A simple indexer for leveldb.

Not ready for production use.

Usage

var index = indexer(dataDB, indexDB);

index.add('typeIndex', function(key, value) {
  return value.type;
});

db.put('0', {type: 'cookie', author: { name: 'cat' }}, function(err) {
  if(err) return console.error(err);

  index.get('typeIndex', 'cookie', function(err, key, value) {
    if(err) return console.error(err);
    console.log("key:", key, "value:", value);
  });
});

See examples/ for details.

If your index function is async, then call .add with the async option set:

index.add('foo', {async: true}, function(key, value, cb) {
  // do your async work here 
  cb(null, "index_key");
});

You can also add simple indexes on properties:

index.add('typeIndex', 'type');

Or even simpler if you want to name them the same as the property:

index.add('type');

Property paths are supported:

index.add('author.name');
// or
index.add('myIndex', 'author.name');

and you can create multiple property indexes in one go:

index.add(['type', 'author.index'])

API

.indexer(db, idb, [opts])

The constructor.

• db: The leveldb or sublevel containing your data.
• idb: The leveldb or sublevel to use for index storage.

Opts:

rebuild: false // force rebuild on init

.add(indexName, [opts], indexFunction)

Add a new index. This function is synchronous.

• indexName: Name of index
• indexFunction: Function that returns the index key

Opts:

 async: false // set to true if indexFunc uses a callback

The indexFunction will receive the arguments like so:

indexFunction(key, value, cb); // cb only present if .add called with {async: true}

For {async: false} the indexFunction must return the index key. For {async: true} the callback must be called with the index key as the second argument and potential errors or null/undefined as the first argument.

If no index should be created for the current db key/value pair then the index key returned by indexFunction must be either undefined or null.

.add(indexName, indexProperty)

TODO not implemented

Add a new index. This function is synchronous.

• indexName: Name of index
• indexProperty: Name of property to index by.

indexProperty can be a path, e.g. creator.name.

.add(indexProperty)

Add a new index. This function is synchronous.

• indexProperty: Name of property to index by.

indexProperty can be a path, e.g. creator.name.

The created index will use indexProperty as its name.

.add(indexPropertyArray)

TODO not implemented

Same as above but takes an array of property names or paths and creates indexes for all of them.

.get(indexName, indexKey, cb)

Use an index to fetch an item from the database.

• indexName: Name of the index to use.
• indexKey: The key in the index.

cb is called like cb(err, key, value) where key and value are the key and value from the actual db (not the index db).

.createReadStream(indexName, opts)

Like the levelup .createReadStream function this returns a stream of keys and values from the actual db but the order and range are based on the specified index.

.del(indexName, cb)

Delete an index. Like .clear but also makes index-level forget the index.

.clear(indexName, cb)

Clear an index. Deletes all of the index's data in the index db.

.clearAll(cb)

Run .clear for all indexes.

.build(indexName, cb)

Build an index from scratch.

Note: You will likely want to .clear the index first or call .rebuild instead.

.buildAll(cb)

Run .build for all indexes.

.rebuild(indexName, cb)

Clear and then build an index.

.rebuildAll(cb)

Call .rebuild for all indexes.

TODO

• Unit tests
• Add option to return a leveldown

License and copyright

License: AGPLv3

Copyright 2016 BioBricks Foundation