unsloth/tests/python/test_e2e_no_torch_sandbox.py
Leo Borcherding 1dd2fc4583
tests: read checked-in files as UTF-8 instead of the platform default (#7438)
* tests: read checked-in files as UTF-8 instead of the platform default

Path.read_text() with no encoding uses locale.getpreferredencoding(), which
is UTF-8 on the Linux runners and cp1252 on a stock Windows install. Nine
module-level reads of checked-in source files were relying on that default.

studio/backend/routes/inference.py carries the DeepSeek tool-call token
regexes, so it holds U+FF5C and U+2581. Under cp1252 that read raised
UnicodeDecodeError on byte 0x81 at position 97806, and because the reads run
at import time it took test_cancel_atomicity.py and test_cancel_id_wiring.py
out at collection, not as failures. Green on CI, permanently broken for a
Windows contributor running the suite locally.

Adds a guard: at module scope there is no tmp_path fixture, so a bare
read_text()/write_text()/open() there is always touching a checked-in file.
That makes the rule mechanical enough to enforce with no allowlist, while
staying quiet about temp-dir I/O inside test bodies where the platform
default is harmless.

The repo already spells this correctly in 464 other places; this only stops
the stragglers coming back.

* tests: cover import-time helper reads and keep the guard py3.9-safe

Follows up on the Codex review:

- add `from __future__ import annotations`, since `str | None` in
  `_offender` is evaluated at import on Python 3.9 and pyproject declares
  requires-python ">=3.9,<3.15".
- widen the guard from module scope to import time. Class bodies and the
  bodies of module-level helpers called from an executing statement run
  during collection too, so `CODE = _extract_mixed_precision_code()` was
  the same hazard as an inline read. `if __name__ == "__main__":` blocks
  are skipped: pytest never executes them.
- scan studio/backend/tests/ as well as tests/. Both trees are collected
  on Windows by separate CI jobs, and the offender that started this,
  test_tool_xml_strip.py reading routes/inference.py, lives there.

Widening it surfaced seven more import-time reads of checked-in sources;
all now name utf-8.

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Harden the import-time encoding guard for PR #7438

Close the detector gaps raised in review, all of which I reproduced against
the actual AST before changing anything.

False negatives (the guard let a real hazard through):
- _is_main_guard ignored the comparison operator, so if __name__ != "__main__"
  counted as script-only even though its body runs at import.
- The else arm of a main guard was discarded with the rest of the If node.
- Decorators and argument defaults on a module-level def were skipped with the
  body, though both are evaluated when the def executes.
- Path.open() in text mode was invisible; only builtin open() was matched.
- encoding = None and encoding = "locale" both re-select the platform default,
  but the keyword merely being present counted as pinned.

False positives (the guard would have blocked a compliant contributor):
- A non-literal mode fell through to the "r" default, so open(p, mode) was
  flagged even when mode is "rb", where adding encoding= is a ValueError and
  there is no edit that satisfies the rule.
- Same for open(*args) and a **kwargs splat, which hide the mode and can hide
  an encoding.
- Lambda bodies and comprehension elements were walked even though neither runs
  at definition.

Verified: still reports the same 22 offenders on unpatched main, green on this
branch and on the tree merged with latest main (557 files), and an adversarial
corpus of 33 cases now scores zero false positives and zero false negatives.
Also corrected two docstring claims: neither collecting job runs on Windows,
and the read is governed by locale.getencoding().

* Walk eager comprehensions and treat io.open as the builtin

Two regressions from the previous commit, both reproduced against the AST
before changing anything.

Lumping list, set and dict comprehensions in with generator expressions was
wrong. Only a genexp is lazy; the other three run their element expression,
their filters and their nested iterators immediately, so
CONTENTS = [p.read_text() for p in PATHS] at module scope is an import-time
read the guard was silently missing. Comprehensions are now walked in full and
only the genexp keeps the outermost-iterable-only treatment.

io was also in the not-a-path-opener list, but io.open is the builtin, with the
same mode position and the same platform default. io.open(CHECKED_IN_FILE) is
exactly the hazard this guard exists for, so it is matched now, with binary
modes and a pinned encoding still exempt. tarfile.open and fitz.open stay
exempt since neither has an encoding to name.

Verified: 13 targeted cases covering all five eager comprehension forms and
io.open in text, binary and pinned shapes all classify correctly; still 22
offenders on unpatched main; green on this branch and on the tree merged with
latest main.

* Close three more walker gaps in the import-time guard

All three reproduced against the AST first.

A generator expression handed straight to a call is consumed there, so
DATA = "".join(p.read_text() for p in paths) runs its element at import. Only
an unconsumed genexp bound to a name stays lazy, so the walker now follows the
consumed ones in full and keeps the outermost-iterable-only treatment for the
rest.

if "__main__" == __name__ is an equivalent and accepted spelling of the main
guard, but requiring __name__ on the left meant its body was treated as
import-time code. That is a false positive on a block pytest never runs, so
both operand orders are recognised now.

The helper table was built from module-level defs only, so a def in a class
body invoked while the class is constructed was never followed, contradicting
the walker's stated coverage of class bodies. Helpers are now collected from
the module body and from class bodies at any nesting.

Verified: 15 targeted cases including all three fixes and the earlier ones
still classify correctly; still 22 offenders on unpatched main; green on this
branch and on the tree merged with latest main.

