datascript - npm Package Versions

0.13.0

• [ BREAKING ] Main namespace to include is now datascript.core, not datascript
• [ BREAKING ] Old datascript.core (internal namespace) was renamed to datascript.db
• [ BREAKING ] datascript.shim (internal namespace) was renamed to datascript.arrays

Motivations:

• Usage of top-level namespaces is discouraged and even generates a warning in CLJS.
• Better sooner than later.
• 0.13 seems like a great number for this sort of release.

Migration procedure:

• Just change (require '[datascript :as d]) to (require '[datascript.core :as d]) and you’re good to go.

For the sake of easy migration, there’re no other changes in this release. Just renamings.

0.12.2

• Fix null pointer exception when contains? is called with an entity (PR #114, thx Kevin Lynagh)

0.12.1

• db-init respects :db/index property

0.12.0

[ BREAKING ] Introducing new :db/index schema attribute:

• Attributes are not put to AVET by default anymore

• Following attributes are put to AVET:

  • :db/index true
  • :db/unique :db.unique/identity
  • :db/unique :db.unique/value
  • :db/valueType :db.type/ref

• All attributes put to AVET should be comparable. Note: maps and list are not comparable by default, vectors are compared value-by-value

• [ BREAKING ] Min/max aggregation functions will only work on a comaparable values

Benefits:

• You can finally store any trash easily and reliably in DataScript database: maps, vectors, functions, JS objects. It doesn’t even have to be comparable. Just do not mark it as :db/index and do not make it :db.cardinality/many
• Faster transactions. There’s ⅓ less indexes to fill
• Good defaults. Most cases where you’ll probably use AVET (lookup by value)—lookup refs, external ids, references—they are all indexed by default

Caveats:

• Your code may break (see below)
• Queries can do lookup by value event without :db/index attribute, but it’ll be slower

Migration procedure:

• Check your code for datoms, seek-datoms calls with :avet index, and index-range call
• Check your queries to see if they utilize AVET index. It happens when you use this pattern: [?free-var <constant-attr> <constant-value>]
• Mark necessary attributes with :db/index true in the schema

0.11.6

• msec argument to squuid (#95, #97)
• lookup refs in JS API pull, pull_many, entity (#94, thx Matt Senior)
• fix in Pull API for reverse non-component attributes (#91, thx Matt Senior)
• Node.js and Browser repls for dev profile (#93)
• Preconditions to validate db/conn arguments (#101)
• Id allocation bug (#66)

0.11.5

• Ported BTSet to JVM
• BTSet and BTSetIter implement ChunkedSeq
• New benchmark runner

This release brings a significant performance boost for JVM version of DataScript (numbers are in comparison to JS/v8 version):

• queries with single clause, no join: ~5–6 times faster
• queries with joins: ~3–4.5 times faster
• transact: ~3 times faster
• init-db: ~3–4 times faster
• rules: ~3-4 times faster

0.11.4

Experimental support for Clojure on JVM:

• sorted-set instead of BTSet
• polymorphic get lookup in query
• no sorting optimisation for init-db

Performance numbers so far (raw data):

• queries with single clause, no join: 30–50% faster
• queries with joins: ~50% slower
• transact: ~3.5–4 times faster
• init-db: ~2–2.5 times slower
• rules: ~2–5 times faster
• db equiv: ~5 times slower
• db hash: ~10 times slower

0.11.3

• Fixed UUID constructor warning under CLJS 0.0-3291 and later

0.11.2

• Support reverse attribute refs in combination with wildcards (issue #80)
• *.cljc sources do not conflict with Clojure now. Code compiles under [lein-cljsbuild "1.0.6"]

0.11.1

• Return nil from entity when passed nil eid (issue #75)