[
https://issues.apache.org/jira/browse/ARROW-2236?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16393009#comment-16393009
]
ASF GitHub Bot commented on ARROW-2236:
---------------------------------------
wesm closed pull request #1683: ARROW-2236: [JS] Add more complete set of
predicates
URL: https://github.com/apache/arrow/pull/1683
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/js/src/Arrow.externs.js b/js/src/Arrow.externs.js
index cf4db9134..be89be152 100644
--- a/js/src/Arrow.externs.js
+++ b/js/src/Arrow.externs.js
@@ -74,17 +74,24 @@ var custom = function () {};
var Value = function() {};
/** @type {?} */
-Value.prototype.gteq;
+Value.prototype.ge;
/** @type {?} */
-Value.prototype.lteq;
+Value.prototype.le;
/** @type {?} */
Value.prototype.eq;
+/** @type {?} */
+Value.prototype.lt;
+/** @type {?} */
+Value.prototype.gt;
+/** @type {?} */
+Value.prototype.ne;
var Col = function() {};
/** @type {?} */
Col.prototype.bind;
var Or = function() {};
var And = function() {};
+var Not = function() {};
var GTeq = function () {};
/** @type {?} */
GTeq.prototype.and;
@@ -108,6 +115,8 @@ Predicate.prototype.and;
/** @type {?} */
Predicate.prototype.or;
/** @type {?} */
+Predicate.prototype.not;
+/** @type {?} */
Predicate.prototype.ands;
var Literal = function() {};
@@ -209,6 +218,8 @@ Int128.prototype.plus
/** @type {?} */
Int128.prototype.hex
+var packBools = function() {};
+
var Type = function() {};
/** @type {?} */
Type.NONE = function() {};
diff --git a/js/src/Arrow.ts b/js/src/Arrow.ts
index 4a0a2ac6d..23e8b9983 100644
--- a/js/src/Arrow.ts
+++ b/js/src/Arrow.ts
@@ -18,7 +18,8 @@
import * as type_ from './type';
import * as data_ from './data';
import * as vector_ from './vector';
-import * as util_ from './util/int';
+import * as util_int_ from './util/int';
+import * as util_bit_ from './util/bit';
import * as visitor_ from './visitor';
import * as view_ from './vector/view';
import * as predicate_ from './predicate';
@@ -40,9 +41,10 @@ export { Table, DataFrame, NextFunc, BindFunc, CountByResult
};
export { Field, Schema, RecordBatch, Vector, Type };
export namespace util {
- export import Uint64 = util_.Uint64;
- export import Int64 = util_.Int64;
- export import Int128 = util_.Int128;
+ export import Uint64 = util_int_.Uint64;
+ export import Int64 = util_int_.Int64;
+ export import Int128 = util_int_.Int128;
+ export import packBools = util_bit_.packBools;
}
export namespace data {
@@ -173,6 +175,7 @@ export namespace predicate {
export import Or = predicate_.Or;
export import Col = predicate_.Col;
export import And = predicate_.And;
+ export import Not = predicate_.Not;
export import GTeq = predicate_.GTeq;
export import LTeq = predicate_.LTeq;
export import Value = predicate_.Value;
@@ -222,16 +225,16 @@ Table['empty'] = Table.empty;
Vector['create'] = Vector.create;
RecordBatch['from'] = RecordBatch.from;
-util_.Uint64['add'] = util_.Uint64.add;
-util_.Uint64['multiply'] = util_.Uint64.multiply;
+util_int_.Uint64['add'] = util_int_.Uint64.add;
+util_int_.Uint64['multiply'] = util_int_.Uint64.multiply;
-util_.Int64['add'] = util_.Int64.add;
-util_.Int64['multiply'] = util_.Int64.multiply;
-util_.Int64['fromString'] = util_.Int64.fromString;
+util_int_.Int64['add'] = util_int_.Int64.add;
+util_int_.Int64['multiply'] = util_int_.Int64.multiply;
+util_int_.Int64['fromString'] = util_int_.Int64.fromString;
-util_.Int128['add'] = util_.Int128.add;
-util_.Int128['multiply'] = util_.Int128.multiply;
-util_.Int128['fromString'] = util_.Int128.fromString;
+util_int_.Int128['add'] = util_int_.Int128.add;
+util_int_.Int128['multiply'] = util_int_.Int128.multiply;
+util_int_.Int128['fromString'] = util_int_.Int128.fromString;
data_.ChunkedData['computeOffsets'] = data_.ChunkedData.computeOffsets;
