For both HUGO and OmniMark the first phase consisted of text matching rules. For the second phase:
For HUGO, it consisted of page layout descriptions.
For OmniMark, it consisted of a rule-based markup language parser, originally for SGML, later for XML and more recently for JSON.
| Beyond OmniMark | Sam Wilmott |
| Slide 5 of 37 | www.wilmott.ca |