# scripts/generate_pr_description.py
import subprocess
import sys
import os
from openai import OpenAI

def get_diff(base_branch: str = "main") -> str:
    result = subprocess.run(
        ["git", "diff", f"origin/{base_branch}...HEAD"],
        capture_output=True,
        text=True
    )
    return result.stdout[:8000]  # token limit safety

def generate_description(diff: str) -> str:
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {
                "role": "system",
                "content": (
                    "You are a senior engineer writing PR descriptions. "
                    "Be concise and technical. Use markdown bullet points. "
                    "Focus on what changed and why, not how."
                )
            },
            {
                "role": "user",
                "content": f"Write a PR description for this diff:\n\n{diff}"
            }
        ],
        max_tokens=600,
        temperature=0.2
    )
    return response.choices[0].message.content

def update_pr_via_gh_cli(body: str):
    pr_number = subprocess.run(
        ["gh", "pr", "view", "--json", "number", "-q", ".number"],
        capture_output=True,
        text=True
    ).stdout.strip()

    subprocess.run(
        ["gh", "pr", "edit", pr_number, "--body", body],
        check=True
    )

if __name__ == "__main__":
    base = sys.argv[1] if len(sys.argv) > 1 else "main"
    diff = get_diff(base)
    description = generate_description(diff)
    update_pr_via_gh_cli(description)
    print("PR description updated.")

SYSTEM_PROMPT = """You write PR descriptions for a team that uses this template:

## TL;DR
One sentence.

## What changed
- Bullet per logical change

## Why
One sentence justification.

## Testing
- How to verify

## Breaking changes
None / [list with migration steps]

Rules:
- Skip sections with nothing to say
- Never explain implementation details
- Always flag breaking changes explicitly
- Keep total length under 250 words"""

def summarize_large_diff(diff: str, max_chunk_lines: int = 200) -> str:
 lines = diff.split("\n")
 file_chunks = []
 current_chunk = []

 for line in lines:
 if line.startswith("diff --git") and current_chunk:
 file_chunks.append("\n".join(current_chunk))
 current_chunk = [line]
 else:
 current_chunk.append(line)

 if current_chunk:
 file_chunks.append("\n".join(current_chunk))

 file_summaries = []
 for chunk in file_chunks:
 summary = client.messages.create(
 model="claude-opus-4-6",
 max_tokens=200,
 messages=[{
 "role": "user",
 "content": f"Summarize this file diff in 2-3 bullets:\n\n{chunk[:3000]}"
 }]
 )
 file_summaries.append(summary.content[0].text)

 combined = "\n\n".join(file_summaries)
 final = client.messages.create(
 model="claude-opus-4-6",
 max_tokens=400,
 messages=[{
 "role": "user",
 "content": f"Write a cohesive PR description from these file summaries:\n\n{combined}"
 }]
 )
 return final.content[0].text

import subprocess

def get_pr_template() -> str:
 try:
 with open(".github/PULL_REQUEST_TEMPLATE.md") as f:
 return f.read()
 except FileNotFoundError:
 return ""

PR_TEMPLATE = get_pr_template()

SYSTEM_PROMPT = f"""Fill in this PR template based on the diff provided.
Keep all section headers exactly as written.
If a section is not applicable, write 'N/A' — do not delete the section.

Template:
{PR_TEMPLATE}"""