diff --git a/js/src/predicate.ts b/js/src/predicate.ts
index b177b4fa7..bff393863 100644
--- a/js/src/predicate.ts
+++ b/js/src/predicate.ts
@@ -26,14 +26,23 @@ export abstract class Value<T> {
if (!(other instanceof Value)) { other = new Literal(other); }
return new Equals(this, other);
}
- lteq(other: Value<T> | T): Predicate {
+ le(other: Value<T> | T): Predicate {
if (!(other instanceof Value)) { other = new Literal(other); }
return new LTeq(this, other);
}
- gteq(other: Value<T> | T): Predicate {
+ ge(other: Value<T> | T): Predicate {
if (!(other instanceof Value)) { other = new Literal(other); }
return new GTeq(this, other);
}
+ lt(other: Value<T> | T): Predicate {
+ return new Not(this.ge(other));
+ }
+ gt(other: Value<T> | T): Predicate {
+ return new Not(this.le(other));
+ }
+ ne(other: Value<T> | T): Predicate {
+ return new Not(this.eq(other));
+ }
}
export class Literal<T= any> extends Value<T> {
@@ -70,6 +79,7 @@ export abstract class Predicate {
abstract bind(batch: RecordBatch): PredicateFunc;
and(expr: Predicate): Predicate { return new And(this, expr); }
or(expr: Predicate): Predicate { return new Or(this, expr); }
+ not(): Predicate { return new Not(this); }
ands(): Predicate[] { return [this]; }
}
@@ -222,6 +232,17 @@ export class GTeq extends ComparisonPredicate {
}
}
+export class Not extends Predicate {
+ constructor(public readonly child: Predicate) {
+ super();
+ }
+
+ bind(batch: RecordBatch) {
+ const func = this.child.bind(batch);
+ return (idx: number, batch: RecordBatch) => !func(idx, batch);
+ }
+}
+
export class CustomPredicate extends Predicate {
constructor(private next: PredicateFunc, private bind_: (batch:
RecordBatch) => void) {
super();
diff --git a/js/test/unit/table-tests.ts b/js/test/unit/table-tests.ts
index 8a433815d..37bbf6242 100644
--- a/js/test/unit/table-tests.ts
+++ b/js/test/unit/table-tests.ts
@@ -327,27 +327,39 @@ describe(`Table`, () => {
const filter_tests = [
{
name: `filter on f32 >= 0`,
- filtered: table.filter(col('f32').gteq(0)),
+ filtered: table.filter(col('f32').ge(0)),
expected: values.filter((row) => row[F32] >= 0)
}, {
name: `filter on 0 <= f32`,
- filtered: table.filter(lit(0).lteq(col('f32'))),
+ filtered: table.filter(lit(0).le(col('f32'))),
expected: values.filter((row) => 0 <= row[F32])
}, {
name: `filter on i32 <= 0`,
- filtered: table.filter(col('i32').lteq(0)),
+ filtered: table.filter(col('i32').le(0)),
expected: values.filter((row) => row[I32] <= 0)
}, {
name: `filter on 0 >= i32`,
- filtered: table.filter(lit(0).gteq(col('i32'))),
+ filtered: table.filter(lit(0).ge(col('i32'))),
expected: values.filter((row) => 0 >= row[I32])
+ }, {
+ name: `filter on f32 < 0`,
+ filtered: table.filter(col('f32').lt(0)),
+ expected: values.filter((row) => row[F32] < 0)
+ }, {
+ name: `filter on i32 > 1 (empty)`,
+ filtered: table.filter(col('i32').gt(0)),
+ expected: values.filter((row) => row[I32] > 0)
}, {
name: `filter on f32 <= -.25 || f3 >= .25`,
- filtered:
table.filter(col('f32').lteq(-.25).or(col('f32').gteq(.25))),
+ filtered:
table.filter(col('f32').le(-.25).or(col('f32').ge(.25))),
expected: values.filter((row) => row[F32] <= -.25 ||
row[F32] >= .25)
+ }, {
+ name: `filter on !(f32 <= -.25 || f3 >= .25) (not)`,
+ filtered:
table.filter(col('f32').le(-.25).or(col('f32').ge(.25)).not()),
+ expected: values.filter((row) => !(row[F32] <= -.25 ||
row[F32] >= .25))
}, {
name: `filter method combines predicates (f32 >= 0 &&
i32 <= 0)`,
- filtered:
table.filter(col('i32').lteq(0)).filter(col('f32').gteq(0)),
+ filtered:
table.filter(col('i32').le(0)).filter(col('f32').ge(0)),
expected: values.filter((row) => row[I32] <= 0 && row[F32]
>= 0)
}, {
name: `filter on dictionary == 'a'`,
@@ -357,13 +369,17 @@ describe(`Table`, () => {
name: `filter on 'a' == dictionary (commutativity)`,
