simple-sqlite3 0.0.7

A simple SQLite3 wrapper for Python

Simple SQLite3

Effortless, Pythonic SQLite database management with a modern API and CLI.

Simple SQLite3 is a lightweight, Pythonic wrapper for Python’s built-in sqlite3 module, making it easy to work with SQLite databases. It offers a user-friendly API for managing tables, inserting and querying data, and exporting results, with built-in support for JSON/CSV/TXT, schema evolution, and a convenient CLI.

---

Features

• Easy-to-use API for SQLite database and table management.
• Command-line interface (CLI) for database operations.
• Support for exporting data to JSON, CSV and TXT formats.
• Robust to both nested and non-nested data, with datetime support.
• Utilities for processing queried results.

---

Installation

Requires Python 3.9+.

The package is available on PyPI:

pip install simple-sqlite3

---

Quick Start

Programmatic Usage

1. Insert and Query

This example demonstrates how to insert multiple rows into a table and query all records.

from simple_sqlite3 import Database

db = Database("database.db")
table = db.table("people")

table.insert(
    [
        {"name": "Amy", "age": 30, "city": "Helsinki"},
        {"name": "Bob", "age": 25, "city": "Cambridge"},
        {"name": "Cat", "age": 20, "city": "Paris"},
    ]
)

results = table.query("SELECT *")

print(results)

Example Output (from print(results)):

[
    {"name": "Amy", "age": 30, "city": "Helsinki"},
    {"name": "Bob", "age": 25, "city": "Cambridge"},
    {"name": "Cat", "age": 20, "city": "Paris"},
]

2. Bulk Insert Large Datasets

This example demonstrates how to efficiently insert many rows using either the insert_fast or insert_many methods. Both are designed for high performance with large datasets:

• insert_fast: Use when your data is a list of dictionaries, and all dictionaries have the same keys (homogeneous). This is ideal for bulk-inserting structured data with column names, and supports schema inference or explicit schema.
• insert_many: Use when your data is a list of tuples/lists (not dicts). This is fastest for simple, flat data and is less flexible (does not handle missing columns or nested data).

Note: Both methods provide significant speed efficiencies for large datasets. Use insert_fast for homogeneous dicts, and insert_many for tuple/list rows.

Example: Using insert_fast (homogeneous dicts)

from simple_sqlite3 import Database

db = Database(":memory:")
table = db.table("people")

data = [
    {"name": "Amy", "age": 30, "city": "Helsinki"},
    {"name": "Bob", "age": 25, "city": "Cambridge"},
    {"name": "Cat", "age": 20, "city": "Paris"},
]

table.insert_fast(data) # Or table.insert(data, fast=True)

Example: Using insert_many (tuples/lists)

from simple_sqlite3 import Database

db = Database(":memory:")
table = db.table("people")

rows = [
    ("Amy", 30, "Helsinki"),
    ("Bob", 25, "Cambridge"),
    ("Cat", 20, "Paris"),
]
columns = ("name", "age", "city")
schema = "name TEXT, age INTEGER, city TEXT" # Optional

table.insert_many(rows, columns, schema=schema)

3. Insert Nested Data

This example demonstrates inserting nested (dictionary) data, which is automatically stored as JSON in SQLite.

from simple_sqlite3 import Database

db = Database("database.db")
table = db.table("nested")

table.insert(
    [
        {
            "country": "Finland",
            "info": {
                "capital": "Helsinki",
                "latitude": 60.1699,
                "longitude": 24.9384,
            },
        },
        {
            "country": "France",
            "info": {
                "capital": "Paris",
                "latitude": 48.8566,
                "longitude": 2.3522,
            },
        },
        {
            "country": "Japan",
            "info": {
                "capital": "Tokyo",
                "latitude": 35.6895,
                "longitude": 139.6917,
            },
        },
    ]
)

4. Insert and Query Timeseries

This example demonstrates how to insert timeseries data using Python datetime objects and conditionally query rows, automatically parsing dates.

