sunholo 0.59.2__py3-none-any.whl → 0.59.3__py3-none-any.whl

sunholo/bots/github_webhook.py

@@ -0,0 +1,264 @@
+# from https://github.com/ray-project/docu-mentor
+import base64
+import httpx
+from dotenv import load_dotenv
+import jwt
+import os
+import time
+
+load_dotenv()
+
+
+
+APP_ID = os.environ.get("APP_ID")
+PRIVATE_KEY = os.environ.get("PRIVATE_KEY", "")
+
+# with open('private-key.pem', 'r') as f:
+#     PRIVATE_KEY = f.read()
+
+def generate_jwt():
+    payload = {
+        "iat": int(time.time()),
+        "exp": int(time.time()) + (10 * 60),
+        "iss": APP_ID,
+    }
+    if PRIVATE_KEY:
+        jwt_token = jwt.encode(payload, PRIVATE_KEY, algorithm="RS256")
+        return jwt_token
+    raise ValueError("PRIVATE_KEY not found.")
+
+
+async def get_installation_access_token(jwt, installation_id):
+    url = f"https://api.github.com/app/installations/{installation_id}/access_tokens"
+    headers = {
+        "Authorization": f"Bearer {jwt}",
+        "Accept": "application/vnd.github.v3+json",
+    }
+    async with httpx.AsyncClient() as client:
+        response = await client.post(url, headers=headers)
+        return response.json()["token"]
+
+
+def get_diff_url(pr):
+    """GitHub 302s to this URL."""
+    original_url = pr.get("url")
+    parts = original_url.split("/")
+    owner, repo, pr_number = parts[-4], parts[-3], parts[-1]
+    return f"https://patch-diff.githubusercontent.com/raw/{owner}/{repo}/pull/{pr_number}.diff"
+
+
+async def get_branch_files(pr, branch, headers):
+    original_url = pr.get("url")
+    parts = original_url.split("/")
+    owner, repo = parts[-4], parts[-3]
+    url = f"https://api.github.com/repos/{owner}/{repo}/git/trees/{branch}?recursive=1"
+    async with httpx.AsyncClient() as client:
+        response = await client.get(url, headers=headers)
+        tree = response.json().get('tree', [])
+        files = {}
+        for item in tree:
+            if item['type'] == 'blob':
+                file_url = item['url']
+                print(file_url)
+                file_response = await client.get(file_url, headers=headers)
+                content = file_response.json().get('content', '')
+                # Decode the base64 content
+                decoded_content = base64.b64decode(content).decode('utf-8')
+                files[item['path']] = decoded_content
+        return files
+
+
+async def get_pr_head_branch(pr, headers):
+    original_url = pr.get("url")
+    parts = original_url.split("/")
+    owner, repo, pr_number = parts[-4], parts[-3], parts[-1]
+    url = f"https://api.github.com/repos/{owner}/{repo}/pulls/{pr_number}"
+
+    async with httpx.AsyncClient() as client:
+        response = await client.get(url, headers=headers)
+
+        # Check if the response is successful
+        if response.status_code != 200:
+            print(f"Error: Received status code {response.status_code}")
+            print("Response body:", response.text)
+            return ''
+
+        # Safely get the 'ref'
+        data = response.json()
+        head_data = data.get('head', {})
+        ref = head_data.get('ref', '')
+        return ref
+
+
+def files_to_diff_dict(diff):
+    files_with_diff = {}
+    current_file = None
+    for line in diff.split("\n"):
+        if line.startswith("diff --git"):
+            current_file = line.split(" ")[2][2:]
+            files_with_diff[current_file] = {"text": []}
+        elif line.startswith("+") and not line.startswith("+++"):
+            files_with_diff[current_file]["text"].append(line[1:])
+    return files_with_diff
+
+
+def parse_diff_to_line_numbers(diff):
+    files_with_line_numbers = {}
+    current_file = None
+    line_number = 0
+    for line in diff.split("\n"):
+        if line.startswith("diff --git"):
+            current_file = line.split(" ")[2][2:]
+            files_with_line_numbers[current_file] = []
+            line_number = 0
+        elif line.startswith("@@"):
+            line_number = int(line.split(" ")[2].split(",")[0][1:]) - 1
+        elif line.startswith("+") and not line.startswith("+++"):
+            files_with_line_numbers[current_file].append(line_number)
+            line_number += 1
+        elif not line.startswith("-"):
+            line_number += 1
+    return files_with_line_numbers
+
+
+def get_context_from_files(files, files_with_line_numbers, context_lines=2):
+    context_data = {}
+    for file, lines in files_with_line_numbers.items():
+        file_content = files[file].split("\n")
+        context_data[file] = []
+        for line in lines:
+            start = max(line - context_lines, 0)
+            end = min(line + context_lines + 1, len(file_content))
+            context_data[file].append('\n'.join(file_content[start:end]))
+    return context_data
+
+app = FastAPI()
+
+
+async def handle_webhook(request: Request):
+    data = await request.json()
+
+    installation = data.get("installation")
+    if installation and installation.get("id"):
