Lark Sheets connector

The Lark Sheets connector allows reading Lark Sheets spreadsheets in Presto.

Lark App

The connector requires a Lark app in order to access the Lark Sheets API.

1. Create a Lark app by clicking the Create Custom App button at Lark Open Platform (for global users) or Feishu Open Platform (for Chinese users). Choose Custom App when creating the app. Get and remember app id and app secret from Credentials and Basic Info page.

2. Grant View, comment, and export Sheets permission to the app at Permission & Scopes page. And publish a new release for the app at App Release page to make the permissions take effect.

Configuration

Create etc/catalog/larksheets.properties to mount the Lark Sheets connector as the larksheets catalog, replacing the properties as appropriate:

connector.name=lark-sheets
app-domain=FEISHU
app-id=example_app_id
app-secret-file=/path/to/app-secret.json

Create app-secret.json as the example below:

{"app-secret": "abcdefghijklmnopqrstuvwxyzabcdef"}

The following configuration properties are available:

Query Model

In the connector, a Lark spreadsheet is mapped to a Presto schema, and each sheet inside the spreadsheet is mapped to a Presto table.

Create a spreadsheet-schema mapping by executing a query like:

CREATE SCHEMA my_ss WITH (TOKEN = 'shtcnBf5pg4BNSkwV2Ku5xwW9Pf')

The token property is taken from the last path segment of the spreadsheet url, using the pattern https://{lark_site_domain}/sheets/{token}?sheet={sheet_id}. For example, shtcnBf5pg4BNSkwV2Ku5xwW9Pf is the token of the spreadsheet at https://test-ch80md45anra.feishu.cn/sheets/shtcnBf5pg4BNSkwV2Ku5xwW9Pf?sheet=MT1p4I

After that, all sheets (tabs) of the spreadsheet are automatically mapped to tables in the schema.

SHOW TABLES FROM my_ss

For each table, its name is taken from the sheet title; its columns are inferred from the head row (the first row of the sheet). Sheets metadata could be listed through the query as:

SELECT * FROM my_ss."$sheets"

In a query statement, a table can be specified in such ways:

-- by sheet title
SELECT * FROM my_ss."number_text"
-- by sheet index
SELECT * FROM my_ss."$0"
-- by sheet id
SELECT * FROM my_ss."@MT1p4I"

As for column infer, columns that have blank value in the head row are ignored. A sheet with duplicate column names can not be queried.

Security Model

Only the spreadsheet that enables Link Sharing can be queried.

The schema is visible and manipulative only to its creator by default. To make the schema visible to others, create it as:

CREATE SCHEMA my_ss WITH (TOKEN = 'shtcnBf5pg4BNSkwV2Ku5xwW9Pf', PUBLIC = true)