sdeul

Structural Data Extractor using LLMs

Installation

$ pip install -U sdeul

Usage

Command Line Interface

1. Create a JSON Schema file for the output

2. Prepare a local model GGUF file or model API key.

3. Extract structural data from given text using sdeul extract.

   Example:

   # Use OpenAI API
   $ sdeul extract --openai-model='gpt-4.1' \
       test/data/medication_history.schema.json \
       test/data/patient_medication_record.txt
   
   # Use Amazon Bedrock API
   $ sdeul extract --bedrock-model='us.anthropic.claude-sonnet-4-20250514-v1:0' \
       test/data/medication_history.schema.json \
       test/data/patient_medication_record.txt
   
   # Use Ollama API
   $ sdeul extract --ollama-model='gemma3:27b' \
       test/data/medication_history.schema.json \
       test/data/patient_medication_record.txt
   
   # Use a Llama.cpp GGUF model file
   $ sdeul extract --llamacpp-model-file='local_llm.gguf' \
       test/data/medication_history.schema.json \
       test/data/patient_medication_record.txt
   

   Expected output:

   {
     "MedicationHistory": [
       {
         "MedicationName": "Lisinopril",
         "Dosage": "10mg daily",
         "Frequency": "daily",
         "Purpose": "hypertension"
       },
       {
         "MedicationName": "Metformin",
         "Dosage": "500mg twice daily",
         "Frequency": "twice daily",
         "Purpose": "type 2 diabetes"
       },
       {
         "MedicationName": "Atorvastatin",
         "Dosage": "20mg at bedtime",
         "Frequency": "at bedtime",
         "Purpose": "high cholesterol"
       }
     ]
   }
   

REST API

SDEUL also provides a REST API for extracting structured data and validating JSON.

1. Start the API server:

   $ sdeul serve
   

2. The API will be available at http://localhost:8000 with the following endpoints:

   • POST /extract - Extract structured data from text
   • POST /validate - Validate JSON data against a schema
   • GET /health - Health check endpoint
   • GET /docs - Interactive API documentation

3. Example API usage:

   # Extract data using OpenAI
   $ curl -X POST "http://localhost:8000/extract" \
     -H "Content-Type: application/json" \
     -d '{
       "text": "Patient is taking Lisinopril 10mg daily for hypertension.",
       "json_schema": {
         "type": "object",
         "properties": {
           "medications": {
             "type": "array",
             "items": {
               "type": "object",
               "properties": {
                 "name": {"type": "string"},
                 "dosage": {"type": "string"},
                 "condition": {"type": "string"}
               }
             }
           }
         }
       },
       "openai_model": "gpt-4o-mini",
       "openai_api_key": "your-api-key"
     }'
   
   # Validate JSON data
   $ curl -X POST "http://localhost:8000/validate" \
     -H "Content-Type: application/json" \
     -d '{
       "data": {"medications": [{"name": "Lisinopril", "dosage": "10mg", "condition": "hypertension"}]},
       "json_schema": {
         "type": "object",
         "properties": {
           "medications": {
             "type": "array",
             "items": {
               "type": "object",
               "properties": {
                 "name": {"type": "string"},
                 "dosage": {"type": "string"},
                 "condition": {"type": "string"}
               }
             }
           }
         }
       }
     }'