+        installation_id = installation.get("id")
+        logger.info(f"Installation ID: {installation_id}")
+
+        JWT_TOKEN = generate_jwt()
+
+        installation_access_token = await get_installation_access_token(
+            JWT_TOKEN, installation_id
+        )
+
+        headers = {
+            "Authorization": f"token {installation_access_token}",
+            "User-Agent": "docu-mentor-bot",
+            "Accept": "application/vnd.github.VERSION.diff",
+        }
+    else:
+        raise ValueError("No app installation found.")
+
+    # If PR exists and is opened
+    if "pull_request" in data.keys() and (
+        data["action"] in ["opened", "reopened"]
+    ):  # use "synchronize" for tracking new commits
+        pr = data.get("pull_request")
+
+        # Greet the user and show instructions.
+        async with httpx.AsyncClient() as client:
+            await client.post(
+                f"{pr['issue_url']}/comments",
+                json={"body": GREETING},
+                headers=headers,
+            )
+        return JSONResponse(content={}, status_code=200)
+
+    # Check if the event is a new or modified issue comment
+    if "issue" in data.keys() and data.get("action") in ["created", "edited"]:
+        issue = data["issue"]
+
+        # Check if the issue is a pull request
+        if "/pull/" in issue["html_url"]:
+            pr = issue.get("pull_request")
+
+            # Get the comment body
+            comment = data.get("comment")
+            comment_body = comment.get("body")
+            # Remove all whitespace characters except for regular spaces
+            comment_body = comment_body.translate(
+                str.maketrans("", "", string.whitespace.replace(" ", ""))
+            )
+
+            # Skip if the bot talks about itself
+            author_handle = comment["user"]["login"]
+
+            # Check if the bot is mentioned in the comment
+            if (
+                author_handle != "docu-mentor[bot]"
+                and "@docu-mentor run" in comment_body
+            ):
+                async with httpx.AsyncClient() as client:
+                    # Fetch diff from GitHub
+                    files_to_keep = comment_body.replace(
+                        "@docu-mentor run", ""
+                    ).split(" ")
+                    files_to_keep = [item for item in files_to_keep if item]
+
+                    logger.info(files_to_keep)
+
+                    url = get_diff_url(pr)
+                    diff_response = await client.get(url, headers=headers)
+                    diff = diff_response.text
+
+                    files_with_lines = parse_diff_to_line_numbers(diff)
+
+                    # Get head branch of the PR
+                    headers["Accept"] = "application/vnd.github.full+json"
+                    head_branch = await get_pr_head_branch(pr, headers)
+
+                    # Get files from head branch
+                    head_branch_files = await get_branch_files(pr, head_branch, headers)
+                    print("HEAD FILES", head_branch_files)
+
+                    # Enrich diff data with context from the head branch.
+                    context_files = get_context_from_files(head_branch_files, files_with_lines)
+
+                    # Filter the dictionary
+                    if files_to_keep:
+                        context_files = {
+                            k: context_files[k]
+                            for k in context_files
+                            if any(sub in k for sub in files_to_keep)
+                        }
+
+                    # Get suggestions from Docu Mentor
+                    content, model, prompt_tokens, completion_tokens = \
+                        ray_mentor(context_files) if ray.is_initialized() else mentor(context_files)
+
+
+                    # Let's comment on the PR
+                    await client.post(
+                        f"{comment['issue_url']}/comments",
+                        json={
+                            "body": f":rocket: Docu Mentor finished "
+                            + "analysing your PR! :rocket:\n\n"
+                            + "Take a look at your results:\n"
+                            + f"{content}\n\n"
+                            + "This bot is  powered by "
+                            + "[Sunholo Multivac](https://www.sunholo.com/).\n"
+                            + f"It used the model {model}, used {prompt_tokens} prompt tokens, "
+                            + f"and {completion_tokens} completion tokens in total."
+                        },
+                        headers=headers,
+                    )
+
+@serve.deployment(route_prefix="/")
+@serve.ingress(app)
+class ServeBot:
+    @app.get("/")
+    async def root(self):
+        return {"message": "Docu Mentor reporting for duty!"}
+
+    @app.post("/webhook/")
+    async def handle_webhook_route(self, request: Request):
+        return await handle_webhook(request)
+

