ci(python): gate the Windows x64 installed wheel

Add node24-win-x64 to the required pull-request and public-release matrices on a native windows-2025 runner, and publish the same win_amd64 artifact from the GitLab tag pipeline. GitHub uses Git Bash for the shared release script while selecting the Windows venv's Scripts/python.exe explicitly; the Linux and macOS legs retain their existing commands and native checks.

Run the complete installed-wheel keyless suite and the trusted two-turn DeepSeek smoke on Windows exactly as on the existing targets. Make the minimal blackbox choose persistent PowerShell on Windows, keep advanced and restart snapshots platform-stable by disabling both one-shot shell variants, locate the generated dsh.exe console command, and validate text lines without assuming POSIX newlines.

Workflow tests pin the four-target matrix, Windows runner and wheel tag, cross-platform venv selection, GitLab publication dependency, and full blackbox invocation. The existing POSIX minimal snapshot changes only its platform-neutral prompt wording; Windows owns a separate model-visible snapshot.
This commit is contained in:
Tianyi Cui
2026-08-24 19:09:40 +08:00
parent ca0b21661e
commit 026a37fc07
8 changed files with 586 additions and 48 deletions
+36 -20
View File
@@ -30,7 +30,7 @@ CODE_PROMPT = "Use run_code to compute the packaged worker smoke value."
CODE_WORKER_TEXT = "code worker smoke ok"
WORKFLOW_PROMPT = "Use workflow to compute the packaged worker smoke value without agents."
WORKFLOW_WORKER_TEXT = "workflow worker smoke ok"
MINIMAL_PROMPT = "Exercise the packaged minimal agent's persistent Bash and string-replacement editor."
MINIMAL_PROMPT = "Exercise the packaged minimal agent's persistent shell and string-replacement editor."
MINIMAL_TEXT = "minimal agent smoke ok"
MINIMAL_EDITOR_PATH_PREFIX = "Editor path: "
FS_SEARCH_PROMPT = "Exercise the packaged filesystem search tools."
@@ -41,11 +41,20 @@ MCP_TEXT = "MCP client smoke ok"
PROFILE_PLUGIN_PROMPT = "Verify the Python-installed dsh profile plugin."
PROFILE_PLUGIN_TEXT = "profile plugin smoke ok"
PROFILE_PLUGIN_MARKER = "PYTHON_INSTALLED_DSH_PROFILE_PLUGIN"
MINIMAL_BASH_COMMAND = (
"counter=$(( ${counter:-0} + 1 )); export counter; "
"printf 'COUNT=%s CWD=%s\\n' \"$counter\" \"$PWD\"; "
"if [ \"$counter\" -eq 1 ]; then cd /tmp; fi"
IS_WINDOWS = sys.platform == "win32"
MINIMAL_SHELL_TOOL = "pwsh" if IS_WINDOWS else "bash"
MINIMAL_SHELL_COMMAND = (
"$global:dshSdkCounter = [int]$global:dshSdkCounter + 1; "
'Write-Output "COUNT=$global:dshSdkCounter CWD=$((Get-Location).Path)"; '
"if ($global:dshSdkCounter -eq 1) { Set-Location $env:TEMP }"
if IS_WINDOWS
else (
"counter=$(( ${counter:-0} + 1 )); export counter; "
"printf 'COUNT=%s CWD=%s\\n' \"$counter\" \"$PWD\"; "
"if [ \"$counter\" -eq 1 ]; then cd /tmp; fi"
)
)
MINIMAL_SHELL_SECOND_CWD = str(Path(tempfile.gettempdir()).resolve()) if IS_WINDOWS else "/tmp"
LEGACY_CUSTOM_DISABLED_ROWS = (
"agent-instructions",
"goal",
@@ -108,6 +117,8 @@ ADVANCED_SNAPSHOT_FILENAMES = ("result.json", "session.jsonl", "session.1.jsonl"
MINIMAL_SNAPSHOT_DIRECTORY = (
Path(__file__).resolve().parent / "snapshots" / "python-sdk-single-exe" / "minimal"
)
if IS_WINDOWS:
MINIMAL_SNAPSHOT_DIRECTORY /= "win-x64"
MINIMAL_SNAPSHOT_FILENAMES = ("model-visible.json",)
RESTART_SNAPSHOT_DIRECTORY = (
Path(__file__).resolve().parent / "snapshots" / "python-sdk-single-exe" / "restart"
@@ -301,8 +312,8 @@ def completion_chunks(body: dict[str, object]) -> list[dict[str, object]]:
if minimal_prompt is not None:
return tool_call_chunks(
"minimal-bash-1",
"bash",
{"command": MINIMAL_BASH_COMMAND},
MINIMAL_SHELL_TOOL,
{"command": MINIMAL_SHELL_COMMAND},
)
scenario_prompts = {
SNAPSHOT_DIRECT_CHILD_PROMPT,
@@ -438,17 +449,18 @@ def minimal_tool_followup(
"""Verify the checked-in minimal composition's PTY and editor."""
if not call_id.startswith("minimal-"):
return None
if call_id == "minimal-bash-1" and tool_name == "bash":
if call_id == "minimal-bash-1" and tool_name == MINIMAL_SHELL_TOOL:
if "COUNT=1" not in tool_text:
raise AssertionError(f"first persistent bash call lost its output: {tool_text}")
raise AssertionError(f"first persistent shell call lost its output: {tool_text}")