from simple_sqlite3 import Database
from datetime import datetime as dt

db = Database("database.db")
table = db.table("timeseries")

table.insert(
    [
        {"date": dt(2024, 6, 1), "value": 1.2345, "pair": "EURUSD"},
        {"date": dt(2024, 6, 2), "value": 1.2350, "pair": "EURUSD"},
        {"date": dt(2024, 6, 3), "value": 1.2360, "pair": "EURUSD"},
        {"date": dt(2024, 6, 4), "value": 1.2375, "pair": "EURUSD"},
        {"date": dt(2024, 6, 1), "value": 109.45, "pair": "USDJPY"},
        {"date": dt(2024, 6, 2), "value": 109.60, "pair": "USDJPY"},
        {"date": dt(2024, 6, 3), "value": 109.75, "pair": "USDJPY"},
        {"date": dt(2024, 6, 4), "value": 109.90, "pair": "USDJPY"},
    ]
)

results = table.query("SELECT date, value WHERE pair = 'EURUSD'", auto_parse_dates=True)

print(results)

Example Output (from print(results)):

[
    {"date": datetime.datetime(2024, 6, 1, 0, 0), "value": 1.2345},
    {"date": datetime.datetime(2024, 6, 2, 0, 0), "value": 1.235},
    {"date": datetime.datetime(2024, 6, 3, 0, 0), "value": 1.236},
    {"date": datetime.datetime(2024, 6, 4, 0, 0), "value": 1.2375},
]

5. Insert Mixed Data

This example demonstrates inserting mixed data, including deeply-nested dictionaries.

from simple_sqlite3 import Database
from datetime import datetime as dt

db = Database("database.db")
table = db.table("mixed_data")

table.insert(
    [
        {
            "date": dt(2024, 6, 1),
            "value": 1.2345,
            "pair": "EURUSD",
            "source": "ECB",
        },
        {
            "date": dt(2024, 6, 1),
            "value": 109.45,
            "pair": "USDJPY",
            "source": "BOJ",
        },
        {
            "date": dt(2024, 6, 2),
            "value": 0.8567,
            "pair": "EURGBP",
            "source": "ECB",
        },
        {
            "date": dt(2024, 6, 2),
            "value": 1.4200,
            "pair": "GBPUSD",
            "source": "FED",
        },
        {
            "date": dt(2024, 6, 2),
            "value": 1.2370,
            "pair": "EURUSD",
            "source": "ECB",
            "meta": {
                "confidence": 0.98,
                "contributors": ["ECB", "Bloomberg"],
                "valuation": {"buy": 0.4, "hold": 0.2, "sell": 0.4},
            },
        },
        {
            "date": dt(2024, 6, 3),
            "value": 109.80,
            "pair": "USDJPY",
            "source": "BOJ",
            "meta": {"confidence": 0.95, "contributors": ["BOJ"]},
        },
    ]
)

6. Insert Data Into Memory and Export as JSON, CSV and TXT

This example demonstrates inserting data into an in-memory database and exporting the table to JSON, CSV, and TXT formats.

from simple_sqlite3 import Database
from datetime import datetime as dt

db = Database(":memory:")
table = db.table("timeseries")

table.insert(
    [
        {"date": dt(2025, 5, 22), "value": 5328, "idx": "S&P 500"},
        {"date": dt(2025, 5, 21), "value": 5421, "idx": "S&P 500"},
        {"date": dt(2025, 5, 22), "value": 5448, "idx": "EURO STOXX 50"},
        {"date": dt(2025, 5, 21), "value": 5452, "idx": "EURO STOXX 50"},
    ]
)

table.export_to_json("timeseries.json")
table.export_to_csv("timeseries.csv")
table.export_to_txt("timeseries.txt")

7. Exporting Queried Results

This example demonstrates how to export queried results using the QueryResultsProcessor utility.

from simple_sqlite3 import Database
from simple_sqlite3.utils import QueryResultsProcessor
from datetime import datetime as dt

db = Database(":memory:")
table = db.table("timeseries")

table.insert(
    [
        {"date": dt(2025, 5, 22), "value": 5328, "idx": "S&P 500"},
        {"date": dt(2025, 5, 21), "value": 5421, "idx": "S&P 500"},
        {"date": dt(2025, 5, 22), "value": 5448, "idx": "EURO STOXX 50"},
        {"date": dt(2025, 5, 21), "value": 5452, "idx": "EURO STOXX 50"},
    ]
)

results = table.query("SELECT * WHERE idx = 'EURO STOXX 50'")

processor = QueryResultsProcessor(results)

processor.to_json("timeseries.json")
processor.to_csv("timeseries.csv")
processor.to_txt("timeseries.txt")

8. Grouping Queried Data

This example demonstrates how to group queried data into a matrix format for easy analysis.

