Document Management

Abstract

This document tries to explore how to build a distributed document management system using Web Standards.

We envision a system where all the meta-information is stored within the HTML as structured RDFa elements, allowing easy navigation with automatic reference and glossary management.

Document structure

For each document, multiple HTML files are provided:

• the main document which includes all the document content;
• an index file that just contains the abstract plus meta-information (e.g. author, title, keywords);
• an optional summary page which imports the index file of the main document and of all referenced documents to provide a nice overview page.

The index file is not meant to be viewed directly by browsers, it just consolidates all the meta-information about the document and is imported by both the main document and the summary page. This index file is the key to be able to relate between documents even when they are managed by different organizations.

Detailed design

The index file is an HTML file containing RDFa meta-data. It should only have minimal external dependencies so that it can be loaded quickly. This is important because a large number of index files may have to be loaded to generate information about referenced documents.

The index file defines as an RDF resource using the [https://www.w3.org/TR/vocab-dcat/ Data Catalog Vocabulary] The document itself should be represented as one dcat:Dataset ressource.

The following information is recommended to be provided for the document:

• Author(s) (dcterms:creator);
• Title(s) (dcterms:title);
• Keywords (dcterms:subject);
• Table of contents (dcterms:tableOfContents);
• Abstract of the document (dcterms:abstract);
• Link to the summary page (dcat:landingPage).

Styling the document

Provide a document template as HTML-import. Customizable properties can be provided via CSS custom variables.

The template should already provide both the template for print as well as a version for interactive use.

Styling the summary

TBD

References to other documents

By importing the .index.html of all referenced documents, it is possible to include meta-information about these documents. (E.g. show a list of references, show hover information / footnotes with a summary of referenced items.)

By not importing the complete referenced document (with its own further links), we can prevent the client from fetching unneeded (and potentially very large) documents.

Each document reference is represented as on import. The name is used as identifier when referencing the reference. TBD: defaults

References to specific items

References to other documents are only allowed when there is already a document-level reference.

Backlinks from other documents

Requires an external index which is obviously dependent on the other documents. This index would have to be managed by some document management system.

Figures

TBD

Interaction with Paged Media

The @page CSS tag only works in the toplevel, not within a web component. Basic page setup for the selected document template has to be done as <style> within a HTML-import.

TBD

Specification of index files

TBD

Relation to HtmlBook

TBD