Docker container that converts SQL Server backup files (.bak) to CSV.

• Restores MSSQL .bak files to a temporary SQL Server instance
• Lists and exports tables to CSV format
• Supports selective table export
• Handles multiple databases and schemas
• Properly quotes and escapes values in CSV output
• Distinguishes between SQL NULL, the string "null", and empty strings
• Outputs CSV files with configurable prefixes and suffixes

• Docker
• Sufficient disk space and memory for database restoration
• .bak file compatible with/upgradable to SQL Server 2022

# Basic usage (auto-detect BAK file, output all tables to mounted bak path)
docker run --rm \
  -v /path/to/backup/directory:/mnt/bak \
  mssql-bak2csv

# Use specified bak file, output selected tables to mounted csv path with given prefix and suffix
docker run --rm \
  -v /path/to/backup/directory:/mnt/bak \
  -v /path/to/output/directory:/mnt/csv \
  mssql-bak2csv \
  --bak-file your-database.bak \
  --tables "table1,table2,table3" \
  --prefix "prefix_" \
  --suffix "_backup"

• --bak-file FILENAME: Name of the .bak file (optional if only one exists)
• --tables TABLE1,TABLE2: Comma-separated list of tables to export (default: all)
• --prefix PREFIX: Prefix for CSV filenames (e.g., 'ABC1_AB_')
• --suffix SUFFIX: Suffix for CSV filenames (e.g., '_backup')
• --help: Show usage information

• Simple name: table1 (uses default database and schema)
• With database: mydatabase.table1 (specifies database, uses default schema)
• Fully qualified: mydatabase.myschema.table1 (specifies database and schema)

CSV files are named using the simplest format that avoids collisions:

• [PREFIX]{table}[SUFFIX].csv (e.g., Users.csv)
• [PREFIX]{schema}_{table}[SUFFIX].csv (e.g., dbo_Users.csv)
• [PREFIX]{database}_{schema}_{table}[SUFFIX].csv (e.g., MyDB_dbo_Users.csv)

Data values, such as images and binary, and floats are exported from sqlcmd as-is to avoid mangling by casting or converting or truncating them.

To ensure reliable parsing and distinction between SQL NULL, the string "null" (case-insensitive), and empty strings, the export logic encodes these values as follows:

• SQL NULL values are exported as an empty field (i.e., nothing between the commas: ,,).
• The string value "null" (case-insensitive, e.g., "NULL", "Null", "null") is always quoted and escaped (unless trailing or leading whitespace is present, in which case the value should be parsed as a string)
• Empty strings are exported as "" (two double quotes), distinguishing them from null values

This helps ensure consumers of the CSV can always distinguish between a true SQL NULL, a string containing "null", and an empty string.

Values that contain the delimiting character (,), line breaks, or quotes are quoted and escaped as follows:

1. Replace any quotes (") in the value with two quotes ("").
2. Wrap the value in quotes (").

• The container uses SQL Server Developer edition by default (not licensed for production use)
• Mounts: backup directory to /mnt/bak, optional output directory to /mnt/csv
• Database restoration is temporary - only CSV files persist between runs

By default, the container generates a secure random password for the SQL Server sa user at runtime.
To specify your own password, set the MSSQL_SA_PASSWORD environment variable when running the container:

docker run --rm \
  -e MSSQL_SA_PASSWORD='YourSecurePassword123!' \
  -v /path/to/backup/directory:/mnt/bak \
  mssql-bak2csv

The password must meet SQL Server complexity requirements (at least 8 characters, including uppercase, lowercase, number, and symbol).

This project uses the official Microsoft SQL Server container image (mcr.microsoft.com/mssql/server:2022-latest) and Microsoft SQL Server tools (mssql-tools, sqlcmd). By using this image and these tools, you agree to the Microsoft SQL Server End User License Agreement (EULA).

The default configuration uses SQL Server Developer edition, which is licensed for development and testing only—not for production workloads. To use another edition, set the MSSQL_PID environment variable as described in the SQL Server container documentation.

This project itself is provided under the MIT License (see LICENSE file).