* Handle positional read_text encodings, lazy generators and nested helpers

* Guard reads reached from test bodies, unbound Path calls and __file__ paths

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Follow derived paths, skip lazy generator helpers, cover compressed openers

* Guard the CLI tests, helper parameters and unbound Path arguments

* Discover test roots and follow literal, in-place and tuple-derived paths

* Identify module openers by import, unwrap starred paths, pin subprocess snippets

* Resolve import origins, seed helper locals, follow named generators and parametrize

* Scope imports lexically, list tracked test files, bind unpacked names

* Resolve aliased openers, keyword-only params, destructured targets, next()

* Pin the encoding on subprocess snippets, workflow lint and CLI output for PR #7438

* Harden the CLI encoding guard against detached streams for PR #7438

* Tighten the encoding guard's path and scope analysis for PR #7438

* Resolve path provenance more precisely and keep POSIX stream encodings for PR #7438

* Resolve qualified path classes and scope conditional imports for PR #7438

* Scope CLI stream setup to the entry point and align two encoding pairs for PR #7438

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
2026-07-26 23:31:56 -07:00

1129 lines
43 KiB
Python

"""E2E sandbox tests for PR #4624: lazy torch imports, CPU fallback, install.sh parsing, NO_TORCH filtering, live server."""
from __future__ import annotations
import os
import shutil
import signal
import subprocess
import sys
import textwrap
import time
from pathlib import Path
from unittest import mock
import pytest
REPO_ROOT = Path(__file__).resolve().parents[2]
STUDIO_DIR = REPO_ROOT / "studio"
BACKEND_DIR = STUDIO_DIR / "backend"
DATASETS_DIR = BACKEND_DIR / "utils" / "datasets"
HARDWARE_DIR = BACKEND_DIR / "utils" / "hardware"
INSTALL_SH = REPO_ROOT / "install.sh"
INSTALL_PY = STUDIO_DIR / "install_python_stack.py"
DATA_COLLATORS = DATASETS_DIR / "data_collators.py"
CHAT_TEMPLATES = DATASETS_DIR / "chat_templates.py"
FORMAT_DETECTION = DATASETS_DIR / "format_detection.py"
MODEL_MAPPINGS = DATASETS_DIR / "model_mappings.py"
VLM_PROCESSING = DATASETS_DIR / "vlm_processing.py"
ITERABLE = DATASETS_DIR / "iterable.py"
HARDWARE_PY = HARDWARE_DIR / "hardware.py"
# Unsloth venv for server tests
STUDIO_VENV = Path.home() / ".unsloth" / "studio" / "unsloth_studio"
sys.path.insert(0, str(STUDIO_DIR))
def _venv_python(venv_dir: Path) -> Path:
"""Return a venv's Python executable path, cross-platform."""
if sys.platform == "win32":
return venv_dir / "Scripts" / "python.exe"
return venv_dir / "bin" / "python"
def _has_uv() -> bool:
return shutil.which("uv") is not None
def _create_no_torch_venv(venv_dir: Path, python_version: str = "3.12") -> Path | None:
"""Create a uv venv with no torch. Returns python path or None."""
result = subprocess.run(
["uv", "venv", str(venv_dir), "--python", python_version],
capture_output = True,
)
if result.returncode != 0:
return None
py = _venv_python(venv_dir)
if not py.exists():
return None
check = subprocess.run([str(py), "-c", "import torch"], capture_output = True)
if check.returncode == 0:
return None
return py
def _run_in_sandbox(
py: str | Path,
code: str,
timeout: int = 60,
env: dict | None = None,
) -> subprocess.CompletedProcess:
"""Run Python code in a sandboxed interpreter."""
return subprocess.run(
[str(py), "-c", code],
capture_output = True,
timeout = timeout,
env = env,
)
def _run_sh(script: str, timeout: int = 30) -> subprocess.CompletedProcess:
"""Run a bash snippet and return the result."""
return subprocess.run(
["bash", "-c", script],
capture_output = True,
timeout = timeout,
)
def _write_loggers_stub(sandbox: Path) -> None:
"""Create a minimal loggers package stub (replaces the structlog-backed real one)."""
loggers_dir = sandbox / "loggers"
loggers_dir.mkdir(exist_ok = True)
(loggers_dir / "__init__.py").write_text(
"from .handlers import get_logger\n__all__ = ['get_logger']\n",
encoding = "utf-8",
)
(loggers_dir / "handlers.py").write_text(
textwrap.dedent("""\
class _Logger:
def info(self, msg, *a, **k): pass
def warning(self, msg, *a, **k): pass
def debug(self, msg, *a, **k): pass
def error(self, msg, *a, **k): pass
def msg(self, msg, *a, **k): pass
def get_logger(name=None):
return _Logger()
"""),
encoding = "utf-8",
)
def _write_structlog_stub(sandbox: Path) -> None:
"""Create a minimal structlog stub."""
structlog_dir = sandbox / "structlog"
structlog_dir.mkdir(exist_ok = True)
(structlog_dir / "__init__.py").write_text(
textwrap.dedent("""\
class _Logger:
def info(self, msg, *a, **k): pass
def warning(self, msg, *a, **k): pass
def debug(self, msg, *a, **k): pass
def error(self, msg, *a, **k): pass
def msg(self, msg, *a, **k): pass
def get_logger(name=None):
return _Logger()
"""),
encoding = "utf-8",
)
def _write_hardware_stub(sandbox: Path) -> None:
