Translate Teradata To Redshift

Translation Mapping Examples

• QUALIFY clause → WHERE with window function result
• TIMESTAMP WITH TIME ZONE → TIMESTAMP (Redshift uses UTC)
• ROW_NUMBER() OVER (PARTITION BY x) → Keep as-is (Redshift supports it)
• Teradata CAST variants → Standard CAST with Redshift data types
• ANSI Temporal tables → Convert to standard query patterns
• FASTLOAD directives → Remove (not applicable to Redshift)

Output Format

Provide the output as a JSON structure with this schema:

{
  "folder_name": [
    {
      "file_name": "string",
      "original_query": "original Teradata query",
      "translated_query": "translated Redshift query",
      "translation_status": "SUCCESS|PARTIAL|FAILED",
      "issues": ["issue1", "issue2"],
      "notes": "explanation of changes made"
    }
  ]
}

Examples

Example 1: Simple QUALIFY Translation

• Input: SELECT id, name, ROW_NUMBER() OVER (ORDER BY id) as rn FROM users QUALIFY ROW_NUMBER() OVER (ORDER BY id) = 1
• Output: SELECT id, name FROM (SELECT id, name, ROW_NUMBER() OVER (ORDER BY id) as rn FROM users) WHERE rn = 1
• Status: SUCCESS

Example 2: TIMESTAMP WITH TIME ZONE

• Input: SELECT created_ts TIMESTAMP WITH TIME ZONE FROM events
• Output: SELECT created_ts FROM events
• Status: SUCCESS with note about timezone conversion

Guidelines

• Handle multi-line queries gracefully
• Preserve query formatting where possible
• If a translation is uncertain, mark as PARTIAL and explain in notes
• Include warnings about features that may affect performance
• Handle nested subqueries properly
• Maintain accurate line mapping between original and translated queries
• If translation completely fails, explain what feature is blocking translation