Skip to content

Latest commit

 

History

History
44 lines (27 loc) · 1.69 KB

README.md

File metadata and controls

44 lines (27 loc) · 1.69 KB
Raw

WBW GSheets Crawler

Crawler engine to ingest WBW gsheets into typesense-server for real-time search.

INSTALLATION

You need typesense key with write access to ingest data into server.

cp .env.example .env
yarn install

Modify TYPESENSE_HOST, TYPESENSE_PORT, TYPESENSE_PROTOCOL and TYPESENSE_KEY afterwards. Then you're good to go.

WRITING CRAWLER

IMPORTANT:
Google sheets must be published to web in order to be crawled

Crawler will read all scripts in metadata directory to intepret sheet structure. Each script represent an index and must contains:

  • schema : typesense schema object. See here for reference.
  • sheetId : a public google-sheets ID i.e.
    https://docs.google.com/spreadsheets/u/1/d/<SHEET_ID>/view
  • indexId : typesense's index name. Must have wbw- prefix.
  • worksheet : List of worksheets in given gsheets

A field named order must be defined manually in the metadata with data int32 data type as sortable field

Every data row, id and sheet fields will be added to mark which worksheet it's originated.

Index will be made automatically when it's not present. To prevent server rejection, crawling process will be executed sequentially (not in parallel).

DEVELOPMENT NOTES

We have two flags to make development easier:

  1. --test_script=SCRIPTNAME will only execute given script in the metadata/ directory
  2. --dry_run to run as dry run mode / not inserting data into typesense

CONTRIBUTING GUIDELINES

Please refers to wargabantuwarga contributing guidelines