Skip to content

Latest commit

 

History

History
143 lines (114 loc) · 5.95 KB

extraction.md

File metadata and controls

143 lines (114 loc) · 5.95 KB
Raw

extraction

This processor encapsulates the extraction phase of ts-transforms.

To use this processor, it is required to use the selection processor as this works by using the metadata that is set in the selection processor. It also requires a file containing all transform rules that is will be preforming and any necessary plugins that those rules require to operate.

The full functionality of ts-transforms is encapsulated in the transform processor.

You would only want to use this processor instead of the fully functional transform processor if you wanted to inject a custom operation between each phase, or to capture the analytics of the records between each phase.

Usage

Example of the rules and plugins to run the extraction phase

Below contains an example of the transform rules, plugins and configurations to perform the extraction phase. Note that the post-process and output phases of ts-transforms are not executed.

Example Job

{
    "name" : "testing",
    "workers" : 1,
    "slicers" : 1,
    "lifecycle" : "once",
    "assets" : [
        "standard"
    ],
    "operations" : [
        {
            "_op": "test-reader"
        },
        {
            "_op": "selection",
            "plugins": ["someAssetId:plugins"],
            "rules": ["someAssetId:transformRules.txt"],
            "type_config": {
                "some": "string",
                "field": "string",
                "field2": "string",
                "date": "date",
                "location": "geo-point"
            },
            "variables": {
                "foo": "data"
            }
        },
        {
            "_op": "extraction",
            "plugins": ["someAssetId:plugins"],
            "rules": ["someAssetId:transformRules.txt"],
            "type_config":  {
                "some": "string",
                "field": "string",
                "field2": "string",
                "date": "date",
                "location": "geo-point"
            },
            "variables": {
                "foo": "data"
            }
        }
    ]
}

Example rules located at someAssetId:transformRules.txt

{ "selector": "some: $foo", "source_field": "field", "target_field": "interm1", "tag": "someTag", "output": false }
{ "selector": "some: $foo", "source_field": "field2", "target_field": "interm2", "tag": "someTag", "output": false }
{ "follow": "someTag", "post_process": "join", "target_field": "final", "delimiter": " " }

{ "source_field": "otherField", "target_field": "lastField", "other_match_required": true }
{ "source_field": "_key", "target_field": "id", "other_match_required": true, "tag": "numberMe" }
{ "follow": "numberMe", "post_process": "number" }

{ "selector": "location: geoBox( top_left: '33.906320, -112.758421' bottom_right: '32.813646,-111.058902')", "source_field": "location", "target_field": "loc" }

{ "selector": "date:[2019-04-16T20:14:44.304Z TO *] AND bytes:>=1000000", "source_field": "date", "target_field": "last_seen", "tag": "tagOp" }
{ "follow": "tagOp", "post_process": "tag" }

Example plugin used by rules above located at someAssetId:plugins

import { DataEntity } from '@terascope/job-components';

class Tag {
    static cardinality = 'one-to-one';

    constructor(operationConfig) {
        this.operationConfig = operationConfig;
    }

    run(doc: DataEntity) {
        doc.wasTagged = true;
        return doc;
    }
}

export default class Plugin {
    init() {
        return {
            tag: Tag,
        };
    }
}

Example of the data and the expected results of the extraction phase, the metadata is what is set by the selection processor

const data = [
    DataEntity.make(
        { some: 'data', field: 'oneThing', field2: 'something' },
        { selectors: ['some: $foo', '*'] }
    ),
    DataEntity.make(
        { location: '33.242, -111.453' },
        { selectors: ["location: geoBox( top_left: '33.906320, -112.758421' bottom_right: '32.813646,-111.058902')", '*'] }
    )
];

const results = await processor.run(data);

results === [
    { interm1: 'oneThing', interm2: 'something' },
    { loc: '33.242, -111.453' }
];

Parameters

Configuration Description Type Notes
_op Name of operation, it must reflect the exact name of the file String required
rules an array of strings that are the locations where rule files. must be specified in "assetName:path" format String[] required
plugins an array of strings that are the locations where plugins reside. must be specified in "assetName:modulePath" format Object[] optional, defaults to []
type_config a schema for the data being consumed. Set the keys to your data field names, with values set to this enum Object optional
variables An object containing any variables for the xlucene rules Object optional