Instructions for writing a custom analyzer

[adelevie]

The title says it all, but it would be really nice to have. I'm new to Clojure and Datomic, but I'm very interested in this project and its potential for analyzing semi-structured legal texts. Building an analyzer that could recognize statute sections as "units":

---

47 USC § 303b - Consideration of children’s television service in broadcast license renewal

• (a) After the standards required by section 303a of this title are in effect, the Commission shall, in its review of any application for renewal of a commercial or noncommercial television broadcast license, consider the extent to which the licensee
  • (1) has complied with such standards; and
  • (2) has served the educational and informational needs of children through the licensee’s overall programming, including programming specifically designed to serve such needs.
• (b) In addition to consideration of the licensee’s programming as required under subsection (a) of this section, the Commission may consider—
  • (1) any special nonbroadcast efforts by the licensee which enhance the educational and informational value of such programming to children; and
  • (2) any special efforts by the licensee to produce or support programming broadcast by another station in the licensee’s marketplace which is specifically designed to serve the educational and informational needs of children.

---

A proper analyzer would recognize 47 USC § 303b as something like a method signature and intelligently track changes inside the "method definition". Even finer-grained units would be especially useful--imagine being able to query for changes down to the subsection such as 47 USC § 303b(b)(2).