MyScale

opencollection: 1.0.0
info:
  name: MyScale SQL (ClickHouse HTTP) Interface
  version: '1.0'
request:
  auth:
    type: basic
    username: '{{username}}'
    password: '{{password}}'
items:
- info:
    name: Query
    type: folder
  items:
  - info:
      name: Ping / health check
      type: http
    http:
      method: GET
      url: https://{{clusterHost}}:8443/
    docs: GET the root of the HTTP interface; returns 'Ok.' when the cluster is reachable.
  - info:
      name: Create table with vector index
      type: http
    http:
      method: POST
      url: https://{{clusterHost}}:8443/
      body:
        type: text
        data: "CREATE TABLE IF NOT EXISTS default.articles\n(\n    id UInt64,\n    title String,\n    content_vector Array(Float32),\n\
          \    CONSTRAINT cons_vec_len CHECK length(content_vector) = 768,\n    VECTOR INDEX article_idx content_vector TYPE\
          \ HNSWFLAT('metric_type=Cosine')\n)\nENGINE = MergeTree ORDER BY id"
    docs: Execute DDL over the ClickHouse HTTP interface to create a table with a VECTOR INDEX.
  - info:
      name: Insert rows
      type: http
    http:
      method: POST
      url: https://{{clusterHost}}:8443/?query=INSERT%20INTO%20default.articles%20FORMAT%20JSONEachRow
      body:
        type: text
        data: '{"id": 1, "title": "example", "content_vector": [0.1, 0.2, 0.3]}'
    docs: Insert data by sending an INSERT ... FORMAT statement; the rows go in the request body.
  - info:
      name: Vector similarity search (distance)
      type: http
    http:
      method: POST
      url: https://{{clusterHost}}:8443/?default_format=JSON
      body:
        type: text
        data: 'SELECT id, title, distance(content_vector, [0.1, 0.2, 0.3]) AS dist

          FROM default.articles

          ORDER BY dist ASC

          LIMIT 10'
    docs: Run a vector similarity search in SQL using the distance() function ordered by distance.