PATCHED_API_KEY integration

patchwork/app.py

@@ -1,5 +1,6 @@
 import importlib
 import json
+import traceback
 from pathlib import Path
 
 import click
@@ -73,9 +74,13 @@ def cli(log: str, patchflow: str, opts: list[str], config: str | None, output: s
         else:
             # treat --key=value as a key-value pair
             inputs[key] = value
-
-    patchflow_instance = patchflow_class(inputs)
-    patchflow_instance.run()
+    try:
+        patchflow_instance = patchflow_class(inputs)
+        patchflow_instance.run()
+    except Exception as e:
+        logger.debug(traceback.format_exc())
+        logger.error(f"Error running patchflow {patchflow}: {e}")
+        exit(1)
 
     data_format_mapping = {
         "yaml": yaml.dump,

patchwork/common/client/scm.py

@@ -137,8 +137,13 @@ def set_pr_description(self, body: str) -> None:
         self._mr.save()
 
     def create_comment(
-        self, path: str, body: str, start_line: int | None = None, end_line: int | None = None
+        self, body: str, path: str | None = None, start_line: int | None = None, end_line: int | None = None
     ) -> str | None:
+        final_body = f"{_COMMENT_MARKER} \n{PullRequestProtocol._apply_pr_template(self, body)}"
+        if path is None:
+            note = self._mr.notes.create({"body": final_body})
+            return f"#note_{note.get_id()}"
+
         while True:
             try:
                 commit = self._mr.commits().next()
@@ -161,7 +166,6 @@ def create_comment(
         head_commit = diff.head_commit_sha
 
         try:
-            final_body = f"{_COMMENT_MARKER} \n{PullRequestProtocol._apply_pr_template(self, body)}"
             discussion = self._mr.discussions.create(
                 {
                     "body": final_body,
@@ -187,14 +191,19 @@ def create_comment(
         return None
 
     def reset_comments(self) -> None:
-        for discussion in self._mr.discussions.list():
+        for discussion in self._mr.discussions.list(iterator=True):
             for note in discussion.attributes["notes"]:
-                if note["type"] == "DiffNote" and note["body"].startswith(_COMMENT_MARKER):
+                if note["body"].startswith(_COMMENT_MARKER):
                     discussion.notes.delete(note["id"])
 
     def file_diffs(self) -> dict[str, str]:
-        files = self._mr.diffs.list()
-        return {file.attributes["new_path"]: file.attributes["diff"] for file in files}
+        diffs = self._mr.diffs.list()
+        latest_diff = max(diffs, key=lambda diff: diff.created_at, default=None)
+        if latest_diff is None:
+            return {}
+
+        files = self._mr.diffs.get(latest_diff.id).diffs
+        return {file["new_path"]: file["diff"] for file in files}
 
 
 class GithubPullRequest(PullRequestProtocol):
@@ -336,7 +345,7 @@ def get_pr_by_url(self, url: str) -> PullRequestProtocol | None:
             logger.error(f"Invalid PR URL: {url}")
             return None
 
-        slug = "/".join(url_parts[-4:-2])
+        slug = "/".join(url_parts[-5:-3])
 
         return self.find_pr_by_id(slug, int(pr_id))
 

patchwork/steps/CallOpenAI/CallOpenAI.py

@@ -10,9 +10,12 @@
 from patchwork.logger import logger
 from patchwork.step import Step
 
+_TOKEN_URL = "https://app.patched.codes/signin"
+_DEFAULT_PATCH_URL = "https://patchwork.patched.codes/v1"
+
 
 class CallOpenAI(Step):
-    required_keys = {"openai_api_key", "prompt_file"}
+    required_keys = {"prompt_file"}
 
     def __init__(self, inputs: dict):
         logger.info(f"Run started {self.__class__.__name__}")
@@ -27,7 +30,29 @@ def __init__(self, inputs: dict):
         self.model_args = {key[len("model_") :]: value for key, value in inputs.items() if key.startswith("model_")}
         self.client_args = {key[len("client_") :]: value for key, value in inputs.items() if key.startswith("client_")}
 
