feat(bitrisescript): scan bitrise.log for PERFHERDER_DATA and dump it…

… to stdout

Perfherder is hardcoded to scan `public/live.log` for PERFHERDER_DATA.
Since this shows up in `<workflow>/bitrise.log`, we need logic to
forward these lines over to the main log.

bitrisescript/src/bitrisescript/bitrise.py

@@ -1,6 +1,7 @@
 import asyncio
 import logging
 import os
+from pathlib import Path
 from pprint import pformat
 from typing import Any
 
@@ -183,6 +184,30 @@ async def download_log(build_slug: str, artifacts_dir: str, poll_interval: int =
         log.error(f"Bitrise has no log for build '{build_slug}'. Please check https://app.bitrise.io/build/{build_slug}")
 
 
+async def dump_perfherder_data(artifacts_dir: str) -> None:
+    """Dumps any detected PERFHERDER_DATA log lines to stdout.
+
+    Perfherder is hardcoded to parse `public/live_backing.log` for lines that
+    start with PERFHERDER_DATA. Scan the bitrise log and forward any such lines
+    to stdout so Perfherder can find them.
+
+    Once bug 1646502 is fixed and all tasks relying on this function have been
+    migrated over to the new ingestion format, we can remove this.
+
+    Args:
+        artifacts_dir (str): Path to the artifact directory.
+    """
+    path = Path(artifacts_dir) / "bitrise.log"
+    if not path.is_file():
+        log.warning(f"Not scanning for Perfherder data, {path} does not exist!")
+        return
+
+    perfherder_data_lines = [line for line in path.read_text().splitlines() if line.startswith("PERFHERDER_DATA")]
+    if perfherder_data_lines:
+        perfherder_data_str = "\n".join(perfherder_data_lines)
+        log.info(f"Found Perfherder data in {path}:\n{perfherder_data_str}")
+
+
 async def download_file(download_url: str, file_destination: str, chunk_size: int = 512) -> None:
     """Download a file.
 
@@ -240,3 +265,4 @@ async def run_build(artifacts_dir: str, workflow_id: str, **build_params: Any) -
     finally:
         log.info(f"Retrieving bitrise log for '{build_slug}'...")
         await download_log(build_slug, artifacts_dir)
+        await dump_perfherder_data(artifacts_dir)

bitrisescript/tests/test_bitrise.py

@@ -1,6 +1,8 @@
-from asyncio import Future
 import inspect
+import logging
+from asyncio import Future
 from contextlib import nullcontext as does_not_raise
+from textwrap import dedent
 
 import pytest
 import pytest_asyncio
@@ -228,6 +230,55 @@ async def test_download_log(mocker, client):
     assert m_download.call_args == ((log_url, f"{artifacts_dir}/bitrise.log"),)
 
 
+@pytest.mark.asyncio
+async def test_dump_perfherder_data(mocker, caplog):
+    caplog.set_level(logging.INFO)
+    artifacts_dir = "artifacts"
+
+    # bitrise.log doesn't exist
+    mock_is_file = mocker.patch.object(bitrise.Path, "is_file")
+    mock_is_file.return_value = False
+    await bitrise.dump_perfherder_data(artifacts_dir)
+    assert "Not scanning for Perfherder data" in caplog.text
+
+    # bitrise.log doesn't contain Perfherder data
+    caplog.clear()
+    mock_is_file.return_value = True
+    mock_read_text = mocker.patch.object(bitrise.Path, "read_text")
+    mock_read_text.return_value = dedent(
+        """
+    INFO does not contain
+    DEBUG any perfherder data
+    """
+    ).strip()
+    await bitrise.dump_perfherder_data(artifacts_dir)
+    assert "Not scanning for Perfherder data" not in caplog.text
+    assert "Found Perfherder data in" not in caplog.text
+
+    # bitrise.log contains Perfherder data
+    caplog.clear()
+    mock_read_text.return_value = dedent(
+        """
+    INFO does contain
+    PERFHERDER_DATA {"foo": "bar"}
+    DEBUG perfherder data
+    PERFHERDER_DATA {"baz": 1}
+    """
+    ).strip()
+    await bitrise.dump_perfherder_data(artifacts_dir)
+    assert "Not scanning for Perfherder data" not in caplog.text
+    assert (
+        dedent(
+            """
+        Found Perfherder data in artifacts/bitrise.log:
+        PERFHERDER_DATA {"foo": "bar"}
+        PERFHERDER_DATA {"baz": 1}
+        """
+        ).strip()
+        in caplog.text
+    )
+
+
 @pytest.mark.asyncio
 async def test_download_file(responses, tmp_path):
     url = "https://example.com/log.txt"