"""Create utils/hardware stub with dataset_map_num_proc."""
hw_dir = sandbox / "utils" / "hardware"
hw_dir.mkdir(parents = True, exist_ok = True)
(sandbox / "utils" / "__init__.py").write_text("", encoding = "utf-8")
(hw_dir / "__init__.py").write_text(
"def dataset_map_num_proc(n=None): return n\n",
encoding = "utf-8",
)
@pytest.fixture(scope = "session")
def repo_root():
return REPO_ROOT
@pytest.fixture
def sandbox_dir(tmp_path):
"""Per-test temporary sandbox directory."""
return tmp_path
@pytest.fixture(params = ["3.12", "3.13"], scope = "module")
def no_torch_venv(request, tmp_path_factory):
"""Temporary uv venv with no torch; 3.12 = Intel Mac default, 3.13 = Apple Silicon/Linux."""
if not _has_uv():
pytest.skip("uv not available")
py_version = request.param
venv_dir = tmp_path_factory.mktemp(f"e2e_no_torch_{py_version}")
py = _create_no_torch_venv(venv_dir, py_version)
if py is None:
pytest.skip(f"Could not create Python {py_version} no-torch venv")
return str(py)
# Group 1: BEFORE vs AFTER -- Import Chain
class TestBeforeAfterImportChain:
"""BEFORE (synthetic top-level torch import) crashes; AFTER (lazy imports) works."""
# -- BEFORE: crashes --
def test_before_chat_templates_crashes(self, no_torch_venv, sandbox_dir):
"""BEFORE: chat_templates.py with top-level IterableDataset import crashes without torch."""
source = CHAT_TEMPLATES.read_text(encoding = "utf-8")
before_source = "from torch.utils.data import IterableDataset\n" + source
before_file = sandbox_dir / "chat_templates_before.py"
before_file.write_text(before_source, encoding = "utf-8")
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: type('L', (), {{'info': lambda s, m: None}})()
sys.modules['loggers'] = loggers
fd = types.ModuleType('format_detection')
fd.detect_dataset_format = fd.detect_multimodal_dataset = fd.detect_custom_format_heuristic = lambda *a, **k: None
sys.modules['format_detection'] = fd
mm = types.ModuleType('model_mappings')
mm.MODEL_TO_TEMPLATE_MAPPER = {{}}
sys.modules['model_mappings'] = mm
source = open({str(before_file)!r}, encoding = "utf-8").read()
source = source.replace('from .format_detection import', 'from format_detection import')
source = source.replace('from .model_mappings import', 'from model_mappings import')
exec(source)
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode != 0, "BEFORE chat_templates.py should crash without torch"
assert b"ModuleNotFoundError" in result.stderr or b"ImportError" in result.stderr
def test_before_data_collators_crashes(self, no_torch_venv, sandbox_dir):
"""BEFORE: data_collators.py with top-level 'import torch' crashes."""
source = DATA_COLLATORS.read_text(encoding = "utf-8")
before_source = "import torch\n" + source
before_file = sandbox_dir / "data_collators_before.py"
before_file.write_text(before_source, encoding = "utf-8")
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: None
sys.modules['loggers'] = loggers
exec(open({str(before_file)!r}, encoding = "utf-8").read())
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode != 0, "BEFORE data_collators.py should crash without torch"
assert b"ModuleNotFoundError" in result.stderr or b"ImportError" in result.stderr
def test_before_full_import_chain_crashes(self, no_torch_venv, sandbox_dir):
"""BEFORE: full utils/datasets/ package with top-level torch imports crashes."""
_write_loggers_stub(sandbox_dir)
_write_hardware_stub(sandbox_dir)
pkg_dir = sandbox_dir / "utils" / "datasets"
pkg_dir.mkdir(parents = True, exist_ok = True)
# Copy torch-free modules as-is
shutil.copy2(FORMAT_DETECTION, pkg_dir / "format_detection.py")
shutil.copy2(MODEL_MAPPINGS, pkg_dir / "model_mappings.py")
shutil.copy2(VLM_PROCESSING, pkg_dir / "vlm_processing.py")
# BEFORE data_collators: prepend top-level 'import torch'
dc_source = DATA_COLLATORS.read_text(encoding = "utf-8")
(pkg_dir / "data_collators.py").write_text(
"import torch\n" + dc_source,
encoding = "utf-8",
)
# BEFORE chat_templates: prepend top-level IterableDataset import
ct_source = CHAT_TEMPLATES.read_text(encoding = "utf-8")
(pkg_dir / "chat_templates.py").write_text(
"from torch.utils.data import IterableDataset\n" + ct_source,
encoding = "utf-8",
)
(pkg_dir / "__init__.py").write_text(
textwrap.dedent("""\
from .format_detection import detect_dataset_format
from .data_collators import DataCollatorSpeechSeq2SeqWithPadding
from .chat_templates import DEFAULT_ALPACA_TEMPLATE
"""),
encoding = "utf-8",
)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
from utils.datasets import detect_dataset_format
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode != 0, "BEFORE full import chain should crash without torch"
assert b"ModuleNotFoundError" in result.stderr or b"ImportError" in result.stderr
# -- AFTER: succeeds --
def test_after_chat_templates_imports(self, no_torch_venv):