filtered: table.filter(lit('a').eq(col('dictionary'))),
expected: values.filter((row) => row[DICT] === 'a')
+ }, {
+ name: `filter on dictionary != 'b'`,
+ filtered: table.filter(col('dictionary').ne('b')),
+ expected: values.filter((row) => row[DICT] !== 'b')
}, {
name: `filter on f32 >= i32`,
- filtered: table.filter(col('f32').gteq(col('i32'))),
+ filtered: table.filter(col('f32').ge(col('i32'))),
expected: values.filter((row) => row[F32] >= row[I32])
}, {
name: `filter on f32 <= i32`,
- filtered: table.filter(col('f32').lteq(col('i32'))),
+ filtered: table.filter(col('f32').le(col('i32'))),
expected: values.filter((row) => row[F32] <= row[I32])
}, {
name: `filter on f32*i32 > 0 (custom predicate)`,
@@ -455,17 +471,17 @@ describe(`Table`, () => {
expect(selected.toString()).toEqual(expected);
});
test(`table.filter(..).count() on always false predicates returns
0`, () => {
- expect(table.filter(col('i32').gteq(100)).count()).toEqual(0);
+ expect(table.filter(col('i32').ge(100)).count()).toEqual(0);
expect(table.filter(col('dictionary').eq('z')).count()).toEqual(0);
});
describe(`lit-lit comparison`, () => {
test(`always-false count() returns 0`, () => {
expect(table.filter(lit('abc').eq('def')).count()).toEqual(0);
- expect(table.filter(lit(0).gteq(1)).count()).toEqual(0);
+ expect(table.filter(lit(0).ge(1)).count()).toEqual(0);
});
test(`always-true count() returns length`, () => {
expect(table.filter(lit('abc').eq('abc')).count()).toEqual(table.length);
-
expect(table.filter(lit(-100).lteq(0)).count()).toEqual(table.length);
+
expect(table.filter(lit(-100).le(0)).count()).toEqual(table.length);
});
});
describe(`col-col comparison`, () => {
diff --git a/js/test/unit/vector-tests.ts b/js/test/unit/vector-tests.ts
index 3eb3fbe01..d25e0e9fd 100644
--- a/js/test/unit/vector-tests.ts
+++ b/js/test/unit/vector-tests.ts
@@ -17,13 +17,13 @@
import { TextEncoder } from 'text-encoding-utf-8';
import Arrow from '../Arrow';
-import { type, TypedArray, TypedArrayConstructor, Vector } from
'../../src/Arrow';
-import { packBools } from '../../src/util/bit'
+import { TypedArray, TypedArrayConstructor } from '../../src/Arrow';
const utf8Encoder = new TextEncoder('utf-8');
+const { packBools } = Arrow.util;
const { BoolData, FlatData, FlatListData, DictionaryData } = Arrow.data;
-const { IntVector, FloatVector, BoolVector, Utf8Vector, DictionaryVector } =
Arrow.vector;
+const { Vector, IntVector, FloatVector, BoolVector, Utf8Vector,
DictionaryVector } = Arrow.vector;
const {
Dictionary, Utf8, Bool,
Float16, Float32, Float64,
@@ -143,7 +143,7 @@ describe('Float16Vector', () => {
const values = concatTyped(Uint16Array, ...bytes);
const vector = bytes
.map((b) => new Uint16Array(b.buffer))
- .map((b) => new FloatVector<type.Float16>(new FlatData(new Float16(),
b.length, null, b)))
+ .map((b) => new FloatVector<Float16>(new FlatData(new Float16(),
b.length, null, b)))
.reduce((v: any, v2) => v.concat(v2));
const n = values.length;
const clamp = (x: number) => (x - 32767) / 32767;
@@ -336,7 +336,7 @@ describe(`DictionaryVector`, () => {
describe(`sliced`, () => {
basicVectorTests(vector.slice(10, 20), values.slice(10,20),
extras);
- })
+ });
});
describe(`index with nullCount > 0`, () => {
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> [JS] Add more complete set of predicates
> ----------------------------------------
>
> Key: ARROW-2236
> URL: https://issues.apache.org/jira/browse/ARROW-2236
> Project: Apache Arrow
> Issue Type: Task
> Components: JavaScript
> Reporter: Brian Hulette
> Assignee: Brian Hulette
> Priority: Major
> Labels: pull-request-available
> Fix For: JS-0.4.0
>
>
> Right now {{arrow.predicate}} only supports ==, >=, <=, &&, and ||
> We should also support !=, <, > at the very least
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
