bloom-lang / bloom-js

Ben Sklaroff's JS implementation, for posterity

Bloom Runtime

Stable Bloom runtime in JavaScript is in node_modules/bloom-runtime-stable. This runs hand-written, JavaScript-syntax Bloom.

Usage:

Include var Bloom = require('bloom-runtime-stable'); at the top of your file. Then write JavaScript-syntax Bloom code as in the examples.

Examples:

$ node examples/bloom_paths_stable.js
$ node examples/decision_tree/decision_tree_stable.js

Experimental Bloom Runtime

Experimental Bloom runtime in JavaScript is in node_modules/bloom-runtime. This runs files generated by the experimental bloom compiler.

Usage:

Include var Bloom = require('bloom-runtime'); at the top of your file. But you should really be generating your target file with the compiler.

Example:

$ node examples/bloom_paths.js

Experiemental Bloom Compiler

Experimental Bloom compiler in JavaScript is in bloom_compiler. This compiles an input file to either JavaScript, to be run with the experimental runtime, or procedural PostgreSQL (PL/pgSQL).

Usage:

Use the script compile_bloom.sh to generate code. To generate PL/pgSQL instead of javascript, use the optional --sql command-line argument.

Examples:

$ ./compile_bloom.sh examples/paths.rb paths.js
$ node paths.js


$ ./compile_bloom.sh --sql examples/paths.rb paths.sql
$ plpgsql paths.sql

Current Issues:

• Compiler only handles pairs, reduce, group, and argmin operations. The rest need to be implemented
• Compiler only handles instantaneous (<=) and deferred (<+-) operators. The rest need to be implemented
• examples/decision_tree/typed_dtree.rb relies on unimplemented features