"""AFTER: PR branch chat_templates.py imports fine without torch."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: type('L', (), {{'info': lambda s, m: None}})()
sys.modules['loggers'] = loggers
fd = types.ModuleType('format_detection')
fd.detect_dataset_format = fd.detect_multimodal_dataset = fd.detect_custom_format_heuristic = lambda *a, **k: None
sys.modules['format_detection'] = fd
mm = types.ModuleType('model_mappings')
mm.MODEL_TO_TEMPLATE_MAPPER = {{}}
sys.modules['model_mappings'] = mm
it = types.ModuleType('iterable')
it.is_streaming_dataset = lambda *a, **k: False
sys.modules['iterable'] = it
source = open({str(CHAT_TEMPLATES)!r}, encoding = "utf-8").read()
source = source.replace('from .format_detection import', 'from format_detection import')
source = source.replace('from .model_mappings import', 'from model_mappings import')
source = source.replace('from .iterable import', 'from iterable import')
exec(source)
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert (
result.returncode == 0
), f"AFTER chat_templates.py should work without torch:\n{result.stderr.decode()}"
assert b"OK" in result.stdout
def test_after_data_collators_imports(self, no_torch_venv):
"""AFTER: PR branch data_collators.py imports fine without torch."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: None
sys.modules['loggers'] = loggers
exec(open({str(DATA_COLLATORS)!r}, encoding = "utf-8").read())
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert (
result.returncode == 0
), f"AFTER data_collators.py should work without torch:\n{result.stderr.decode()}"
assert b"OK" in result.stdout
def test_after_full_import_chain_imports(self, no_torch_venv, sandbox_dir):
"""AFTER: full utils/datasets/ package imports fine without torch."""
_write_loggers_stub(sandbox_dir)
_write_hardware_stub(sandbox_dir)
pkg_dir = sandbox_dir / "utils" / "datasets"
pkg_dir.mkdir(parents = True, exist_ok = True)
# Copy AFTER versions (PR branch -- no top-level torch)
for src in [
FORMAT_DETECTION,
MODEL_MAPPINGS,
VLM_PROCESSING,
DATA_COLLATORS,
CHAT_TEMPLATES,
ITERABLE,
]:
if src.exists():
shutil.copy2(src, pkg_dir / src.name)
(pkg_dir / "__init__.py").write_text(
textwrap.dedent("""\
from .format_detection import detect_dataset_format, detect_custom_format_heuristic
from .model_mappings import MODEL_TO_TEMPLATE_MAPPER
from .chat_templates import DEFAULT_ALPACA_TEMPLATE, get_dataset_info_summary
from .data_collators import (
DataCollatorSpeechSeq2SeqWithPadding,
DeepSeekOCRDataCollator,
VLMDataCollator,
)
from .vlm_processing import generate_smart_vlm_instruction
"""),
encoding = "utf-8",
)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
from utils.datasets import (
detect_dataset_format,
DEFAULT_ALPACA_TEMPLATE,
DataCollatorSpeechSeq2SeqWithPadding,
DeepSeekOCRDataCollator,
VLMDataCollator,
generate_smart_vlm_instruction,
)
assert 'Instruction' in DEFAULT_ALPACA_TEMPLATE
print("OK: full import chain succeeded")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert (
result.returncode == 0
), f"AFTER full import chain should work:\n{result.stderr.decode()}"
assert b"OK: full import chain succeeded" in result.stdout
# Group 2: Dataclass Instantiation
class TestDataclassInstantiation:
"""Dataclass collators instantiate and constants are accessible without torch."""
def test_speech_collator_instantiate(self, no_torch_venv):
"""DataCollatorSpeechSeq2SeqWithPadding(processor=None) succeeds."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: None
sys.modules['loggers'] = loggers
exec(open({str(DATA_COLLATORS)!r}, encoding = "utf-8").read())
obj = DataCollatorSpeechSeq2SeqWithPadding(processor=None)
assert obj.processor is None
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
def test_deepseek_ocr_collator_instantiate(self, no_torch_venv):
"""DeepSeekOCRDataCollator has correct default field values."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: None
sys.modules['loggers'] = loggers
exec(open({str(DATA_COLLATORS)!r}, encoding = "utf-8").read())
obj = DeepSeekOCRDataCollator(processor=None)
assert obj.processor is None
assert obj.max_length == 2048
assert obj.ignore_index == -100
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
def test_vlm_collator_instantiate(self, no_torch_venv):
"""VLMDataCollator has correct default field values."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: None
sys.modules['loggers'] = loggers
exec(open({str(DATA_COLLATORS)!r}, encoding = "utf-8").read())
obj = VLMDataCollator(processor=None)
assert obj.processor is None
assert obj.max_length == 2048
assert obj.mask_input_tokens is True
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
def test_alpaca_template_accessible(self, no_torch_venv):
"""DEFAULT_ALPACA_TEMPLATE constant is accessible and contains 'Instruction'."""
code = textwrap.dedent(f"""\
import sys, types
loggers = types.ModuleType('loggers')
loggers.get_logger = lambda n: type('L', (), {{'info': lambda s, m: None}})()
sys.modules['loggers'] = loggers