sunholo/utils/config_schema.py

@@ -52,7 +52,8 @@ VAC_SUBCONFIG_SCHEMA = {
                 "cluster": {"type": "string"},
                 "instance": {"type": "string"},
                 "database": {"type": "string"}
-            }
+            },
+            "required": ["project_id", "region", "cluster", "instance", "database"]
         },
         "secrets": {
             "type": "array",
@@ -87,7 +88,7 @@ VAC_CONFIG_SCHEMA = {
             }
         }
     },
-    "required": ["kind", "apiVersion", "gcp_config", "vac"]
+    "required": ["kind", "apiVersion", "vac"]
 }
 
 PROMPT_CONFIG_SCHEMA = {

{sunholo-0.59.2.dist-info → sunholo-0.59.3.dist-info}/METADATA

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: sunholo
-Version: 0.59.2
+Version: 0.59.3
 Summary: Large Language Model DevOps - a package to help deploy LLMs to the Cloud.
 Home-page: https://github.com/sunholo-data/sunholo-py
-Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.59.2.tar.gz
+Download-URL: https://github.com/sunholo-data/sunholo-py/archive/refs/tags/v0.59.3.tar.gz
 Author: Holosun ApS
 Author-email: multivac@sunholo.com
 License: Apache License, Version 2.0

{sunholo-0.59.2.dist-info → sunholo-0.59.3.dist-info}/RECORD

@@ -20,6 +20,7 @@ sunholo/auth/__init__.py,sha256=4owDjSaWYkbTlPK47UHTOC0gCWbZsqn4ZIEw5NWZTlg,28
 sunholo/auth/run.py,sha256=4fhNhDgtMtBHc08AywDmtazQPQ2560at6EMLQbMwTIo,2846
 sunholo/bots/__init__.py,sha256=EMFd7e2z68l6pzYOnkzHbLd2xJRvxTKFRNCTuhZ8hIw,130
 sunholo/bots/discord.py,sha256=cCFae5K1BCa6JVkWGLh_iZ9qFO1JpXb6K4eJrlDfEro,2442
+sunholo/bots/github_webhook.py,sha256=5pQPRLM_wxxcILVaIzUDV8Kt7Arcm2dL1r1kMMHA524,9629
 sunholo/bots/webapp.py,sha256=EIMxdAJ_xtufwJmvnn7N_Fb_1hZ9DjhJ0Kf_hp02vEU,1926
 sunholo/chunker/__init__.py,sha256=UhQBZTKwDfBXm0TPv4LvsGc5pdUGCbYzi3qPTOkU4gw,55
 sunholo/chunker/data_to_embed_pubsub.py,sha256=t-pWNYv2mnwVAkMcIOK2CrIb3yr2aS9iAdtryk7hT8o,2931
@@ -85,14 +86,14 @@ sunholo/summarise/summarise.py,sha256=C3HhjepTjUhUC8FLk4jMQIBvq1BcORniwuTFHjPVhV
 sunholo/utils/__init__.py,sha256=G11nN_6ATjxpuMfG_BvcUr9UU8onPIgkpTK6CjOcbr8,48
 sunholo/utils/big_context.py,sha256=qHYtds4Ecf9eZRHVqXho4_q8Je7HD44-vS6RJ6s9Z0Q,5387
 sunholo/utils/config.py,sha256=Ve1sb68Av9_SPGqXs33g5FAJSIQ3GODoeuUCW3MNCwU,8802
-sunholo/utils/config_schema.py,sha256=bx3SHHuZ3SCOOXNRU91Mk-b4pHXt7D-EElv7Q85gAdw,3758
+sunholo/utils/config_schema.py,sha256=Rkw5nVHcCtIQH_sH5ZUiDaxW1TOjUmzsDwHhrfiWeqQ,3829
 sunholo/utils/gcp.py,sha256=B2G1YKjeD7X9dqO86Jrp2vPuFwZ223Xl5Tg09Ndw-oc,5760
 sunholo/utils/parsers.py,sha256=OrHmASqIbI45atVOhiGodgLvnfrzkvVzyHnSvAXD89I,3841
 sunholo/vertex/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 sunholo/vertex/init_vertex.py,sha256=JDMUaBRdednzbKF-5p33qqLit2LMsvgvWW-NRz0AqO0,1801
-sunholo-0.59.2.dist-info/LICENSE.txt,sha256=SdE3QjnD3GEmqqg9EX3TM9f7WmtOzqS1KJve8rhbYmU,11345
-sunholo-0.59.2.dist-info/METADATA,sha256=YcmQZ5MGbmH8xVGiJJEqVuK0DEr8ak2BM3nsQl3gfWc,7903
-sunholo-0.59.2.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-sunholo-0.59.2.dist-info/entry_points.txt,sha256=bZuN5AIHingMPt4Ro1b_T-FnQvZ3teBes-3OyO0asl4,49
-sunholo-0.59.2.dist-info/top_level.txt,sha256=wt5tadn5--5JrZsjJz2LceoUvcrIvxjHJe-RxuudxAk,8
-sunholo-0.59.2.dist-info/RECORD,,
+sunholo-0.59.3.dist-info/LICENSE.txt,sha256=SdE3QjnD3GEmqqg9EX3TM9f7WmtOzqS1KJve8rhbYmU,11345
+sunholo-0.59.3.dist-info/METADATA,sha256=XAF0rooghlKcIPwDYbNNHt6dv5k06WF_PyV1t6GKl9E,7903
+sunholo-0.59.3.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+sunholo-0.59.3.dist-info/entry_points.txt,sha256=bZuN5AIHingMPt4Ro1b_T-FnQvZ3teBes-3OyO0asl4,49
+sunholo-0.59.3.dist-info/top_level.txt,sha256=wt5tadn5--5JrZsjJz2LceoUvcrIvxjHJe-RxuudxAk,8
+sunholo-0.59.3.dist-info/RECORD,,