
• Reference implementations to access the service
  via some kind of XML datastructures

5
How do we achieve this?

• Component 1 The service
• written in Perl
• Object-oriented design
• accessible using some XML formatted requests
• publicly available interface definitions
• Status
• Almost complete
• Additional work to make the delivery more robust
• Additional work to support more export formats
  (i.e. some XML, dvi, ps and so)
• Performance fine tuning

6
How do we achieve this?

• Component 2 The web interface
• accesses the service using XML
• presents the results using some kind of style
  sheet (CSS / XSL(T))
• Interface reference implementation
• Status
• Rudimentary, but fully functional
• Need some more optical tuning
• Optimization of the workflow to make the use more
  intuitive

7
Advantages and resulting features

• Usage of the service is independent from the
  TeXDocC web interface, thus
• easy and seamless integration into other services
• Easy to provide additional interfaces
  (OS-specific, or extensions to existing
  interfaces)
• Changes to the service (except to the protocol
  itself) dont interfere with the interface(s).
• Feature Providing RSS feeds based on MSC
  classification
• Register for your favorite MSC classifications
  and get your customized RSS feeds with
  HTML-rendered abstract
• Superior overview of new articles you may be
  interested in! Much more flexible in comparison
  to e-mail notifications
• Direct retrieval of the document from your
  favorite RSS feed reader

8
Advantages and resulting features

• Discussion Synchronize with other e-print
  servers
• Decentralized grids
• Distributed searches among other servers which
  implement a suitable protocol
• Distributed storage
• Discussion, Test SOAP Interface
• Standard interface for XML-RPC-Calls
• Easy integration into application which are SOAP
  capable
• Verify and publish directly out of your favorite
  (La)TeX editor (a short set of delivery scripts
  has already been developed and successfully
  tested on Mac OS X and an enhanced version of
  TeXShop)

9
The flexible metadata model of TeXDocC

• Automatic generation of metadata from prior user
  data and document specific data
• Reinsertion of the full metadata set into the
  (La)TeX document
• Customized metadata via special tags in the
  (La)TeX document
• Non-interactive delivery of (La)TeX documents
  possible due to integration of the metadata into
  the document itself
• Authors/Libraries can add extra metadata
• For example Own publication or revision numbers
• No collision with the standard metadata (the
  model uses internal and external or additional
  metadata)
• Special interfaces allow searching these metadata
  in own applications, too
• Libraries could outsource the storage of (La)TeX
  documents to the TeXDocC document server but
  could include specific metadata (i.e. search all
  documents delivered by a given library)
• Libraries could add their default metadata to the
  documents, too.

10
Short technical presentation
11

• Thank you!