`POST /select/`¶

Executes a SQL-style SELECT query against an ingested namespace. Supports column projection, WHERE filtering, ORDER BY, LIMIT, OFFSET, and DISTINCT.

Request¶

Query Parameters:

Parameter	Description
`db_name`	Name of the database
`namespace`	Namespace to query

Body:

Parameter	Type	Required	Default	Description
`columns`	list of string	❌	all	Columns to return. If empty, all columns are returned
`where`	list of condition	❌	—	Filter conditions
`logical_op`	string	❌	`"and"`	How to combine WHERE conditions: `"and"` or `"or"`
`order_by`	list of order	❌	—	Sort fields
`limit`	int	❌	—	Max rows to return
`offset`	int	❌	`0`	Number of rows to skip
`distinct`	bool	❌	`false`	Remove duplicate rows

WHERE condition object:

Field	Type	Description
`field`	string	Column name to filter on
`operator`	string	One of: `eq`, `ne`, `gt`, `gte`, `lt`, `lte`, `like`, `in`, `notin`, `isnull`, `isnotnull`
`value`	string or list	Value to compare against. Use a list for `in` / `notin`

ORDER BY object:

Field	Type	Description
`field`	string	Column to sort by
`descending`	bool	`true` for descending, `false` for ascending (default)

Behavior¶

Filters are applied in order against RocksDB row metadata. Numeric comparisons (gt, gte, lt, lte) automatically cast values to float. String operations (like) support % wildcards. All conditions are combined using logical_op (and by default).

Responses¶

200 OK¶

{
  "total_rows": 3,
  "execution_time_ms": 12.4,
  "rows": [
    { "id": "E001", "name": "Alice", "department": "Engineering", "salary": "120000" },
    { "id": "E004", "name": "Dave",  "department": "Engineering", "salary": "95000"  }
  ]
}

Field	Description
`total_rows`	Number of rows returned
`execution_time_ms`	Query execution time
`rows`	Array of row objects

Error Responses¶

Status	Condition
`404`	Namespace not found
`500`	Unexpected internal error

Notes¶

like operator supports % prefix (%value), suffix (value%), and both (%value%).
Numeric comparisons work on string-stored numbers — they are cast to float at query time.
Column projection happens after filtering, so WHERE can reference columns not in columns.

Example¶

import requests

SERVER_URL = "http://18.220.128.24:8000"
API_KEY    = "yourapitoken"

def select(db_name: str, namespace: str, query: dict) -> dict:
    response = requests.post(
        f"{SERVER_URL}/select/",
        headers={"X-API-Key": API_KEY},
        params={"db_name": db_name, "namespace": namespace},
        json=query,
    )
    response.raise_for_status()
    return response.json()


result = select("my_db", "employees", {
    "columns": ["name", "department", "salary"],
    "where": [
        {"field": "department", "operator": "eq",  "value": "Engineering"},
        {"field": "salary",     "operator": "gte", "value": "100000"},
    ],
    "logical_op": "and",
    "order_by": [{"field": "salary", "descending": True}],
    "limit": 5,
})
print(result)

Expected output:

{
  "total_rows": 2,
  "execution_time_ms": 8.1,
  "rows": [
    { "name": "Alice", "department": "Engineering", "salary": "120000" },
    { "name": "Jack",  "department": "Engineering", "salary": "102000" }
  ]
}

POST /select/¶