Corpus Protocol
If you want me to run your rules, this is the shape I need them in. If it matches this, it runs. The whole template is at the bottom of this page in one block. Copy it and paste it into a spreadsheet.
1 · Your rules, one row per record
Each row is one thing your world cites: a section, a control, an article. Every row has the same six columns.
| field | what goes in it | example |
|---|---|---|
| citation | The record exactly as people in your field cite it. No two records share one. | Sec. 181.001 |
| cites_group | The citation without the record's own number. | Sec. 181 |
| number | The record's own number, as text. | 001 |
| container | The part or chapter it sits in. | Chapter 181 |
| heading | Its heading. | Definitions |
| body | The full text of the record. | In this chapter: ... |
2 · About your rules
Fill in the right-hand column.
| field | what I need | your answer |
|---|---|---|
| name | What the rules are called | |
| jurisdiction | Where they apply | |
| record_level | What one record is (section, control, article) | |
| a_citation_reads | The one pattern, e.g. {cites_group}.{number} | |
| published_at | Where the official text lives | |
| held_on_the_basis_of | Your right to the text (public, licensed, your own) | |
| publisher_says_there_are | How many records the publisher counts | |
| publisher_count_came_from | Where that count comes from | |
| absent_citations | Two or more citations in the right shape that do not exist | |
| records_you_left_out | How many you left out, and why | |
| records_you_left_out_are | Their citations |
3 · Your questions, and your answers
Write the questions your people actually ask. Twenty is enough; fifty is better. Include some whose answer is not in your rules, so you can see it decline. You write the answers. I don't.
| question | should_it_answer | answer_is_in | note |
|---|---|---|---|
| How many days does a provider have to supply a record after a written request? | yes | Sec. 181.102 | |
| What are the rules on public urination? | no | not in these rules |
4 · Getting it to me
Email me first. I'll send you the standard agreements. Once they're signed, you send the spreadsheet, I run it through the engine, and you get back every answer exactly as it came out, fingerprinted, with its source.
Luke Thorburn
Luke.C.Thorburn@gmail.com / Luke@PennoyerSystems.com
214-789-0990
5 · The whole template
Copy the block and paste it into a spreadsheet.
| RULES FOR WHOEVER FILLS THIS IN | |||||
| 1. One row per record: one section, control or article that people cite. | |||||
| 2. Format the number column as Text before typing, so 001 stays 001 and 56.10 stays 56.10. | |||||
| 3. citation = cites_group + number, the same way on every row. AC-2(1) is AC + 2(1), not AC-2 + 1. | |||||
| 4. No two rows share a citation. citation, cites_group, number and body are never empty. | |||||
| 5. Questions: ask by content and put the citation in answer_is_in, never in the question. Include some whose answer is not in your rules (should_it_answer = no). | |||||
| 6. Save as a spreadsheet. Email Luke@PennoyerSystems.com first; we agree terms, then you send it. | |||||
| YOUR RULES (one row per record) | |||||
| citation | cites_group | number | container | heading | body |
| Sec. 181.001 | Sec. 181 | 001 | Chapter 181 | Definitions | In this chapter: ... |
| ABOUT YOUR RULES | |||||
| field | your answer | ||||
| name | |||||
| jurisdiction | |||||
| record_level | |||||
| a_citation_reads | |||||
| published_at | |||||
| held_on_the_basis_of | |||||
| publisher_says_there_are | |||||
| publisher_count_came_from | |||||
| absent_citations | |||||
| records_you_left_out | |||||
| records_you_left_out_are | |||||
| YOUR QUESTIONS | |||||
| question | should_it_answer | answer_is_in | note | ||
| How many days does a provider have to supply a record after a written request? | yes | Sec. 181.102 | |||
| What are the rules on public urination? | no | not in these rules | |||