That much I can do, at least for version one of the open source engine. I can do this right here, in this post and get it out of the way. But it might be informal. Remember when I write this, the universe of G is made of maybe, millions of nested ontology graphs, in SQL tables. Where do these millions of G come from? Mainly sel convolutions of G on itself. G trying to coax new nodes of raw test into ontology form and generating Gs in abundance. Typically, about 5% of the Gs around the company will be human designed, mainly to manage dictionary tables and text reduction. So the Ontology to SQL interface is about grabbing a G and executing its operators with SQL prebuilt routines. OK, I can do that.
What are the main APIs?
Link, when one G jumps to another, change the table name parameter.
Exit, the convolution is done, insert Gout onto a table store. The table name is the first node of G3.
Methods to help traversal:
OR select, with or without match, step or block mode
AND select, same as above.
For the default grammar, there is not much else. This is version 1.0, and I assume all table stores are local.
No comments:
Post a Comment