sirajchokshi / csv-api

Instant JSON API from CSV data. Supports querying, mocking HTTP status, and hidden columns.

csv-api

Mock API tool that reads from a CSV and generates a JSON API at runtime. Custom identifiers allow for querying specific rows, hidden columns, and mock HTTP errors.

Generated API

route	description
/csv?url=...	Returns an index based on all rows in the CSV at the url param
/csv?url=...&{field}={value}	Returns 1 result where a row has a column {field} with value {value}. Only works on unique columns.
/healthcheck	returns 'ok' if the API is up

Schema

Reserved Column Names

Reserved column names provide special utility and start with a $.

value	description	example
$http	override http status, allows for forcing errors on entries	500

Identifier/Unique Columns

Columns names prefixed with an at-sign (@) can be used for lookups.

Example

@deviceId,name
1523,Roku 4K
1524,Apple TV

fetch("/csv?url=...&deviceId=1523"); // results for deviceId #1523
fetch("/csv?url=...&name=Apple%20TV"); // 404

Hidden Columns

Column names starting with a dollar sign ($) will never be returned. They are also used for reserved names, but can be used to embed any other data.

Shallow Columns

Fields by default will show up in the index (/csv). If you would like a field to only be available when requesting the individual ID prefix it with an exclamation point (!). The prefix is removed when the field is actually returned.

Example

@id,name,!age,$http
1,Alice,28,
2,Bob,26,404

When querying the /csv route without specifying an id, the API returns:

[
  {
    "id": "1",
    "name": "Alice"
  },
  {
    "id": "2",
    "name": "Bob"
  }
]

However, if an id is specified in the query, for example /csv?url=???&id=1, the API returns (note that the ! is stripped from the final field):

{
  "id": "1",
  "name": "Alice",
  "age": 28
}