gjm
6143bb9ab2
支持直接审查 PDF 与位图图纸，全链路离线（不连 CAD、不连 8080 端口）。 渲染在内存内完成，复用既有墨迹热力图 + 自适应瓦片规划。 - 新增 dwg_review/sources.py - detect_source / plan_file / collect 统一入口 - PdfReader：pypdfium2 惰性导入，页序即图框序，处理 /Rotate 与 crop 边距 - ImageReader：世界单位 = 源像素，snap_box 整数对齐保证窗口与像素严格一致 - ink_grid：纯 PIL 灰度阈值 → 占用网格（不引入 numpy） - ReviewContext.create(connect_cad=False)：离线时 client/tools 均为 None - 全链路 ctx.tools is None 守卫：review 只读子循环 / inspect_region / annotator / checks 的窗口类检查 / 绘图区最大化 - checklist.build_system_prompt(allow_tools=False, offline=True)：离线提示词 - __main__：位置参数 source，按扩展名自动判定源类型并决定是否连 CAD - config 新增 8 个常量（PDF_MAX_PAGES / 文本最小字数 / 各源缩放上限 / 墨迹阈值等） - tiling.content_predicate 与 heatmap.cell_box 提升为公开函数 验证： - pytest 457 passed - verify_pdf_source.py：PDF 23/23、图片 9/9（不需 CAD），含实测瓦片字高像素与 旋转映射自检（文本区墨迹 0.311 vs 忽略 /Rotate 反向对照 0.145） - 真实 LLM 端到端：20 页 HVAC 图纸前 3 页跑通，结论具体可核验且未编造 handle 已知局限：页边线在墨迹网格里会把整页连成单簇，瓦片常退化为整页象限（CAD 侧 靠 drop_oversized 丢图框块，离线侧暂无等价手段），留待后续打磨。
463 lines
17 KiB
Python
463 lines
17 KiB
Python
"""`dwg_review.phases.evidence` 单元测试。"""
|
|
|
|
import base64
|
|
import io
|
|
|
|
import pytest
|
|
|
|
from dwg_review.phases.evidence import (
|
|
_apply_keep,
|
|
_content_predicate,
|
|
_is_truncated,
|
|
_plan_tiles,
|
|
_text_boxes,
|
|
bbox_to_window,
|
|
collect_frame,
|
|
maximized_drawing_area,
|
|
)
|
|
from tests.fakes import (
|
|
STANDARD_TOOLS,
|
|
FakeMcpClient,
|
|
err,
|
|
frame,
|
|
make_ctx,
|
|
ok_image,
|
|
ok_text,
|
|
tool_schema,
|
|
)
|
|
|
|
|
|
def test_bbox_to_window_mapping():
|
|
assert bbox_to_window([1, 2, 3, 4]) == {
|
|
"min_x": 1.0,
|
|
"min_y": 2.0,
|
|
"max_x": 3.0,
|
|
"max_y": 4.0,
|
|
}
|
|
|
|
|
|
def test_bbox_to_window_non_numeric_known_gap():
|
|
"""已知瑕疵：4 元素但含非数值的 bbox 会抛 ValueError，上游 collect_frame 只判长度。"""
|
|
with pytest.raises(ValueError):
|
|
bbox_to_window([0, "a", 10, 10])
|
|
|
|
|
|
def test_collect_frame_returns_none_for_missing_bbox():
|
|
ctx = make_ctx(client=FakeMcpClient())
|
|
assert collect_frame(ctx, 1, {}) is None
|
|
assert collect_frame(ctx, 1, {"bbox": None}) is None
|
|
|
|
|
|
def test_collect_frame_returns_none_for_bad_bbox_length():
|
|
ctx = make_ctx(client=FakeMcpClient())
|
|
assert collect_frame(ctx, 1, {"bbox": [0, 0, 10]}) is None
|
|
|
|
|
|
def build_evidence_client(image=ok_image("QUJD")):
|
|
return FakeMcpClient(
|
|
responses={
|
|
"query_entities": ok_text({"total": 5, "items": [{"handle": "A"}]}),
|
|
"get_text_content": ok_text({"total": 1, "items": [{"text": "你好"}]}),
|
|
"get_dimensions": ok_text({"total": 2, "items": []}),
|
|
"get_block_references": ok_text({"total": 0, "items": []}),
|
|
"get_polylines_info": ok_text({"total": 3, "items": []}),
|
|
"get_viewport_screenshot": image,
|
|
}
|
|
)
|
|
|
|
|
|
def test_collect_frame_assembles_evidence():
|
|
ctx = make_ctx(client=build_evidence_client())
|
|
ev = collect_frame(ctx, 2, frame([0, 0, 100, 50]))
|
|
assert ev["index"] == 2
|
|
assert ev["window"] == {"min_x": 0.0, "min_y": 0.0, "max_x": 100.0, "max_y": 50.0}
|
|
assert ev["entities"]["total"] == 5
|
|
assert ev["texts"]["total"] == 1
|
|
assert ev["dimensions"]["total"] == 2
|
|
assert ev["blocks"]["total"] == 0
|
|
assert ev["polylines"]["total"] == 3
|
|
assert ev["source_type"] == "cad"
|
|
# 截图登记「实际可见世界范围」：拿不到绘图区像素时退回「窗口外扩 margin」的近似值
|
|
assert ev["images"] == [
|
|
{
|
|
"b64": "QUJD",
|
|
"mime": "image/png",
|
|
"label": "overall",
|
|
"window": (-2.0, -1.0, 102.0, 51.0),
|
|
}
|
|
]
|
|
|
|
|
|
def test_collect_frame_calls_expected_tools():
|
|
ctx = make_ctx(client=build_evidence_client(), frame_limit=99)
|
|
collect_frame(ctx, 1, frame([1, 2, 3, 4]))
|
|
client = ctx.client
|
|
|
|
win = {"min_x": 1.0, "min_y": 2.0, "max_x": 3.0, "max_y": 4.0}
|
|
assert client.arguments_of("zoom_to_window") == dict(win, margin=0.02)
|
|
for name in (
|
|
"query_entities",
|
|
"get_text_content",
|
|
"get_dimensions",
|
|
"get_block_references",
|
|
"get_polylines_info",
|
|
):
|
|
assert client.arguments_of(name) == {"window": win, "limit": 99}
|
|
assert client.arguments_of("get_viewport_screenshot") == {}
|
|
|
|
|
|
def test_collect_frame_without_image():
|
|
ctx = make_ctx(client=build_evidence_client(image=ok_text({})))
|
|
ev = collect_frame(ctx, 1, frame([0, 0, 10, 10]))
|
|
assert ev["images"] == []
|
|
|
|
|
|
def test_collect_frame_tolerates_empty_structured_data():
|
|
ctx = make_ctx(client=FakeMcpClient()) # 全部默认 ok_text({})
|
|
ev = collect_frame(ctx, 1, frame([0, 0, 10, 10]))
|
|
assert ev["entities"] == {}
|
|
assert ev["images"] == []
|
|
|
|
|
|
def test_sub_windows_2x2():
|
|
from dwg_review.phases.evidence import sub_windows
|
|
|
|
win = {"min_x": 0.0, "min_y": 0.0, "max_x": 10.0, "max_y": 10.0}
|
|
subs = sub_windows(win, 4)
|
|
assert len(subs) == 4
|
|
# 行优先：左上 → 右上 → 左下 → 右下
|
|
assert subs[0] == {"min_x": 0.0, "min_y": 5.0, "max_x": 5.0, "max_y": 10.0}
|
|
assert subs[1] == {"min_x": 5.0, "min_y": 5.0, "max_x": 10.0, "max_y": 10.0}
|
|
assert subs[2] == {"min_x": 0.0, "min_y": 0.0, "max_x": 5.0, "max_y": 5.0}
|
|
assert subs[3] == {"min_x": 5.0, "min_y": 0.0, "max_x": 10.0, "max_y": 5.0}
|
|
|
|
|
|
def test_sub_windows_disabled():
|
|
from dwg_review.phases.evidence import sub_windows
|
|
|
|
win = {"min_x": 0.0, "min_y": 0.0, "max_x": 10.0, "max_y": 10.0}
|
|
assert sub_windows(win, 0) == []
|
|
assert sub_windows(win, 1) == []
|
|
|
|
|
|
def test_collect_frame_captures_multiple_images(monkeypatch):
|
|
from dwg_review import config
|
|
from dwg_review.phases.evidence import collect_frame
|
|
|
|
monkeypatch.setattr(config, "EVIDENCE_DETAIL_TILES", 4)
|
|