fd = types.ModuleType('format_detection')
fd.detect_dataset_format = fd.detect_multimodal_dataset = fd.detect_custom_format_heuristic = lambda *a, **k: None
sys.modules['format_detection'] = fd
mm = types.ModuleType('model_mappings')
mm.MODEL_TO_TEMPLATE_MAPPER = {{}}
sys.modules['model_mappings'] = mm
it = types.ModuleType('iterable')
it.is_streaming_dataset = lambda *a, **k: False
sys.modules['iterable'] = it
ns = {{}}
source = open({str(CHAT_TEMPLATES)!r}, encoding = "utf-8").read()
source = source.replace('from .format_detection import', 'from format_detection import')
source = source.replace('from .model_mappings import', 'from model_mappings import')
source = source.replace('from .iterable import', 'from iterable import')
exec(source, ns)
assert 'Instruction' in ns['DEFAULT_ALPACA_TEMPLATE']
print("OK")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
# Group 3: Edge Cases -- Partial/Broken Torch
class TestEdgeCasesBrokenTorch:
"""Behavior with fake or broken torch modules on sys.path."""
def test_fake_broken_torch_module(self, no_torch_venv, sandbox_dir):
"""Fake torch raising RuntimeError on import: data_collators.py (no top-level torch) still loads."""
torch_dir = sandbox_dir / "torch"
torch_dir.mkdir()
(torch_dir / "__init__.py").write_text(
'raise RuntimeError("CUDA not found")\n',
encoding = "utf-8",
)
_write_loggers_stub(sandbox_dir)
shutil.copy2(DATA_COLLATORS, sandbox_dir / "data_collators.py")
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
exec(open({str(sandbox_dir / 'data_collators.py')!r}, encoding = "utf-8").read())
obj = DataCollatorSpeechSeq2SeqWithPadding(processor=None)
print("OK: data_collators works despite broken torch on sys.path")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Should work with broken torch:\n{result.stderr.decode()}"
assert b"OK:" in result.stdout
def test_torch_import_error_hardware_fallback(self, no_torch_venv, sandbox_dir):
"""Fake torch raising ImportError: detect_hardware() falls back to CPU."""
torch_dir = sandbox_dir / "torch"
torch_dir.mkdir()
(torch_dir / "__init__.py").write_text(
'raise ImportError("No torch binary")\n',
encoding = "utf-8",
)
_write_loggers_stub(sandbox_dir)
_write_structlog_stub(sandbox_dir)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
source = open({str(HARDWARE_PY)!r}, encoding = "utf-8").read()
ns = {{'__name__': '__test__'}}
exec(source, ns)
result = ns['detect_hardware']()
assert result == ns['DeviceType'].CPU, f"Expected CPU, got {{result}}"
print("OK: detect_hardware returned CPU")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert (
result.returncode == 0
), f"detect_hardware should fallback to CPU:\n{result.stderr.decode()}"
assert b"OK: detect_hardware returned CPU" in result.stdout
def test_fake_torch_no_cuda(self, no_torch_venv, sandbox_dir):
"""Fake torch imports OK but cuda.is_available() is False: detect_hardware() falls back to CPU."""
torch_dir = sandbox_dir / "torch"
torch_dir.mkdir()
(torch_dir / "__init__.py").write_text(
textwrap.dedent("""\
class _Cuda:
@staticmethod
def is_available():
return False
cuda = _Cuda()
class version:
cuda = None
"""),
encoding = "utf-8",
)
_write_loggers_stub(sandbox_dir)
_write_structlog_stub(sandbox_dir)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
source = open({str(HARDWARE_PY)!r}, encoding = "utf-8").read()
ns = {{'__name__': '__test__'}}
exec(source, ns)
result = ns['detect_hardware']()
assert result == ns['DeviceType'].CPU, f"Expected CPU, got {{result}}"
print("OK: detect_hardware returned CPU with fake torch (no CUDA)")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Should fall back to CPU:\n{result.stderr.decode()}"
assert b"OK:" in result.stdout
def test_lazy_torch_fails_at_call_time_not_import_time(self, no_torch_venv, sandbox_dir):
"""apply_chat_template_to_dataset imports without torch; the lazy import fails at call time, not import time."""
_write_loggers_stub(sandbox_dir)
code = textwrap.dedent(f"""\
import sys, types
sys.path.insert(0, {str(sandbox_dir)!r})
fd = types.ModuleType('format_detection')
fd.detect_dataset_format = fd.detect_multimodal_dataset = fd.detect_custom_format_heuristic = lambda *a, **k: None
sys.modules['format_detection'] = fd
mm = types.ModuleType('model_mappings')
mm.MODEL_TO_TEMPLATE_MAPPER = {{}}
sys.modules['model_mappings'] = mm
it = types.ModuleType('iterable')
it.is_streaming_dataset = lambda *a, **k: False
sys.modules['iterable'] = it
ns = {{}}
source = open({str(CHAT_TEMPLATES)!r}, encoding = "utf-8").read()
source = source.replace('from .format_detection import', 'from format_detection import')
source = source.replace('from .model_mappings import', 'from model_mappings import')
source = source.replace('from .iterable import', 'from iterable import')
exec(source, ns)
# Import succeeds -- this is the fix
assert 'apply_chat_template_to_dataset' in ns
print("OK: import succeeded")
