Data Type Mappings :: Synthesized Docs

Overview

When transforming data between databases, the platform automatically maps source data types to equivalent destination types. This ensures data compatibility while preserving semantics and precision.

Type Mapping Strategy

The platform uses these rules for type mapping:

Preserve semantics: Ensure data meaning is maintained
Preserve precision: Don’t lose decimal places or string length
Use native types: Prefer destination database’s native types
Handle edge cases: NULL values, special characters, timezone data

Common Type Mappings

String Types

Source Type	PostgreSQL	MySQL
VARCHAR(N)	VARCHAR(N)	VARCHAR(N)
CHAR(N)	CHAR(N)	CHAR(N)
TEXT	TEXT	TEXT / LONGTEXT
CLOB	TEXT	LONGTEXT

Source Type

PostgreSQL

MySQL

VARCHAR(N)

CHAR(N)

TEXT

TEXT / LONGTEXT

CLOB

TEXT

LONGTEXT

Numeric Types

Source	PostgreSQL	MySQL	Oracle
INTEGER	INTEGER	INT	NUMBER(10)
BIGINT	BIGINT	BIGINT	NUMBER(19)
DECIMAL(P,S)	NUMERIC(P,S)	DECIMAL(P,S)	NUMBER(P,S)
FLOAT	REAL	FLOAT	BINARY_FLOAT
DOUBLE	DOUBLE PRECISION	DOUBLE	BINARY_DOUBLE

Source

PostgreSQL

MySQL

Oracle

INTEGER

INT

NUMBER(10)

BIGINT

NUMBER(19)

DECIMAL(P,S)

NUMERIC(P,S)

DECIMAL(P,S)

NUMBER(P,S)

FLOAT

REAL

FLOAT

BINARY_FLOAT

DOUBLE

DOUBLE PRECISION

DOUBLE

BINARY_DOUBLE

Date/Time Types

Source	PostgreSQL	MySQL
DATE	DATE	DATE
TIME	TIME	TIME
DATETIME	TIMESTAMP	DATETIME
TIMESTAMP	TIMESTAMP	TIMESTAMP

Source

PostgreSQL

MySQL

DATE

TIME

DATETIME

TIMESTAMP

DATETIME

TIMESTAMP

Boolean Types

Source	PostgreSQL	MySQL	Oracle
BOOLEAN	BOOLEAN	TINYINT(1)	NUMBER(1)
BIT	BOOLEAN	BIT	NUMBER(1)

Source

PostgreSQL

MySQL

Oracle

BOOLEAN

TINYINT(1)

NUMBER(1)

BIT

BOOLEAN

BIT

NUMBER(1)

Database-Specific Types

PostgreSQL Specific

Arrays:

INTEGER[] → Supported in PostgreSQL only
         → Converted to JSON in other databases

JSON/JSONB:

JSONB → Supported in PostgreSQL
      → JSON in MySQL
      → CLOB in Oracle

UUID:

UUID → UUID in PostgreSQL
     → CHAR(36) in MySQL/Oracle

See: Type Handling

Oracle Specific

NUMBER:

NUMBER → NUMERIC in PostgreSQL
       → DECIMAL in MySQL

CLOB/BLOB:

CLOB → TEXT in PostgreSQL
     → LONGTEXT in MySQL

MySQL Specific

TINYINT:

TINYINT(1) → BOOLEAN in PostgreSQL
           → NUMBER(1) in Oracle

See: MySQL TINYINT

ENUM:

ENUM('A','B','C') → VARCHAR with CHECK constraint

Type Conversion During Transformation

The platform handles type conversions automatically:

Implicit Conversions

String → Number:

# Original column: VARCHAR containing "12345"
# Destination: INTEGER
# Platform converts: "12345" → 12345

Number → String:

# Original: INTEGER 12345
# Destination: VARCHAR
# Platform converts: 12345 → "12345"

Explicit Conversions

Use Value Mapping for custom conversions:

- columns: ["status_code"]
  type: ValueMapping
  params:
    mappings:
      "1": "ACTIVE"
      "0": "INACTIVE"

See: Value Mapping

Handling Special Cases

NULL Values

Preserved across all type conversions:

NULL → NULL (regardless of type)

Empty Strings

"" (empty string) → "" in VARCHAR
                  → NULL in some databases (Oracle)

Date Formats

The platform uses ISO 8601 format internally:

2024-01-15T14:30:00Z

Time Zones

Timestamps converted to destination timezone if specified.

Custom Type Handling

For unsupported types, use Scripting Transformer:

- columns: ["custom_type_column"]
  type: Scripting
  params:
    mode: MASKING
    code: |
      (ctx, originalRecord) => {
        const value = originalRecord.get('custom_type_column');
        // Custom conversion logic
        return convertCustomType(value);
      }

Best Practices

Test type mappings on small datasets first
Check precision for DECIMAL types (don’t lose decimal places)
Validate NULL handling for NOT NULL constraints
Use explicit mappings for critical business data
Document custom conversions in workflow config

Troubleshooting

Common type-related issues: