Hyperpiler
An improved method for generating complex formal language documents from simple input values, reducing the barrier to formal communications. The techniques described may be applied to different domains to generate different types of documents requiring formal language. For illustration, this disclosure focuses on generating a computer program document as programming languages are among the most formal.
1. A computer-implemented method for generating a document within a targeted domain, the method comprising:
accessing a specification for the document, the specification comprising a plurality of specification entities that comprise values to be included in the document;
applying a domain ruleset to convert the plurality of specification entities to a plurality of segment entities that comprise the values from the specification entities, wherein the domain ruleset comprises predefined rules for processing specification entities for documents within the targeted domain;
processing the plurality of segment entities to populate corresponding predefined segments from a segment set by inserting values from the segment entities into the predefined segments; wherein the predefined segments in the segment set comprise fragments of a document within the targeted domain and further comprise value insertion points for the insertion of values into the predefined segments; and
assembling the segments according to the ruleset, thereby generating the document.
2. The computer-implemented method of claim 1 wherein at least some of the segment entities each comprise values from two different specification entities.
3. The computer-implemented method of claim 1 wherein assembling the segments comprises arranging at least some of the segments according to values in the corresponding segment entities.
4. The computer-implemented method of claim 3 wherein at least some values in the corresponding segment entities used for arranging the segments are determined at least in part by values in the specification entities.
5. The computer-implemented method of claim 3 wherein at least some of the segment entities comprise values assigning parent-child relationships among the segment entities; and assembling the segments comprises recursively nesting child segments into value insertion points in parent segments according to the parent-child relationships in the corresponding segment entities.
6. The computer-implemented method of claim 1 wherein the targeted domain is a computer program in a computer programming language, the predefined segments comprise program instructions in the computer programming language, and the generated document is a computer program in the computer programming language.
7. The computer-implemented method of claim 6 further comprising: compiling the generated computer program.
8. The computer-implemented method of claim 1 wherein the targeted domain is a legal document, the predefined segments comprise legal language, and the generated document is a legal document.
9. A computer system for generating a document within a targeted domain, the computer system comprising:
a data storage device that stores:
a specification for the document, the specification comprising a plurality of specification entities that comprise values to be included in the document;
a domain ruleset comprising predefined rules for processing specification entities for documents within the targeted domain; and
a segment set comprising predefined segments, the predefined segments comprising fragments of a document within the targeted domain and further comprising value insertion points for the insertion of values into the segment; and
a processor executing program instructions configured to:
apply the domain ruleset to convert the plurality of specification entities to a plurality of segment entities that comprise the values from the specification entities;
process the plurality of segment entities to populate the segments by inserting values from the segment entities; and
assemble the segments according to the ruleset, thereby generating the document.
10. A non-transitory computer-readable storage medium storing executable computer program instructions for generating a document within a targeted domain, the instructions executable by a computer system and causing the computer system to perform a method comprising:
accessing a specification for the document, the specification comprising a plurality of specification entities that comprise values to be included in the document;
applying a domain ruleset to convert the plurality of specification entities to a plurality of segment entities that comprise the values from the specification entities, wherein the domain ruleset comprises predefined rules for processing specification entities for documents within the targeted domain;
processing the plurality of segment entities to populate the segments by inserting values from the segment entities; wherein the predefined segments in the domain segment set comprise fragments of documents within the targeted domain and further comprise value insertion points for the insertion of values into the segment; and
assembling the segments according to the ruleset, thereby generating the document.