# Calling alpaca branch triggers lazy torch import inside the try block.
# The function catches the error and returns it in the errors list.
dataset_info = {{
'dataset': type('D', (), {{'map': lambda *a, **k: None}})(),
'final_format': 'alpaca',
'chat_column': None,
'is_standardized': True,
'warnings': [],
}}
result = ns['apply_chat_template_to_dataset'](dataset_info, None)
# The function has a try/except that catches the error gracefully
if not result['success']:
print("OK: call-time failure caught gracefully")
else:
print("OK: call succeeded (unexpected but not a crash)")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Should not crash at import time:\n{result.stderr.decode()}"
assert b"OK: import succeeded" in result.stdout
# Group 4: Hardware Detection Without Torch
class TestHardwareDetectionNoTorch:
"""Hardware module works without torch, falling back to CPU."""
def test_detect_hardware_no_torch(self, no_torch_venv, sandbox_dir):
"""detect_hardware() returns CPU when torch is not installed."""
_write_loggers_stub(sandbox_dir)
_write_structlog_stub(sandbox_dir)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
source = open({str(HARDWARE_PY)!r}, encoding = "utf-8").read()
ns = {{'__name__': '__test__'}}
exec(source, ns)
device = ns['detect_hardware']()
assert device == ns['DeviceType'].CPU
assert ns['CHAT_ONLY'] is True
print("OK: detect_hardware returned CPU, CHAT_ONLY=True")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
assert b"OK:" in result.stdout
def test_get_package_versions_no_torch(self, no_torch_venv, sandbox_dir):
"""get_package_versions() returns torch=None, cuda=None without torch."""
_write_loggers_stub(sandbox_dir)
_write_structlog_stub(sandbox_dir)
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
source = open({str(HARDWARE_PY)!r}, encoding = "utf-8").read()
ns = {{'__name__': '__test__'}}
exec(source, ns)
versions = ns['get_package_versions']()
assert versions['torch'] is None, f"Expected torch=None, got {{versions['torch']}}"
assert versions['cuda'] is None, f"Expected cuda=None, got {{versions['cuda']}}"
print("OK: torch=None, cuda=None")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
assert b"OK:" in result.stdout
def test_hardware_module_import_no_torch(self, no_torch_venv, sandbox_dir):
"""Hardware module imports and detect_hardware is callable without torch."""
_write_loggers_stub(sandbox_dir)
_write_structlog_stub(sandbox_dir)
_write_hardware_stub(sandbox_dir)
# Copy the real hardware module into a sandbox package
hw_sandbox = sandbox_dir / "hw_pkg"
hw_sandbox.mkdir()
(hw_sandbox / "__init__.py").write_text("", encoding = "utf-8")
shutil.copy2(HARDWARE_PY, hw_sandbox / "hardware.py")
code = textwrap.dedent(f"""\
import sys
sys.path.insert(0, {str(sandbox_dir)!r})
source = open({str(hw_sandbox / 'hardware.py')!r}, encoding = "utf-8").read()
ns = {{'__name__': '__test__'}}
exec(source, ns)
assert callable(ns['detect_hardware'])
assert callable(ns['get_package_versions'])
assert callable(ns['is_apple_silicon'])
print("OK: all hardware functions accessible")
""")
result = _run_in_sandbox(no_torch_venv, code)
assert result.returncode == 0, f"Failed:\n{result.stderr.decode()}"
assert b"OK:" in result.stdout
# Group 5: install.sh Logic (via bash subprocess)
class TestInstallShLogic:
"""install.sh flag parsing, platform detection, and guard logic."""
@pytest.fixture(autouse = True)
def _check_install_sh(self):
if not INSTALL_SH.is_file():
pytest.skip("install.sh not found")
def test_python_flag_parsing(self):
"""--python flag correctly sets _USER_PYTHON."""
script = textwrap.dedent("""\
_USER_PYTHON=""
_next_is_python=false
for arg in "$@"; do
if [ "$_next_is_python" = true ]; then
_USER_PYTHON="$arg"
_next_is_python=false
continue
fi
case "$arg" in
--python) _next_is_python=true ;;
esac
done
echo "$_USER_PYTHON"
""")
# --python 3.12
r = _run_sh(f"{script}" + "\n", timeout = 10)
r = subprocess.run(
["bash", "-c", script + "\n", "_", "--python", "3.12"],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"3.12"
# --local --python 3.11
r = subprocess.run(
["bash", "-c", script + "\n", "_", "--local", "--python", "3.11"],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"3.11"
# no --python flag
r = subprocess.run(
["bash", "-c", script + "\n", "_", "--local"],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b""
def test_python_flag_missing_arg_errors(self):
"""--python without a version argument triggers an error."""
script = textwrap.dedent("""\
set -e
_USER_PYTHON=""
_next_is_python=false
for arg in "$@"; do
if [ "$_next_is_python" = true ]; then
_USER_PYTHON="$arg"
_next_is_python=false
continue
fi
case "$arg" in
--python) _next_is_python=true ;;
esac
done
if [ "$_next_is_python" = true ]; then
echo "ERROR: --python requires a version argument" >&2
exit 1
fi
echo "$_USER_PYTHON"
""")
r = subprocess.run(
["bash", "-c", script + "\n", "_", "--python"],
capture_output = True,