ctx = make_ctx(client=build_evidence_client())
|
|
ev = collect_frame(ctx, 1, frame([0, 0, 100, 50]))
|
|
assert [img["label"] for img in ev["images"]] == [
|
|
"overall",
|
|
"tile-1",
|
|
"tile-2",
|
|
"tile-3",
|
|
"tile-4",
|
|
]
|
|
|
|
|
|
def test_collect_dispatches_cad_source():
|
|
from dwg_review.phases.evidence import collect
|
|
|
|
ctx = make_ctx(client=build_evidence_client())
|
|
ev = collect(ctx, 1, frame([0, 0, 10, 10]))
|
|
assert ev is not None
|
|
assert ev["source_type"] == "cad"
|
|
|
|
|
|
def test_collect_returns_none_for_unsupported_source():
|
|
from dwg_review.phases.evidence import collect
|
|
|
|
ctx = make_ctx(client=FakeMcpClient())
|
|
# 未知来源（非 cad/pdf/image）→ 无法取证
|
|
assert collect(ctx, 1, {"source_type": "svg", "bbox": [0, 0, 1, 1]}) is None
|
|
|
|
|
|
def test_collect_dispatches_pdf_source_to_sources(monkeypatch):
|
|
"""pdf / image 来源交给 `sources.collect`（此处打桩，避免真去解析文件）。"""
|
|
from dwg_review.phases import evidence as ev_mod
|
|
|
|
calls = []
|
|
|
|
def _fake(ctx, index, unit):
|
|
calls.append((index, unit.get("source_type")))
|
|
return {"source_type": unit.get("source_type"), "index": index}
|
|
|
|
monkeypatch.setattr(ev_mod.sources, "collect", _fake)
|
|
ctx = make_ctx(client=FakeMcpClient())
|
|
out = ev_mod.collect(ctx, 3, {"source_type": "pdf", "bbox": [0, 0, 1, 1]})
|
|
assert out == {"source_type": "pdf", "index": 3}
|
|
assert calls == [(3, "pdf")]
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# 自适应瓦片规划
|
|
# ---------------------------------------------------------------------------
|
|
@pytest.fixture
|
|
def tiling_env(monkeypatch):
|
|
"""把瓦片相关配置钉死，使用例不受外部环境变量影响。"""
|
|
from dwg_review import config
|
|
|
|
for name, value in {
|
|
"TILING_ENABLED": True,
|
|
"TEXT_HEIGHT_MM": 2.0,
|
|
"TEXT_READABLE_PX": 12,
|
|
"LINE_READABLE_PX": 6,
|
|
"TILE_MAX_COUNT": 36,
|
|
"TILE_OVERLAP": 0.0,
|
|
"HEATMAP_CELL_DIV": 32,
|
|
"HEATMAP_CLUSTER_DILATE": 0,
|
|
"HEATMAP_MAX_AREA_FRAC": 0.5,
|
|
"IMAGE_MAX_EDGE": 1280,
|
|
"EVIDENCE_MAX_IMAGES": 5,
|
|
"EVIDENCE_DETAIL_TILES": 4,
|
|
}.items():
|
|
monkeypatch.setattr(config, name, value)
|
|
return config
|
|
|
|
|
|
def _png_b64(w=2000, h=1414):
|
|
from PIL import Image
|
|
|
|
buf = io.BytesIO()
|
|
Image.new("RGB", (w, h), "black").save(buf, format="PNG")
|
|
return base64.b64encode(buf.getvalue()).decode("ascii")
|
|
|
|
|
|
def _tiling_client(b64, entity_type="LINE"):
|
|
return FakeMcpClient(
|
|
responses={
|
|
"query_entities": ok_text(
|
|
{"total": 1, "items": [{"handle": "A", "type": entity_type,
|
|
"bbox": [0, 0, 60000, 20000]}]}
|
|
),
|
|
"get_text_content": ok_text({"total": 0, "items": []}),
|
|
"get_dimensions": ok_text({"total": 0, "items": []}),
|
|
"get_block_references": ok_text({"total": 0, "items": []}),
|
|
