Get All Questions

Method

REST methods are used. The API is read-only; there are no mutating endpoints.

HTTP Method

GET: Reads from pre-aggregated views

Naming Convention

To keep query parameters and node names consistent and readable, request URIs and JSON nodes use snake_case.

Endpoints follow the hierarchy /api/catalogs/{catalog_name}/schemas/{schema_name}/tables/{table_name}/..., mirroring the Unity Catalog structure directly in the URI.

Request and Response

Headers

Request Headers

Content-Type: application/json

There is no application-specific authentication header (see "Authentication").

Response Headers

Content-Type: application/json

Get All Questions

Returns every question in the given table, with its rows grouped per question. The frontend grid and heatmap views use this endpoint.

URI

/api/catalogs/{catalog_name}/schemas/{schema_name}/tables/{table_name}/all_questions

Path and Query Parameters

Being a GET, there is no request body.

Kind	Name	Description
Path	catalog_name	Unity Catalog catalog name. The frontend hardcodes `cs`.
Path	schema_name	Schema name. The frontend hardcodes `cs_dm`.
Path	table_name	Table name, selected in the UI.
Query	question_type	Filters by `Text Select` / `Free Text`. No filter when omitted.

Validation Rules

Field	Rule
catalog_name	Required. `^[A-Za-z0-9_]+$`. A violation returns 400.
schema_name	Required. `^[A-Za-z0-9_]+$`. A violation returns 400.
table_name	Required. `^[A-Za-z0-9_]+$`. A violation returns 400.
question_type	Optional. Passed to SQL as a `?` placeholder, so no format constraint.

Note that aggregation queries always filter by is_comp = 'すべての有効回答'. That condition is not parameterized.

Response (200 OK)

The response is JSON. The schema field is schema_name in Pydantic, but alias="schema" renders it as schema in JSON.

{
  "catalog": "cs",
  "schema": "cs_dm",
  "table": "dm_vis_jreslso_tokyo23_comp_test",
  "questions": [
    {
      "question": "現在の勤務形態について教えてください",
      "question_number": "03",
      "question_type": "Text Select",
      "data": [
        {
          "question_number": "03",
          "question_type": "Text Select",
          "question": "現在の勤務形態について教えてください",
          "choice_text": "出社のみ",
          "sum_answer": 412,
          "total": 1000,
          "rate": 0.412,
          "n": 1000,
          "is_comp": "すべての有効回答"
        }
      ]
    }
  ]
}

Each row in data uses extra="allow", so table-specific extra columns may appear as-is.

Other Endpoints

URI	Summary	Main Parameters
`GET /api/health`	Health check; returns `{"status": "healthy"}`	-
`GET /api/catalogs/{catalog_name}/schemas`	Lists schemas in the catalog (via the Workspace API)	-
`GET /api/catalogs/{catalog_name}/schemas/{schema_name}/tables`	Lists tables in the schema (via the Workspace API)	-
`GET /api/catalogs/{c}/schemas/{s}/tables/{t}/questions`	Lists questions in the table (`SELECT DISTINCT`)	`question_type`
`GET /api/catalogs/{c}/schemas/{s}/tables/{t}/questions/{question}`	Returns the rows of a single question	-
`GET /api/data_sql`	Development sample; returns `question_number='03'` from `cs.cs_dm.dm_vis_jreslso_tokyo23_comp_test`	-
`GET /api/dm_vis_jreslso_tokyo23_comp_test`	Development sample; returns the same view with filters	`question_number` (default `03`), `is_comp` (default `すべての有効回答`)

Full details for every endpoint are available in the Swagger UI (http://localhost:8000/api/docs).

Authentication

There is no application-specific authentication. End-user access control is delegated to Databricks Apps.

The app connects to Databricks with an OAuth service principal (M2M) using DATABRICKS_CLIENT_ID / DATABRICKS_CLIENT_SECRET. The data visible through the app is bounded by the Unity Catalog grants held by that service principal.

Error Handling

Status codes on error are as follows. Each handler in routes/ maps them from the exception type.

Description	Status Code	Status Name
Identifier violates the allowlist (`ValueError`)	400	Bad Request
FastAPI request validation failure	422	Unprocessable Entity
Internal server error	500	Internal Server Error
Missing configuration or uninstalled dependency (`RuntimeError`)	503	Service Unavailable

Processing Flow

Sequence Diagram

sequenceDiagram
    participant Client
    participant API as FastAPI (routes)
    participant SQL as backend/sql
    participant Warehouse as SQL Warehouse

    Client->>API: GET .../all_questions?question_type=Text Select
    API->>SQL: asyncio.to_thread(fetch_all_questions)
    SQL->>SQL: Validate identifiers via fully_qualified_name
    SQL->>Warehouse: SELECT ... WHERE is_comp = 'すべての有効回答' AND question_type = ?
    Warehouse-->>SQL: Rows
    SQL->>SQL: Group by question
    SQL-->>API: List[Dict]
    API-->>Client: 200 OK (AllQuestionsResponse)