Full text search
Parallel processing model
Prerequisite Concepts		Related Syntax

OmniMark has the capability to run processes in parallel. Such processes run as co-routines, not threads. Parallel processing occurs whenever a program is generating markup which is then fed to the markup processor. This occurs in up-translate and context-translate programs, in programs that do subdocument processing, and in programs that use an input function to supply input to the markup processor in a do xml-parse and do sgml-parse.

Running two processes in parallel has three principal advantages:

• It eliminates the need to buffer generated markup before processing it with the markup processor.
• It allows the markup processor to report errors in generated markup as that markup is generated, allowing you to use the markup processor to simplify error detection and correction in programs which generate markup.
• It allows you to use the current state of the parse tree (as expressed in the execution state of the markup processor) to guide markup generation, allowing you to do context-sensitive markup building.

For the most part, you do not need to concern yourself with the mechanics of parallel processing. OmniMark handles all the details for you. However, there are some important restrictions you need to be aware of:

• Each domain has its own groups.
• save and save-clear operations are specific to the domain in which they occur.
• A process which is supplying input to the markup processor cannot itself invoke the markup processor. Thus you cannot use do xml-parse and do sgml-parse in an input function or in the find rules of an up-translate or context-translate program (unless those rules are processing the results of a submit action initiated in a markup rule, and therefore acting on the output of the markup processor rather than the input).
• OmniMark runs parallel processes in separate processing domains. Stream variables belong to the domain in which they are opened. If you want a process in one domain to be able to write to a stream opened in another domain, you need to open that stream with the domain-free modifier. Note that it may not always be obvious which domain a certain piece of code is running in. For instance, a find rule that could be fired either by the main input of a cross-translate or up-translate program, or by input from a submit action in a markup rule would be in one domain in the first instance and in another domain in the second instance.
• using nested-referents can only be used in one domain at a time.

Prerequisite Concepts      Pattern and markup processors

Related Syntax    context-translate    do sgml-parse    do xml-parse    domain-free    up-translate

----

[CONTENTS] [CONCEPTS] [SYNTAX] [LIBRARIES] [SAMPLES] [ERRORS] [INDEX]