Puddle-corpus
Corpus representaion stored in JSON and wrapped into Corpus CRUD API
Features:
[X] Minimum dependencies
[X] .json <-> .corpus file conversion utility
[X] Pretty prints .json output to make it easier to track with git
[X] Uses STDIO and STDOUT as input and output
[X] Automatically determines type of input data (json or corpus)
[X] Corpus file may have comments starting with #
[X] Normalize corpus (i.e. cat corpus.dump | sort | uniq)
[X] Stores data as JSON
[X] Conforms to Corpus CRUD API
To be implemented:
[ ] Validate variable appearance constraints
[ ] Duplicate lines are removed.
[ ] No variable is DEFINE'd multiple times,
[ ] Every free variable is DEFINE'd once.
Installation:
npm install puddle-corpus
npm test # optional
Usage:
var corpus = require(‘puddle-corpus’)(corpus.json)
In corpus variable you will have Corpus CRUD API object which will use given file as a DB. Referr to puddle-crud module for API documentation.
JSON file format:
[{id: code, id2: code2, id3: code3 ...}]
.json<=>.corpus conversion utility
//Takes file on STDIO and outputs to STDOUT
//Input format is determined automatically by calling JSON.parse()
node convert < corpus/main.json > corpus/main.corpus
node convert < corpus/main.corpus > corpus/main.json
Corpus file format
The corpus file format is as follows:
- One line per statement,
- Statements are either ASSERT or DEFINE
- No trailing newline,
- Lines are sorted lexicographically,
In addition, each file should satisfy the constraints:
- Duplicate lines are removed.
- No variable is DEFINE'd multiple times,
- Every free variable is DEFINE'd once.
Contributors
- Fritz Obermeyer https://github.com/fritzo
- Yan T. https://github.com/yangit
Puddle was factored out of Pomagma in 2014.
License
Copyright 2013-2014 Fritz Obermeyer.
Puddle is licensed under the MIT license.