UPDATE

Author: mazzasaverio

data/custom_extraction_20250628_215530.json

@@ -0,0 +1,16 @@
+{
+  "headline": "Tech Giant Unveils New Smartphone with Innovative Features",
+  "summary": "Today, a leading technology company announced the release of its latest smartphone model. Packed with state-of-the-art technology, including a groundbreaking camera and enhanced battery life, this new device promises to set a new benchmark in the industry. During a launch event, the CEO discussed the advanced features and their potential impact on daily life.",
+  "publication_date": "November 15, 2023",
+  "author": "Jamie Smith",
+  "location": "California, USA",
+  "key_people": [
+    "CEO John Doe",
+    "Product Director Emily Davis"
+  ],
+  "organizations": [
+    "ABC Technology Corp"
+  ],
+  "category": "Technology",
+  "sentiment": "Positive"
+}

examples/news_schema.json

@@ -44,6 +44,6 @@
         "description": "Overall sentiment of the article (positive, negative, neutral)"
       }
     },
-    "required": ["headline", "summary", "key_people", "organizations"],
+    "required": ["headline", "summary", "publication_date", "author", "location", "key_people", "organizations", "category", "sentiment"],
     "additionalProperties": false
   }

src/structured_output_cookbook/cli.py

@@ -2,6 +2,7 @@
 
 import json
 import sys
+from datetime import datetime
 from pathlib import Path
 from typing import Optional
 
@@ -19,6 +20,37 @@
 }
 
 
+def save_extraction_result(
+    result_data: dict,
+    template_name: str,
+    output_path: Optional[str] = None,
+    data_dir: str = "data"
+) -> str:
+    """Save extraction result to file and return the path."""
+    if output_path:
+        # Use specified path
+        save_path = Path(output_path)
+    else:
+        # Generate automatic path in data directory
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        filename = f"{template_name}_extraction_{timestamp}.json"
+        
+        data_path = Path(data_dir)
+        data_path.mkdir(exist_ok=True)
+        save_path = data_path / filename
+    
+    # Ensure parent directory exists
+    save_path.parent.mkdir(parents=True, exist_ok=True)
+    
+    # Save the file
+    save_path.write_text(
+        json.dumps(result_data, indent=2, ensure_ascii=False),
+        encoding="utf-8"
+    )
+    
+    return str(save_path)
+
+
 @click.group()
 @click.option("--debug", is_flag=True, help="Enable debug logging")
 @click.pass_context
@@ -47,16 +79,20 @@ def list_templates() -> None:
 @click.argument("template", type=click.Choice(list(TEMPLATES.keys())))
 @click.option("--input-file", "-i", type=click.Path(exists=True), help="Input text file")
 @click.option("--text", "-t", help="Input text directly")
-@click.option("--output", "-o", type=click.Path(), help="Output JSON file")
+@click.option("--output", "-o", type=click.Path(), help="Output JSON file (default: auto-generated in data/)")
+@click.option("--data-dir", default="data", help="Directory for auto-generated outputs")
 @click.option("--pretty", is_flag=True, help="Pretty print JSON output")
+@click.option("--no-save", is_flag=True, help="Don't save to file, only print to stdout")
 @click.pass_context
 def extract(
     ctx: click.Context,
     template: str,
     input_file: Optional[str],
     text: Optional[str],
     output: Optional[str],
-    pretty: bool
+    data_dir: str,
+    pretty: bool,
+    no_save: bool
 ) -> None:
     """Extract data using a predefined template."""
     logger = ctx.obj["logger"]
@@ -82,20 +118,29 @@ def extract(
         click.echo(f"Extraction failed: {result.error}", err=True)
         sys.exit(1)
 
+    # Ensure we have data
+    if result.data is None:
+        click.echo("Error: Extraction succeeded but no data returned", err=True)
+        sys.exit(1)
+    
     # Format output
     indent = 2 if pretty else None
     output_json = json.dumps(result.data, indent=indent, ensure_ascii=False)
 
-    # Write output
-    if output:
-        Path(output).write_text(output_json, encoding="utf-8")
-        click.echo(f"Results saved to {output}")
-    else:
+    # Save to file unless --no-save is specified
+    if not no_save:
+        save_path = save_extraction_result(result.data, template, output, data_dir)
+        click.echo(f"✅ Results saved to {save_path}")
+    
+    # Always print to stdout if no output file specified or if pretty print requested
+    if not output or pretty or no_save:
+        click.echo("📄 Extraction Result:")
         click.echo(output_json)
 
     # Show stats
     if result.tokens_used:
-        logger.info(f"Tokens used: {result.tokens_used}")
+        click.echo(f"📊 Tokens used: {result.tokens_used}")
+        click.echo(f"💰 Estimated cost: ~${(result.tokens_used * 0.00001):.4f}")  # Rough estimate
 
 
 @main.command()
@@ -104,8 +149,10 @@ def extract(
 @click.option("--prompt", help="System prompt text")
 @click.option("--input-file", "-i", type=click.Path(exists=True), help="Input text file")
 @click.option("--text", "-t", help="Input text directly")
-@click.option("--output", "-o", type=click.Path(), help="Output JSON file")
+@click.option("--output", "-o", type=click.Path(), help="Output JSON file (default: auto-generated in data/)")
+@click.option("--data-dir", default="data", help="Directory for auto-generated outputs")
 @click.option("--pretty", is_flag=True, help="Pretty print JSON output")
+@click.option("--no-save", is_flag=True, help="Don't save to file, only print to stdout")
 @click.pass_context
 def extract_custom(
     ctx: click.Context,
@@ -115,7 +162,9 @@ def extract_custom(
     input_file: Optional[str],
     text: Optional[str],
     output: Optional[str],
-    pretty: bool
+    data_dir: str,
+    pretty: bool,
+    no_save: bool
 ) -> None:
     """Extract data using a custom JSON schema."""
     logger = ctx.obj["logger"]
@@ -156,20 +205,29 @@ def extract_custom(
         click.echo(f"Extraction failed: {result.error}", err=True)
         sys.exit(1)
 
+    # Ensure we have data
+    if result.data is None:
+        click.echo("Error: Extraction succeeded but no data returned", err=True)
+        sys.exit(1)
+    
     # Format output
     indent = 2 if pretty else None
     output_json = json.dumps(result.data, indent=indent, ensure_ascii=False)
 
-    # Write output
-    if output:
-        Path(output).write_text(output_json, encoding="utf-8")
-        click.echo(f"Results saved to {output}")
-    else:
+    # Save to file unless --no-save is specified
+    if not no_save:
+        save_path = save_extraction_result(result.data, "custom", output, data_dir)
+        click.echo(f"✅ Results saved to {save_path}")
+    
+    # Always print to stdout if no output file specified or if pretty print requested
+    if not output or pretty or no_save:
+        click.echo("📄 Extraction Result:")
         click.echo(output_json)
 
     # Show stats
     if result.tokens_used:
-        logger.info(f"Tokens used: {result.tokens_used}")
+        click.echo(f"📊 Tokens used: {result.tokens_used}")
+        click.echo(f"💰 Estimated cost: ~${(result.tokens_used * 0.00001):.4f}")  # Rough estimate
 
 
 if __name__ == "__main__":