timeout = 10,
)
assert r.returncode != 0
assert b"ERROR" in r.stderr
def test_python_version_resolution(self):
"""Python version defaults to 3.12 on Intel Mac, 3.13 elsewhere.
--python overrides both."""
script = textwrap.dedent("""\
MAC_INTEL="$1"
_USER_PYTHON="$2"
if [ -n "$_USER_PYTHON" ]; then
PYTHON_VERSION="$_USER_PYTHON"
elif [ "$MAC_INTEL" = true ]; then
PYTHON_VERSION="3.12"
else
PYTHON_VERSION="3.13"
fi
echo "$PYTHON_VERSION"
""")
# Intel Mac, no override
r = subprocess.run(
["bash", "-c", script + "\n", "_", "true", ""],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"3.12"
# non-Intel, no override
r = subprocess.run(
["bash", "-c", script + "\n", "_", "false", ""],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"3.13"
# Intel Mac with --python override
r = subprocess.run(
["bash", "-c", script + "\n", "_", "true", "3.11"],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"3.11"
def test_mac_intel_detection_snippet(self):
"""Architecture detection sets MAC_INTEL correctly for different platforms."""
script = textwrap.dedent("""\
OS="$1"
_ARCH="$2"
MAC_INTEL=false
if [ "$OS" = "macos" ] && [ "$_ARCH" = "x86_64" ]; then
MAC_INTEL=true
fi
echo "$MAC_INTEL"
""")
cases = [
(("macos", "x86_64"), b"true"),
(("macos", "arm64"), b"false"),
(("linux", "x86_64"), b"false"),
(("linux", "aarch64"), b"false"),
]
for (os_val, arch), expected in cases:
r = subprocess.run(
["bash", "-c", script + "\n", "_", os_val, arch],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == expected, (
f"MAC_INTEL for ({os_val}, {arch}): "
f"expected {expected!r}, got {r.stdout.strip()!r}"
)
def test_stale_venv_guard_respects_override(self):
"""When _USER_PYTHON is set, the stale venv recreation guard is skipped."""
script = textwrap.dedent("""\
MAC_INTEL=true
_USER_PYTHON="$1"
_VENV_EXISTS=true # simulate existing venv
SHOULD_RECREATE=false
if [ "$MAC_INTEL" = true ] && [ -z "$_USER_PYTHON" ] && [ "$_VENV_EXISTS" = true ]; then
SHOULD_RECREATE=true
fi
echo "$SHOULD_RECREATE"
""")
# with override: should NOT recreate
r = subprocess.run(
["bash", "-c", script + "\n", "_", "3.11"],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"false"
# without override: SHOULD recreate
r = subprocess.run(
["bash", "-c", script + "\n", "_", ""],
capture_output = True,
timeout = 10,
)
assert r.stdout.strip() == b"true"
# Group 6: install_python_stack.py NO_TORCH Filtering
class TestInstallPythonStackFiltering:
"""NO_TORCH filtering logic in install_python_stack.py."""
@pytest.fixture(autouse = True)
def _check_install_py(self):
if not INSTALL_PY.is_file():
pytest.skip("install_python_stack.py not found")
def test_filter_requirements_removes_torch_deps(self):
"""_filter_requirements removes all NO_TORCH_SKIP_PACKAGES from a real extras file."""
import install_python_stack as ips
extras = STUDIO_DIR / "backend" / "requirements" / "extras.txt"
if not extras.is_file():
pytest.skip("extras.txt not found")
result_path = ips._filter_requirements(extras, ips.NO_TORCH_SKIP_PACKAGES)
filtered = Path(result_path).read_text(encoding = "utf-8").lower()
lines = [
l.strip() for l in filtered.splitlines() if l.strip() and not l.strip().startswith("#")
]
for pkg in ips.NO_TORCH_SKIP_PACKAGES:
assert not any(
l.startswith(pkg) for l in lines
), f"{pkg} should be removed from extras.txt"
def test_filter_requirements_preserves_non_torch(self):
"""Non-torch packages survive NO_TORCH filtering."""
import install_python_stack as ips
extras = STUDIO_DIR / "backend" / "requirements" / "extras.txt"
if not extras.is_file():
pytest.skip("extras.txt not found")
result_path = ips._filter_requirements(extras, ips.NO_TORCH_SKIP_PACKAGES)
filtered_text = Path(result_path).read_text(encoding = "utf-8").lower()
must_survive = ["scikit-learn", "loguru", "tiktoken", "einops"]
original_text = extras.read_text(encoding = "utf-8").lower()
for pkg in must_survive:
if pkg in original_text:
assert pkg in filtered_text, f"{pkg} should survive NO_TORCH filtering"
def test_infer_no_torch_env_var_overrides_platform(self):
"""UNSLOTH_NO_TORCH=true on Linux -> True; =false on Intel Mac -> False."""
import install_python_stack as ips
# Explicit true on Linux
with (
mock.patch.dict(os.environ, {"UNSLOTH_NO_TORCH": "true"}),
mock.patch.object(ips, "IS_MAC_INTEL", False),
):
assert ips._infer_no_torch() is True
# explicit false on Intel Mac
with (
mock.patch.dict(os.environ, {"UNSLOTH_NO_TORCH": "false"}),
mock.patch.object(ips, "IS_MAC_INTEL", True),
):
assert ips._infer_no_torch() is False
# Unset on Intel Mac -> True (platform fallback)
env = os.environ.copy()
env.pop("UNSLOTH_NO_TORCH", None)
with (
mock.patch.dict(os.environ, env, clear = True),
mock.patch.object(ips, "IS_MAC_INTEL", True),
):
assert ips._infer_no_torch() is True
def test_no_torch_skips_overrides_and_triton(self):
"""When NO_TORCH=True, overrides.txt and triton are skipped (source guard check)."""
