sql-pydb 0.3.0

A Python package for simplified SQL operations across MSSQL, PostgreSQL, MySQL, and SQLite.

SQLPy

SQLPy (sql_pydb) is a Python library to bridge the gap between Python and Database Tables.

---

Instalation Guide

pip install sql_pydb

• sql_pydb requires the following packages: Pandas, Numpy, pyodbc, python-dateutil

Overview:

• You can connect to MSSQL and SQLite.
• You can perform operations such as:
  • Create Database(s)
  • Create Table(s)
  • Drop Column(s)
  • Add Column(s)
  • Alter Column Types (If database supports it)
    • Performs a Try Cast check to be safe. (i.e. VARCHAR to INTEGER)
  • Read from Table
  • Delete from Table
  • Drop Table(s)

Usage

1. Example: Upload Pandas DataFrame to a MSSQL Database

from sql_pydb.Database import DatabaseType 
from sql_pydb.Table import TableActions

table = TableActions(database=DATABASE, table_name=TABLE_NAME, 
                driver="ODBC Driver 17 for SQL Server",
                server="url,port",
                username="admin",
                password="admin",
                db_type=DatabaseType.MSSQL)

df = pd.read_csv("...file.csv")
table.identify_schema(df)   # Figures out the best column types. Only needed 1 time per TableActions instance
table.sync_schema(update_column_types=True, add_new_columns=True, delete_old_columns=False) # Optional
status = table.insert_df(df, batch=1_000)
assert status == True

2. Example: Upload Pandas DataFrame to a SQLite Database

from sql_pydb.Database import DatabaseType 
from sql_pydb.Table import TableActions

table = TableActions(table_name=TABLE_NAME, 
                sqlite_path = SQLITE_PATH,  # "./path/file.db"
                db_type=DatabaseType.SQLITE)

df = pd.read_csv("...file.csv")
table.identify_schema(df)   # Figures out the best column types. Only needed 1 time per TableActions instance
table.sync_schema(update_column_types=True, add_new_columns=True, delete_old_columns=False) # Optional
status = table.insert_df(df, batch=1_000)
assert status == True

3. Example: Query Table

table = TableActions(...)
# ...
command = table.generate_select_all_sql()
df = table.run_query(command)   # Returns pd.DataFrame

4. Example: Execute Transactions

table = TableActions(...)
# ...
command1 = table.generate_add_columns_sql()
command2 = table.generate_drop_column_sql(column_name="UNUSED_COLUMN")
# ...
commandX = table.generate_insert_sql(df_new_records)
status = table.run_transaction([command1, command2, ..., commandX])
assert status == True

5. Example: Create Table

from sql_pydb.Column import Column
from sql_pydb.Table import TableActions

table = TableActions(...)
# ...
user_defined_columns = [
    Column("FIRST_NAME", data_type="VARCHAR", size=100),
    ...
    Column("ADDRESS", data_type="VARCHAR"), # Default 255 for strings
]
table.identify_schema(df, user_defined_columns)
# Option 1: Let it automatically sync with the new defined columns
status = table.sync_schema(update_column_types=True, add_new_columns=True, delete_old_columns=False)
# Option 2: Manually create the table
command = table.generate_create_table_sql()
status = table.run_transaction([command])
assert status == True

• You may choose to replace user_defined_columns with a pd.DataFrame with ["COLUMN", "DATA_TYPE", "SIZE"] columns. ["DATABASE", "TABLE"] columns may also be passed in.

---

Support:

• If you have a question or need additional support, please create an issue ticket on SQLPy GitHub Repo.