-        self.openai_api_key = inputs["openai_api_key"]
+        openai_key = inputs.get("openai_api_key") or os.environ.get("OPENAI_API_KEY")
+        if openai_key is not None:
+            self.openai_api_key = openai_key
+
+        patched_key = inputs.get("patched_api_key")
+        if patched_key is not None:
+            self.openai_api_key = patched_key
+            self.client_args["base_url"] = _DEFAULT_PATCH_URL
+
+        if self.openai_api_key is None:
+            raise ValueError(
+                f"Model API key not found.\n"
+                f'Please login at: "{_TOKEN_URL}",\n'
+                "Please go to the Integration's tab and generate an API key.\n"
+                "Please copy the access token that is generated, "
+                "and add `--patched_api_key=<token>` to the command line.\n"
+                "\n"
+                "If you are using a OpenAI API Key, please set `--openai_api_key=<token>`.\n"
+            )
+
+        if not self.openai_api_key:
+            raise ValueError('Missing required data: "openai_api_key"')
+
         self.prompt_file = Path(inputs["prompt_file"])
         if not self.prompt_file.is_file():
             raise ValueError(f'Unable to find Prompt file: "{self.prompt_file}"')

patchwork/steps/ExtractCode/ExtractCode.py

@@ -93,6 +93,76 @@ def resolve_artifact_location(
     return None
 
 
+def transform_sarif_results(
+    sarif_data: dict, base_path: Path, context_length: int, vulnerability_limit: int
+) -> dict[tuple[str, int, int, int], list[str]]:
+    # Process each result in SARIF data
+    grouped_messages = defaultdict(list)
+    vulnerability_count = 0
+    for run_idx, run in enumerate(sarif_data.get("runs", [])):
+        artifact_locations = [
+            parse_sarif_location(base_path, artifact["location"]["uri"]) for artifact in run.get("artifacts", [])
+        ]
+
+        for result_idx, result in enumerate(run.get("results", [])):
+            for location_idx, location in enumerate(result.get("locations", [])):
+                physical_location = location.get("physicalLocation", {})
+
+                artifact_location = physical_location.get("artifactLocation", {})
+                uri = resolve_artifact_location(base_path, artifact_location, artifact_locations)
+                if uri is None:
+                    logger.warn(
+                        f'Unable to find file for ".runs[{run_idx}].results[{result_idx}].locations[{location_idx}]"'
+                    )
+                    continue
+
+                region = physical_location.get("region", {})
+                start_line = region.get("startLine", 1)
+                end_line = region.get("endLine", start_line)
+                start_line = start_line - 1
+
+                # Generate file path assuming code is in the current working directory
+                file_path = str(uri.relative_to(base_path))
+
+                # Extract lines from the code file
+                logger.info(f"Extracting context for {file_path} at {start_line}:{end_line}")
+                try:
+                    with open_with_chardet(file_path, "r") as file:
+                        src = file.read()
+
+                    source_lines = src.splitlines(keepends=True)
+                    context_start, context_end = get_source_code_context(
+                        file_path, source_lines, start_line, end_line, context_length
+                    )
+
+                    source_code_context = None
+                    if context_start is not None and context_end is not None:
+                        source_code_context = "".join(source_lines[context_start:context_end])
+
+                except FileNotFoundError:
+                    context_start = None
+                    context_end = None
+                    source_code_context = None
+                    logger.info(f"File not found in the current working directory: {file_path}")
+
+                if source_code_context is None:
+                    logger.info(f"No context found for {file_path} at {start_line}:{end_line}")
+                    continue
+
+                start = context_start if context_start is not None else start_line
+                end = context_end if context_end is not None else end_line
+
+                grouped_messages[(uri, start, end, source_code_context)].append(
+                    result.get("message", {}).get("text", "")
+                )
+
+                vulnerability_count = vulnerability_count + 1
+                if 0 < vulnerability_limit <= vulnerability_count:
+                    return grouped_messages
+
+    return grouped_messages
+
+
 class ExtractCode(Step):
     required_keys = {"sarif_file_path"}
 
@@ -112,7 +182,6 @@ def __init__(self, inputs: dict):
         self.vulnerability_limit = inputs.get("vulnerability_limit", 10)
 