import install_python_stack as ips
source = Path(ips.__file__).read_text(encoding = "utf-8")
assert "if NO_TORCH:" in source, "NO_TORCH guard not found in install_python_stack.py"
# macOS guard for triton
assert (
"not IS_WINDOWS and not IS_MACOS" in source
), "'not IS_WINDOWS and not IS_MACOS' guard for triton not found"
# Group 7: Live Server Startup -- Heavyweight
def _studio_venv_python() -> Path | None:
"""Return the studio venv Python path, or None if not found."""
py = _venv_python(STUDIO_VENV)
if py.exists():
return py
return None
def _server_port() -> int:
"""Find an available port for the test server."""
import socket
with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
s.bind(("", 0))
return s.getsockname()[1]
server = pytest.mark.server
@server
class TestLiveServerStartup:
"""Live server startup against the existing Unsloth venv with torch made unimportable (pytest -m server)."""
@pytest.fixture(autouse = True)
def _check_studio_venv(self):
py = _studio_venv_python()
if py is None:
pytest.skip("Unsloth venv not found at ~/.unsloth/studio/unsloth_studio")
@pytest.fixture(scope = "class")
def server_process(self):
"""Start the studio backend server without torch, yield (proc, port), then stop."""
py = _studio_venv_python()
if py is None:
pytest.skip("Unsloth venv not found")
port = _server_port()
backend_dir = BACKEND_DIR
check = subprocess.run(
[str(py), "-c", "import torch; print(torch.__version__)"],
capture_output = True,
)
torch_was_installed = check.returncode == 0
torch_version = check.stdout.decode().strip() if torch_was_installed else None
if torch_was_installed:
subprocess.run(
[
str(py),
"-m",
"pip",
"uninstall",
"-y",
"torch",
"torchvision",
"torchaudio",
],
capture_output = True,
timeout = 120,
)
env = os.environ.copy()
env["PYTHONPATH"] = str(backend_dir)
proc = subprocess.Popen(
[str(py), str(backend_dir / "run.py"), "--port", str(port)],
env = env,
stdout = subprocess.PIPE,
stderr = subprocess.PIPE,
cwd = str(backend_dir),
)
# Wait for server to be ready (poll /api/health)
import urllib.request
import urllib.error
ready = False
for _ in range(30):
time.sleep(1)
try:
resp = urllib.request.urlopen(f"http://127.0.0.1:{port}/api/health", timeout = 2)
if resp.status == 200:
ready = True
break
except (urllib.error.URLError, ConnectionRefusedError, OSError):
continue
if not ready:
stdout, stderr = proc.communicate(timeout = 5)
if torch_was_installed and torch_version:
subprocess.run(
[
str(py),
"-m",
"pip",
"install",
f"torch=={torch_version}",
"torchvision",
"torchaudio",
],
capture_output = True,
timeout = 300,
)
server_output = stdout.decode(errors = "replace") + stderr.decode(errors = "replace")
pytest.skip(f"Server failed to start within 30 seconds. Output:\n{server_output}")
yield proc, port
# Cleanup: stop server, reinstall torch
proc.terminate()
try:
proc.wait(timeout = 10)
except subprocess.TimeoutExpired:
proc.kill()
proc.wait(timeout = 5)
if torch_was_installed and torch_version:
subprocess.run(
[
str(py),
"-m",
"pip",
"install",
f"torch=={torch_version}",
"torchvision",
"torchaudio",
],
capture_output = True,
timeout = 300,
)
def test_server_starts_without_torch(self, server_process):
"""Server responds to /api/health with chat_only: true."""
import json
import urllib.request
_, port = server_process
resp = urllib.request.urlopen(f"http://127.0.0.1:{port}/api/health", timeout = 5)
data = json.loads(resp.read())
assert data["status"] == "healthy"
assert data["chat_only"] is True
def test_all_routes_registered(self, server_process):
"""OpenAPI spec shows >= 20 paths (server started fully)."""
import json
import urllib.request
_, port = server_process
resp = urllib.request.urlopen(f"http://127.0.0.1:{port}/openapi.json", timeout = 5)
spec = json.loads(resp.read())
assert (
len(spec.get("paths", {})) >= 20
), f"Expected >= 20 routes, got {len(spec.get('paths', {}))}"
def test_hardware_endpoint_no_torch(self, server_process):
"""GET /api/system/hardware returns torch=null, gpu_name=null."""
import json
import urllib.request
_, port = server_process
resp = urllib.request.urlopen(
f"http://127.0.0.1:{port}/api/system/hardware",
timeout = 5,
)
data = json.loads(resp.read())
versions = data.get("versions", {})
assert versions.get("torch") is None
assert versions.get("cuda") is None
def test_server_survives_multiple_requests(self, server_process):
"""Hit 5 different endpoints. Server PID should still be alive after."""
import urllib.request
import urllib.error
proc, port = server_process
endpoints = [
"/api/health",
"/openapi.json",
"/api/system/hardware",
"/api/health",
"/docs",
]
for ep in endpoints:
try:
urllib.request.urlopen(f"http://127.0.0.1:{port}{ep}", timeout = 5)
except urllib.error.HTTPError:
pass # 4xx/5xx fine -- server didn't crash
except urllib.error.URLError:
pytest.fail(f"Server stopped responding at {ep}")
assert proc.poll() is None, "Server process should still be running"