Postprocessing

Post-processing Node

This node performs post-processing tasks such as replacing or cleaning specific words or patterns within text. It is mainly used to refine the output of an LLM, identifying and transforming specific formats of numbers or strings such as phone numbers or email addresses.

Target Pattern

Enter the value (pattern) you want to replace.

Post-processing Result

Enter the resulting value after the target pattern has been replaced.

Basic Usage

Enter the value/pattern to be replaced in Target Pattern.

Enter the replacement string in Post-processing Result.

Click Add to register the rule. You can register multiple rules as needed.

Usage Examples

1) Masking Sensitive Information

Regex Replacement

Target Pattern: \b\d{3}-\d{4}-\d{4}\b → Result: ***-****-****

Target Pattern: \b[0-9a-z._%+-]+@[0-9a-z.-]+\.[a-z]{2,}\b → Result: [email masked]

2) Format Cleanup (Improving Readability)

Regex Replacement

Target Pattern: \s{2,} → Result: (single space)

Target Pattern: \s*\n\s* → Result: \n (removes unnecessary blank lines)

3) Handling Forbidden/Sensitive Words

Word Replacement

Target Pattern: predefined forbidden-word list → Result: [filtered]

5) Simple Text Transformations

Regex Replacement

Target Pattern: (?i)\btrue\b → Result: Yes

Target Pattern: (?i)\bfalse\b → Result: No

When needed, use word replacement for exact matches and regex replacement for pattern-wide modifications to maximize efficiency.

Post-processing Node#

Target Pattern#

Post-processing Result#

Basic Usage#

Usage Examples#

1) Masking Sensitive Information#

2) Format Cleanup (Improving Readability)#

3) Handling Forbidden/Sensitive Words#

5) Simple Text Transformations#

Post-processing Node

Target Pattern

Post-processing Result

Basic Usage

Usage Examples

1) Masking Sensitive Information

2) Format Cleanup (Improving Readability)

3) Handling Forbidden/Sensitive Words

5) Simple Text Transformations