add src/main.py

src/main.py

@@ -0,0 +1,104 @@
+"""Main module for task 69dd4221f309a98be0006b2e.
+
+This module defines a Pydantic model representing a task card and a function
+`parse_task_description` that takes a natural‑language description of a
+task and returns a validated instance of the model.
+
+The implementation uses LangChain's structured output parser to guarantee
+that the LLM returns data in the expected format.
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import List, Optional
+
+from langchain_core.prompts import PromptTemplate
+from langchain_openai import ChatOpenAI
+from langchain_core.output_parsers import PydanticOutputParser
+from pydantic import BaseModel, Field
+
+# ---------------------------------------------------------------------------
+# Pydantic model
+# ---------------------------------------------------------------------------
+class TaskCard(BaseModel):
+    """Structured representation of a task description.
+
+    The field names are chosen to be concise yet expressive.  All fields are
+    optional because the LLM may not mention every piece of information.
+    """
+
+    title: Optional[str] = Field(None, description="Short title of the task")
+    subject: Optional[str] = Field(None, description="Subject or domain of the task")
+    deadline_hint: Optional[str] = Field(
+        None, description="Human‑readable hint about the deadline"
+    )
+    deliverable_type: Optional[str] = Field(
+        None, description="What is expected to be submitted (report, code, etc.)"
+    )
+    grading_hints: Optional[List[str]] = Field(
+        None, description="List of hints about grading criteria"
+    )
+
+# ---------------------------------------------------------------------------
+# LangChain chain
+# ---------------------------------------------------------------------------
+# LLM – you can change the model name or temperature via environment
+# variables or by passing arguments to ChatOpenAI.
+llm = ChatOpenAI(temperature=0.0)
+
+parser = PydanticOutputParser(pydantic_object=TaskCard)
+
+prompt = PromptTemplate(
+    template=(
+        "You are an assistant that extracts structured information from a short "
+        "task description. Return the data in the following JSON format: "
+        "{format_instructions}\n\nInput: {input}\nOutput:"
+    ),
+    input_variables=["input"],
+    partial_variables={"format_instructions": parser.get_format_instructions()},
+)
+
+chain = prompt | llm | parser
+
+# ---------------------------------------------------------------------------
+# Public API
+# ---------------------------------------------------------------------------
+
+def parse_task_description(description: str) -> TaskCard:
+    """Parse a natural‑language task description.
+
+    Parameters
+    ----------
+    description: str
+        One‑sentence or short paragraph describing the task.
+
+    Returns
+    -------
+    TaskCard
+        Validated Pydantic model with extracted fields.
+    """
+    return chain.invoke({"input": description})
+
+# ---------------------------------------------------------------------------
+# Demo / CLI
+# ---------------------------------------------------------------------------
+if __name__ == "__main__":
+    import argparse
+    import json
+
+    parser_cli = argparse.ArgumentParser(description="Parse a task description.")
+    parser_cli.add_argument("description", type=str, help="Task description to parse")
+    args = parser_cli.parse_args()
+
+    card = parse_task_description(args.description)
+    print("Parsed card:")
+    print(json.dumps(card.model_dump(), indent=2, ensure_ascii=False))
+    print("\nHuman‑readable summary:")
+    print(
+        f"Title: {card.title or 'N/A'}\n"
+        f"Subject: {card.subject or 'N/A'}\n"
+        f"Deadline hint: {card.deadline_hint or 'N/A'}\n"
+        f"Deliverable: {card.deliverable_type or 'N/A'}\n"
+        f"Grading hints: {', '.join(card.grading_hints) if card.grading_hints else 'N/A'}"
+    )