from simple_sqlite3 import Database
from simple_sqlite3.utils import QueryResultsProcessor
from datetime import datetime as dt

db = Database(":memory:")
table = db.table("timeseries")

table.insert(
    [
        {"date": dt(2024, 6, 1), "value": 1.2345, "pair": "EURUSD"},
        {"date": dt(2024, 6, 2), "value": 1.2350, "pair": "EURUSD"},
        {"date": dt(2024, 6, 3), "value": 1.2360, "pair": "EURUSD"},
        {"date": dt(2024, 6, 4), "value": 1.2375, "pair": "EURUSD"},
        {"date": dt(2024, 6, 1), "value": 109.45, "pair": "USDJPY"},
        {"date": dt(2024, 6, 2), "value": 109.60, "pair": "USDJPY"},
        {"date": dt(2024, 6, 3), "value": 109.75, "pair": "USDJPY"},
        {"date": dt(2024, 6, 4), "value": 109.90, "pair": "USDJPY"},
    ]
)

results = table.query("SELECT *", auto_parse_dates=True)

processor = QueryResultsProcessor(results)

results_matrix_format = processor.to_matrix_format(
    index_key="date", group_key="pair", value_key="value"
)

print(results_matrix_format)

Example Output (from print(results)):

{
    "index": [
        datetime.datetime(2024, 6, 1, 0, 0),
        datetime.datetime(2024, 6, 2, 0, 0),
        datetime.datetime(2024, 6, 3, 0, 0),
        datetime.datetime(2024, 6, 4, 0, 0),
    ],
    "columns": ["EURUSD", "USDJPY"],
    "values": [[1.2345, 109.45], [1.235, 109.6], [1.236, 109.75], [1.2375, 109.9]],
}

---

CLI Usage

The CLI is installed automatically with the package:

pip install simple-sqlite3

You can run the CLI using the db command (if your Python scripts directory is in your PATH or you have an active virtual environment), or with:

python -m simple_sqlite3.cli

CLI Command Overview

Command	Description
db --help	Show help and available commands
query	Query records from a table
insert	Insert data from a file (CSV, JSON, TXT)
export	Export table data to a file (CSV, JSON, TXT)
rename-column	Rename a column in a table
rename-columns	Rename multiple columns in a table
delete-column	Delete a column from a table
delete-columns	Delete multiple columns from a table
rename-table	Rename a table
delete-duplicates	Remove duplicate rows from a table
delete-table	Delete a table from the database
delete-database	Delete the entire database file
vacuum	Reclaim unused space and optimize the database file

Common Flags

Short	Long	Description
-d	--database	Path to the SQLite database
-t	--table	Name of the table
-f	--file	Input/output file path
-s	--sql	SQL query to execute
-F	--force	Force action without confirmation

---

Show CLI help

Displays help and available commands.

db --help

Insert data from a JSON file into a table

Inserts data from timeseries.json into the timeseries table in database.db.

db insert -d database.db -t timeseries -f timeseries.json

If you don't have a timeseries.json file, you can create one with the following example content:

[
    {"date": "2025-05-22", "value": 5328, "idx": "S&P 500"},
    {"date": "2025-05-21", "value": 5421, "idx": "S&P 500"},
    {"date": "2025-05-22", "value": 5448, "idx": "EURO STOXX 50"},
    {"date": "2025-05-21", "value": 5452, "idx": "EURO STOXX 50"}
]

Query all rows from a table

Queries all rows from the timeseries table.

db query -d database.db -t timeseries -s "SELECT *"

Remove duplicate rows from a table

Removes duplicate rows from the timeseries table.

db delete-duplicates -d database.db -t timeseries

Export a table to CSV format

Exports the timeseries table to timeseries.csv.

db export -d database.db -t timeseries -f timeseries.csv

Delete a table from the database

Deletes the timeseries table from database.db. Use -F to skip confirmation.

db delete-table -d database.db -t timeseries -F

Delete the entire database file

Deletes the database.db file. Use -F to skip confirmation.

db delete-database -d database.db -F

---

Advanced Features

• Automatic WAL Mode: Write-Ahead Logging for better concurrency (default).
• Schema Evolution: New columns are added automatically on insert if force=True (default).
• Batch Export: Efficiently export large tables in batches to avoid memory issues.

---

License

This project is developed by Rob Suomi and licensed under the MIT License.
See the LICENSE file for details.