The parse pipeline would take user input as typed on topics text entries and meta data entry panes, perform any required validation, enrichment or processing of the input text and then, if all is well, pass the resulting Topic, MetaData and attachments down to the StorageSubsystem.

Implementation Specific Notes

BeijingRelease Implementation

I'm not sure, the parse pipeline could be said to validate the forms, but it does not enrich or further process the input text.

Apart from this the parse pipeline does not exist in Beijing

CairoRelease Implementation

DakarRelease Implementation

Types of Enrichment

Validation

For example, invoking preview for NonExistantWikiWordReferences

Enrichment

Enrichment would mean TWiki adding to or changing what has been written in the topic text before it gets saved. If this happens Preview should normally be invoked to show the user the enrichment. Examples of enrichment include:

• Automatic Spelling corrections
• Adding fields to a table where values can be derived but should be stored for other purposes. e.g.
  • Deriving Address Information from postcode or Map reference information from longitude/latitude reading.
• The on-the-fly subdefinition of another WikiWord. e.g. in a document describing WordA you could write "See also WordB %DEFINE{WordB: here is a definition for WordB}%". When WordA is saved the DEFINE directive means the 'here is a definition for WordB' does not get put into TopicA, instead that is saved in the WordB topic as if it had been created separately. WordA just contains "See also WordB"

Processing

Processing would mean recognising WikiControlLanguage commands written into the topic text and executing those rather than saving the text directly.

-- MartinCleaver - 15 Sep 2002, 21 Apr 2003