return tool_call_chunks(
"minimal-bash-2",
"bash",
{"command": MINIMAL_BASH_COMMAND},
MINIMAL_SHELL_TOOL,
{"command": MINIMAL_SHELL_COMMAND},
)
if call_id == "minimal-bash-2" and tool_name == "bash":
if "COUNT=2 CWD=/tmp" not in tool_text:
raise AssertionError(f"persistent bash did not retain state: {tool_text}")
if call_id == "minimal-bash-2" and tool_name == MINIMAL_SHELL_TOOL:
expected = f"COUNT=2 CWD={MINIMAL_SHELL_SECOND_CWD}"
if expected.lower() not in tool_text.lower():
raise AssertionError(f"persistent shell did not retain state: {tool_text}")
messages = body.get("messages")
if not isinstance(messages, list):
raise AssertionError("persistent editor smoke request has no messages")
@@ -800,8 +812,9 @@ def smoke_sdk_live() -> None:
sessions = dsh_home / "sessions"
marker = root / "live-api-marker.txt"
session_id = "installed-wheel-live-api"
shell_tool = "pwsh" if IS_WINDOWS else "bash"
create_prompt = (
"Use the bash tool to create the file at the absolute path below with exactly one line "
f"Use the {shell_tool} tool to create the file at the absolute path below with exactly one line "
f"containing {LIVE_API_SENTINEL}. Then reply with exactly {LIVE_API_SENTINEL}.\n{marker}"
)
verify_prompt = (
@@ -845,8 +858,8 @@ def smoke_sdk_live() -> None:
raise AssertionError(f"{label} turn returned {result.final_response!r}")
if not marker.is_file():
raise AssertionError(f"real-model tool turn did not create {marker}")
if marker.read_bytes() != f"{LIVE_API_SENTINEL}\n".encode():
raise AssertionError(f"real-model tool turn wrote unexpected bytes to {marker}")
if marker.read_text(encoding="utf-8").splitlines() != [LIVE_API_SENTINEL]:
raise AssertionError(f"real-model tool turn wrote unexpected text to {marker}")
assert_zstd_session_log(sessions)
@@ -921,6 +934,7 @@ def smoke_sdk_custom(base_url: str, executable: Path) -> None:
{"id": "session-log-deepseek", "config": {"enabled": True}},
*({"id": row_id, "disabled": True} for row_id in LEGACY_CUSTOM_DISABLED_ROWS),
{"id": "tool-bash", "disabled": True},
{"id": "tool-pwsh", "disabled": True},
{
"id": "tool-subagent",
"config": {
@@ -989,7 +1003,7 @@ def smoke_sdk_minimal(base_url: str, executable: Path, update_snapshots: bool) -
raise AssertionError(f"minimal agent run emitted no final response: {result.events}")
if editor_path.read_text() != "created by packaged editor\n":
raise AssertionError(f"packaged editor wrote unexpected content: {editor_path.read_text()!r}")
assert_session_log(sessions, root, MINIMAL_TEXT, "COUNT=1", "COUNT=2 CWD=/tmp")
assert_session_log(sessions, root, MINIMAL_TEXT, "COUNT=1", "COUNT=2")
files = build_minimal_snapshot_files(MockModelHandler.requests[first_request:], root)
compare_snapshot_files(
@@ -1105,7 +1119,7 @@ def smoke_sdk_profile_plugin(base_url: str) -> None:
"insert": [{"id": "python-sdk-blackbox-plugin", "name": "dsh-python-blackbox-plugin"}],
}], indent=2))
dsh = Path(sysconfig.get_path("scripts")) / "dsh"
dsh = Path(sysconfig.get_path("scripts")) / ("dsh.exe" if IS_WINDOWS else "dsh")
environment = {**os.environ, "DSH_HOME": str(dsh_home)}
installed = subprocess.run(
[str(dsh), "plugin", "--profile", "sdk", "add", f"file:{plugin}"],
@@ -1170,6 +1184,7 @@ def smoke_sdk_snapshot(base_url: str, executable: Path, update_snapshots: bool)
{"id": "session-log-deepseek", "config": {"enabled": True}},
*({"id": row_id, "disabled": True} for row_id in LEGACY_CUSTOM_DISABLED_ROWS),
{"id": "tool-bash", "disabled": True},
{"id": "tool-pwsh", "disabled": True},
{
"id": "tool-subagent",
"config": {
@@ -1243,6 +1258,7 @@ def smoke_sdk_restart_snapshot(base_url: str, executable: Path, update_snapshots
{"id": "session-log-deepseek", "config": {"enabled": True}},
*({"id": row_id, "disabled": True} for row_id in LEGACY_CUSTOM_DISABLED_ROWS),
{"id": "tool-bash", "disabled": True},
{"id": "tool-pwsh", "disabled": True},
{
"id": "tool-subagent",
"config": {
@@ -1758,7 +1774,7 @@ def compare_snapshot_files(
if update:
directory.mkdir(parents=True, exist_ok=True)
for name, content in files.items():
(directory / name).write_text(content, encoding="utf-8")
(directory / name).write_text(content, encoding="utf-8", newline="\n")
print(f"smoke-python-runtime: updated snapshots in {directory}")
existing = {