mirror of
https://github.com/TencentCloud/Octop.git
synced 2026-10-31 08:03:19 +08:00
jubaoliangandCursor
5e7a8ba956
Ship a coordinated polish pass: container-aware workspace roots, SkillHub UTF-8 repair and expert avatars, safer HITL resume/queue behavior, friendlier skill slash tokens, and catalog “show more” UX with harness bumps. Co-authored-by: Cursor <cursoragent@cursor.com>
62 lines
2.1 KiB
Python
62 lines
2.1 KiB
Python
"""Tests for UTF-8 coercion of third-party package text files."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import pytest
|
|
|
|
from octop.infra.utils.utf8_text import (
|
|
InvalidSkillManifestEncodingError,
|
|
coerce_utf8_text_bytes,
|
|
looks_like_text_path,
|
|
repair_known_utf8_corruption,
|
|
require_utf8_skill_manifest,
|
|
)
|
|
|
|
|
|
def test_looks_like_text_path() -> None:
|
|
assert looks_like_text_path("SKILL.md")
|
|
assert looks_like_text_path("references/guide.yaml")
|
|
assert not looks_like_text_path("assets/icon.png")
|
|
|
|
|
|
def test_coerce_repairs_skillhub_le_corruption() -> None:
|
|
# Upstream ecommerce-copy-humanizer-zh shipped ``≤`` as ``e2 6a 24``.
|
|
raw = "每句 ".encode() + b"\xe2j$" + "15 字".encode()
|
|
fixed = coerce_utf8_text_bytes(raw, path="notes.md")
|
|
assert fixed.decode("utf-8") == "每句 ≤15 字"
|
|
|
|
|
|
def test_coerce_leaves_valid_utf8_unchanged() -> None:
|
|
raw = "hello ≤ world".encode()
|
|
assert coerce_utf8_text_bytes(raw, path="SKILL.md") == raw
|
|
|
|
|
|
def test_coerce_skips_binary_paths() -> None:
|
|
raw = b"\xe2j$binary"
|
|
assert coerce_utf8_text_bytes(raw, path="icon.png") == raw
|
|
|
|
|
|
def test_coerce_replaces_unknown_invalid_sequences_for_aux_text() -> None:
|
|
raw = b"ok\xffstill"
|
|
fixed = coerce_utf8_text_bytes(raw, path="notes.txt")
|
|
assert fixed.decode("utf-8") == "ok\ufffdstill"
|
|
|
|
|
|
def test_repair_known_returns_none_for_unknown_corruption() -> None:
|
|
assert repair_known_utf8_corruption(b"ok\xffstill") is None
|
|
|
|
|
|
def test_require_utf8_skill_manifest_repairs_known_corruption() -> None:
|
|
raw = "每句 ".encode() + b"\xe2j$" + "15 字".encode()
|
|
assert require_utf8_skill_manifest(raw).decode("utf-8") == "每句 ≤15 字"
|
|
|
|
|
|
def test_require_utf8_skill_manifest_rejects_unknown_corruption() -> None:
|
|
with pytest.raises(InvalidSkillManifestEncodingError):
|
|
require_utf8_skill_manifest(b"---\nname: bad\n---\n\xff\xfe")
|
|
|
|
|
|
def test_coerce_skill_manifest_rejects_unknown_corruption() -> None:
|
|
with pytest.raises(InvalidSkillManifestEncodingError):
|
|
coerce_utf8_text_bytes(b"---\nname: bad\n---\n\xff\xfe", path="SKILL.md")
|