add functions to evaluate feedstock sustainability

sequestrae_engine/cli/commands.py

@@ -64,3 +64,35 @@ def extract_audit_information_command(api_key, project_dir, limit=100):
 
     print(f"Successfully processed {markdown_count} markdown files")
     return 0
+
+
+def evaluate_feedstock_sustainability_command(api_key, project_dir, limit=100):
+    if limit is None:
+        limit = 100
+    print(f"Max number of files to process: {limit}")
+
+    if not api_key:
+        print("Error: MISTRAL_API_KEY is required")
+        return 1
+
+    extractor = AuditReportExtractor(api_key=api_key)
+    project_path = Path(project_dir)
+
+    if not project_path.exists():
+        print(f"Error: Directory not found at {project_path}")
+        return 1
+
+    markdown_count = 0
+    for folder_path in project_path.iterdir():
+        if not folder_path.name.startswith(".") and folder_path.is_dir() and markdown_count < limit:
+            for md_path in folder_path.glob("*.md"):
+                if md_path.is_file() and "report" in md_path.stem.lower():
+                    markdown_count += 1
+                    try:
+                        extractor.analyze_feedstock_sustainability(audit_path=md_path)
+                    except Exception as e:
+                        print(f"Error processing {md_path}: {str(e)}")
+                    time.sleep(1)  # Sleep for 1 second to avoid rate limiting
+
+    print(f"Successfully processed {markdown_count} markdown files")
+    return 0

sequestrae_engine/cli/main.py

@@ -39,6 +39,25 @@ def main():
         )
     )
 
+    # Evaluate feedstock sustainability command
+    evaluate_feedstock_parser = subparsers.add_parser(
+        "evaluate-feedstock", help="Evaluate feedstock sustainability from markdown report files"
+    )
+    evaluate_feedstock_parser.add_argument(
+        "--mistral-api-key", help="Mistral API key", required=True
+    )
+    evaluate_feedstock_parser.add_argument(
+        "--project-dir", help="Project data directory", required=True
+    )
+    evaluate_feedstock_parser.add_argument(
+        "--limit", help="Maximum number of files to process", type=int
+    )
+    evaluate_feedstock_parser.set_defaults(
+        func=lambda args: commands.evaluate_feedstock_sustainability_command(
+            args.mistral_api_key, args.project_dir, args.limit
+        )
+    )
+
     args = parser.parse_args()
     if hasattr(args, "func"):
         exit_status = args.func(args)

sequestrae_engine/document_parsing/extractors.py

@@ -2,14 +2,25 @@
 import logging
 import os
 import re
+import time
+from typing import Dict, List
 
 from mistralai import Mistral
 
+from sequestrae_engine.core.utilities import load_json_file
+
+# Configure logging
 logging.basicConfig(
     level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
 )
 logger = logging.getLogger(__name__)
 
+# Define paths relative to this file
+SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
+SYSTEM_PROMPT_PATH = os.path.join(SCRIPT_DIR, "prompts/system_prompt.txt")
+FEEDSTOCK_PROMPT_PATH = os.path.join(SCRIPT_DIR, "prompts/feedstock_prompt.txt")
+FEEDSTOCK_CRITERIA_PATH = os.path.join(SCRIPT_DIR, "prompts/feedstock_criteria.json")
+
 
 def read_markdown_file(filepath: str) -> str:
     """
@@ -65,10 +76,7 @@ def parse_audit_report(self, audit_report_path, output_folder_path=None, overwri
         return audit_report_dict
 
     def extract_audit_report_data(self, report_content):
-        # load the system prompt file
-        script_dir = os.path.dirname(os.path.abspath(__file__))
-        prompt_path = os.path.join(script_dir, "system_prompt.txt")
-        with open(prompt_path, "r") as file:
+        with open(SYSTEM_PROMPT_PATH, "r") as file:
             system_prompt = file.read()
 
         user_message_template = """
@@ -109,3 +117,87 @@ def extract_audit_report_data(self, report_content):
         audit_dict = json.loads(audit_info)
 
         return audit_dict
+
+    def analyze_feedstock_sustainability(
+        self, audit_path: str, output_folder_path=None, overwrite=False
+    ) -> List[Dict]:
+        """
+        Analyze feedstock sustainability from audit report using Mistral LLM.
+
+        Args:
+            audit_path: Path to the audit report file
+            output_folder_path: Optional path to output folder. If None, uses same directory as input
+            overwrite: Whether to overwrite existing output file
+
+        Returns:
+            List of analysis results for each topic
+        """
+        # Load prompts and criteria
+        with open(FEEDSTOCK_PROMPT_PATH, "r") as f:
+            context_content = f.read()
+        with open(audit_path, "r") as file:
+            audit_report = file.read()
+
+        criteria_guideline = load_json_file(FEEDSTOCK_CRITERIA_PATH)
+
+        # Template for the full message
+        full_message_template = """
+        {context_content}
+
+        **Criteria Guideline**
+
+        Topic: {topic}
+        {questions}
+
+        **Audit report**
+
+        {audit_report}
+        """
+
+        result_list = []
+
+        # Process each topic in the criteria guideline
+        for topic in criteria_guideline.keys():
+            start_time = time.time()
+            questions = criteria_guideline.get(topic)
+
+            full_message = full_message_template.format(
+                context_content=context_content,
+                topic=topic,
+                questions=questions,
+                audit_report=audit_report,
+            )
+
+            messages = [{"role": "user", "content": full_message}]
+
+            chat_response = self.mistral_client.chat.complete(
+                model=self.model, messages=messages, response_format={"type": "json_object"}
+            )
+
+            response_content = chat_response.choices[0].message.content
+            response_content_dict = json.loads(response_content)
+            result_list.append(response_content_dict)
+
+            logger.info(f'Processed topic "{topic}" in {time.time() - start_time:.2f}s')
+            time.sleep(1)  # Rate limiting
+
+        # Determine output path
+        input_dir = os.path.dirname(audit_path)
+        input_filename = os.path.basename(audit_path)
+        output_filename = os.path.splitext(input_filename)[0] + "_feedstock_analysis.json"
+
+        output_path = os.path.join(
+            output_folder_path if output_folder_path else input_dir, output_filename
+        )
+
+        # Check if file exists and overwrite is False
+        if os.path.exists(output_path) and not overwrite:
+            logger.info(f"Output file already exists at {output_path} and overwrite=False")
+            return result_list
+
+        # Save the JSON file
+        with open(output_path, "w", encoding="utf-8") as f:
+            json.dump(result_list, f, indent=2)
+
+        logger.info(f"Feedstock analysis complete. Results saved to {output_path}")
+        return result_list