         # Prepare for data extraction
-        self.extracted_data = []
         self.extracted_code_contexts = []
 
     def run(self) -> dict:
@@ -122,77 +191,8 @@ def run(self) -> dict:
 
         vulnerability_count = 0
         base_path = Path.cwd()
-        # Process each result in SARIF data
-        grouped_messages = defaultdict(list)
-        for run_idx, run in enumerate(sarif_data.get("runs", [])):
-            artifact_locations = [
-                parse_sarif_location(base_path, artifact["location"]["uri"]) for artifact in run.get("artifacts", [])
-            ]
-
-            for result_idx, result in enumerate(run.get("results", [])):
-                for location_idx, location in enumerate(result.get("locations", [])):
-                    physical_location = location.get("physicalLocation", {})
-
-                    artifact_location = physical_location.get("artifactLocation", {})
-                    uri = resolve_artifact_location(base_path, artifact_location, artifact_locations)
-                    if uri is None:
-                        logger.warn(
-                            f'Unable to find file for ".runs[{run_idx}].results[{result_idx}].locations[{location_idx}]"'
-                        )
-                        continue
-
-                    region = physical_location.get("region", {})
-                    start_line = region.get("startLine", 1)
-                    end_line = region.get("endLine", start_line)
-                    start_line = start_line - 1
-
-                    # Generate file path assuming code is in the current working directory
-                    file_path = str(uri.relative_to(base_path))
-
-                    # Extract lines from the code file
-                    logger.info(f"Extracting context for {file_path} at {start_line}:{end_line}")
-                    try:
-                        with open_with_chardet(file_path, "r") as file:
-                            src = file.read()
-
-                        source_lines = src.splitlines(keepends=True)
-                        context_start, context_end = get_source_code_context(
-                            file_path, source_lines, start_line, end_line, self.context_length
-                        )
-
-                        source_code_context = None
-                        if context_start is not None and context_end is not None:
-                            source_code_context = "".join(source_lines[context_start:context_end])
-
-                    except FileNotFoundError:
-                        context_start = None
-                        context_end = None
-                        source_code_context = None
-                        logger.info(f"File not found in the current working directory: {file_path}")
-
-                    if source_code_context is None:
-                        logger.info(f"No context found for {file_path} at {start_line}:{end_line}")
-                        continue
-
-                    start = context_start if context_start is not None else start_line
-                    end = context_end if context_end is not None else end_line
-                    self.extracted_data.append(
-                        {
-                            "affectedCode": source_code_context,
-                            "startLine": start,
-                            "endLine": end,
-                            "uri": file_path,
-                            "messageText": result.get("message", {}).get("text", ""),
-                        }
-                    )
-
-                    grouped_messages[(uri, start, end, source_code_context)].append(
-                        result.get("message", {}).get("text", "")
-                    )
 
-                    vulnerability_count = vulnerability_count + 1
-                    if 0 < self.vulnerability_limit <= vulnerability_count:
-                        break
+        grouped_messages = transform_sarif_results(sarif_data, base_path, self.context_length, self.vulnerability_limit)
 
         self.extracted_code_contexts = [
             {

patchwork/steps/ModifyCode/ModifyCode.py

@@ -74,7 +74,9 @@ def run(self) -> dict:
         code_snippets = load_json_file(self.code_snippets_path)
 
         modified_code_files = []
-        sorted_list = sorted(zip(code_snippets, self.extracted_responses), key=lambda x: x[0]["startLine"], reverse=True)
+        sorted_list = sorted(
+            zip(code_snippets, self.extracted_responses), key=lambda x: x[0]["startLine"], reverse=True
+        )
         for code_snippet, extracted_response in sorted_list:
             uri = code_snippet["uri"]
             start_line = code_snippet["startLine"]

patchwork/steps/ReadPRDiffs/ReadPRDiffs.py

@@ -5,6 +5,26 @@
 from patchwork.logger import logger
 from patchwork.step import Step
 
+_IGNORED_EXTENSIONS = [
+    ".png",
+    ".jpg",
+    ".jpeg",
+    ".gif",
+    ".svg",
+    ".pdf",
+    ".docx",
+    ".xlsx",
+    ".pptx",
+    ".zip",
+    ".tar",
+    ".gz",
+    ".lock",
+]
+
+
+def filter_by_extension(file, extensions):
+    return any(file.endswith(ext) for ext in extensions)
+
 
 class ReadPRDiffs(Step):
     required_keys = {"pr_url"}
@@ -30,6 +50,8 @@ def __init__(self, inputs: dict):
     def run(self) -> dict:
         prompt_values = []
         for path, diffs in self.pr.file_diffs().items():
+            if filter_by_extension(path, _IGNORED_EXTENSIONS):
+                continue
             prompt_values.append(dict(path=path, diff=diffs))
 
         prompt_value_file = tempfile.mktemp(".json")