Add Tracing (#203)

* add telemetry

* better v1 cgroups

* version bump and trace cli version

* fix not accepting pathlike

Author: CTY-git

patchwork/app.py

@@ -115,6 +115,9 @@ def list_option_callback(ctx: click.Context, param: click.Parameter, value: str
     "data_format", "--format", type=click.Choice(["yaml", "json"]), default="json", help="Format of the output file."
 )
 @click.option("patched_api_key", "--patched_api_key", help="API key to use with the patched.codes service.")
+@click.option(
+    "disable_telemetry", "--disable-telemetry", is_flag=True, help="Disable telemetry.", default=False
+)
 def cli(
     log: str,
     patchflow: str,
@@ -123,6 +126,7 @@ def cli(
     output: str | None,
     data_format: str,
     patched_api_key: str | None,
+    disable_telemetry: bool,
 ):
     if "::" in patchflow:
         module_path, _, patchflow_name = patchflow.partition("::")
@@ -187,7 +191,11 @@ def cli(
 
     try:
         repo = Repo(Path.cwd(), search_parent_directories=True)
-        with PatchedClient(inputs.get("patched_api_key")).patched_telemetry(patchflow_name, repo, {}):
+        patched = PatchedClient(inputs.get("patched_api_key"))
+        if not disable_telemetry:
+            patched.send_public_telemetry(patchflow_name, inputs)
+
+        with patched.patched_telemetry(patchflow_name, repo, {}):
             patchflow_instance = patchflow_class(inputs)
             patchflow_instance.run()
     except Exception as e:

patchwork/common/client/patched.py

@@ -1,7 +1,16 @@
+import asyncio
 import atexit
 import contextlib
+import hashlib
+import http.client
+import json
+import platform
 import socket
 import sys
+import uuid
+from importlib import metadata
+from threading import Thread
+from typing import Any
 
 import click
 import requests
@@ -10,7 +19,7 @@
 from requests.adapters import DEFAULT_POOLBLOCK, HTTPAdapter
 from urllib3 import HTTPConnectionPool, HTTPSConnectionPool, PoolManager
 
-from patchwork.common.utils import get_current_branch
+from patchwork.common.utils import get_current_branch, is_container
 from patchwork.logger import logger
 
 
@@ -64,6 +73,7 @@ def init_poolmanager(self, connections, maxsize, block=DEFAULT_POOLBLOCK, **pool
 class PatchedClient(click.ParamType):
     TOKEN_URL = "https://app.patched.codes/signin"
     DEFAULT_PATCH_URL = "https://patchwork.patched.codes"
+    ALLOWED_TELEMETRY_KEYS = {"model",}
 
     def __init__(self, access_token: str, url: str = DEFAULT_PATCH_URL):
         self.access_token = access_token
@@ -119,6 +129,41 @@ def test_token(self) -> bool:
 
         return body["msg"] == "ok"
 
+    def __handle_telemetry_inputs(self, inputs: dict[str, Any]) -> dict:
+        diff_keys = set(inputs.keys()).difference(self.ALLOWED_TELEMETRY_KEYS)
+
+        inputs_copy = inputs.copy()
+        for key in diff_keys:
+            inputs_copy[key] = True
+
+        return inputs_copy
+
+    async def _public_telemetry(self, patchflow: str, inputs: dict[str, Any]):
+        requests.post(
+            url=self.url + "/v1/telemetry/",
+            headers={"Authorization": f"Bearer {self.access_token}"},
+            json=dict(
+                client_id=hashlib.sha256(str(uuid.getnode()).encode()).hexdigest(),
+                patchflow=patchflow,
+                inputs=self.__handle_telemetry_inputs(inputs),
+                environment=dict(
+                    system=platform.system(),
+                    release=platform.release(),
+                    machine=platform.machine(),
+                    python_version=platform.python_version(),
+                    cli_version=metadata.version("patchwork-cli"),
+                    is_container=is_container(),
+                ),
+            ),
+        )
+
+    def send_public_telemetry(self, patchflow: str, inputs: dict):
+        try:
+            _thread = Thread(target=asyncio.run, args=(self._public_telemetry(patchflow, inputs),))
+            _thread.start()
+        except Exception as e:
+            logger.debug(f"Failed to send public telemetry: {e}")
+
     @contextlib.contextmanager
     def patched_telemetry(self, patchflow: str, repo: Repo, inputs: dict):
         if not self.access_token:

patchwork/common/utils.py

@@ -166,3 +166,24 @@ def get_current_branch(repo: Repo) -> Head:
 
 def get_required_keys(cls: TypedDict) -> set:
     return getattr(cls, "__required_keys__", set())
+
+
+def is_container() -> bool:
+    test_files = ["/.dockerenv", "/run/.containerenv"]
+    if any(Path(file).exists() for file in test_files):
+        return True
+
+    cgroup_v1 = Path("/proc/self/cgroup")
+    if cgroup_v1.exists():
+        with cgroup_v1.open() as f:
+            lines = f.readlines()
+            for line in lines:
+                # format is `hierachy_id:controllers:pathname`
+                # cgroup v2 is `0::/`
+                hierachy_id, _, rest = line.partition(":")
+                controllers, _, pathname = rest.partition(":")
+                if hierachy_id != "0" and len(controllers) > 0:
+                    return True
+
+    # TODO: cgroup v2 detection
+    return False

patchwork/steps/CallCode2Prompt/CallCode2Prompt.py

@@ -1,5 +1,6 @@
 import os
 import subprocess
+from pathlib import Path
 
 from patchwork.logger import logger
 from patchwork.step import Step
@@ -19,7 +20,7 @@ def __init__(self, inputs: dict):
         self.folder_path = inputs[FOLDER_PATH]
         self.filter = inputs.get("filter", None)
         self.suppress_comments = inputs.get("suppress_comments", False)
-        self.code_file_path = str(self.folder_path / "README.md")
+        self.code_file_path = str(Path(self.folder_path) / "README.md")
         # Check if the file exists
         if not os.path.exists(self.code_file_path):
             # The file does not exist, create it by opening it in append mode and then closing it

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "patchwork-cli"
-version = "0.0.16"
+version = "0.0.17"
 description = ""
 authors = ["patched.codes"]
 license = "AGPL"