"get_polylines_info": ok_text({"total": 0, "items": []}),
|
|
"get_viewport_screenshot": ok_image(b64),
|
|
}
|
|
)
|
|
|
|
|
|
def test_is_truncated():
|
|
assert _is_truncated({"total": 5, "items": [1]}) is True
|
|
assert _is_truncated({"total": 1, "items": [1]}) is False
|
|
assert _is_truncated({"items": [1, 2]}) is False
|
|
assert _is_truncated(None, {}, {"total": 3, "items": []}) is True
|
|
assert _is_truncated() is False
|
|
|
|
|
|
def test_text_boxes_includes_std_and_custom():
|
|
entities = {"items": [
|
|
{"type": "LINE", "bbox": [0, 0, 1, 1]},
|
|
{"type": "AECDLABEL", "bbox": [2, 2, 3, 3]},
|
|
{"type": "TEXT", "bbox": [4, 4, 5, 5]},
|
|
]}
|
|
texts = {"items": [{"bbox": [6, 6, 7, 7]}]}
|
|
dimensions = {"items": [{"bbox": [8, 8, 9, 9]}]}
|
|
boxes = _text_boxes(entities, texts, dimensions)
|
|
for expected in ((2, 2, 3, 3), (4, 4, 5, 5), (6, 6, 7, 7), (8, 8, 9, 9)):
|
|
assert tuple(float(v) for v in expected) in boxes
|
|
assert (0.0, 0.0, 1.0, 1.0) not in boxes # 普通线不算文字
|
|
|
|
|
|
def test_plan_tiles_degrades(tiling_env):
|
|
win = {"min_x": 0.0, "min_y": 0.0, "max_x": 100.0, "max_y": 50.0}
|
|
f = frame([0, 0, 100, 50], sheet_size="A0", sheet_scale=100.0)
|
|
boxes = [(0.0, 0.0, 10.0, 10.0)]
|
|
assert _plan_tiles(win, f, boxes, [], False, (100, 50)) != [] # 正常路径非空
|
|
assert _plan_tiles(win, None, boxes, [], False, (100, 50)) == [] # 缺图框
|
|
assert _plan_tiles(win, f, [], [], False, (100, 50)) == [] # 无实体
|
|
assert _plan_tiles(win, f, boxes, [], False, None) == [] # 无像素
|
|
assert _plan_tiles(win, f, boxes, [], True, (100, 50)) == [] # 数据截断
|
|
# 比例未知（sheet_scale=0 且 sheet_size 不在纸张表）→ 无法规划
|
|
unknown = frame([0, 0, 100, 50], sheet_size="ZZ", sheet_scale=0.0)
|
|
assert _plan_tiles(win, unknown, boxes, [], False, (100, 50)) == []
|
|
|
|
|
|
def test_plan_tiles_disabled_by_config(tiling_env, monkeypatch):
|
|
tiling_env.TILING_ENABLED = False
|
|
win = {"min_x": 0.0, "min_y": 0.0, "max_x": 100.0, "max_y": 50.0}
|
|
f = frame([0, 0, 100, 50], sheet_size="A0", sheet_scale=100.0)
|
|
assert _plan_tiles(win, f, [(0.0, 0.0, 10.0, 10.0)], [], False, (100, 50)) == []
|
|
|
|
|
|
def test_capture_images_tiles_line_cluster(tiling_env):
|
|
"""线条簇：单张可覆盖 → 2 张瓦片（行优先）。"""
|
|
pytest.importorskip("PIL")
|
|
ctx = make_ctx(client=_tiling_client(_png_b64()))
|
|
ev = collect_frame(
|
|
ctx, 1, frame([0, 0, 120000, 40000], sheet_size="A0", sheet_scale=100.0)
|
|
)
|
|
assert [i["label"] for i in ev["images"]] == ["overall", "tile-1", "tile-2"]
|
|
|
|
|
|
def test_capture_images_tiles_text_cluster_is_finer(tiling_env):
|
|
"""文字簇用更细尺度 → 瓦片更多，并受 EVIDENCE_MAX_IMAGES 截断。"""
|
|
pytest.importorskip("PIL")
|
|
ctx = make_ctx(client=_tiling_client(_png_b64(), entity_type="AECDLABEL"))
|
|
ev = collect_frame(
|
|
ctx, 1, frame([0, 0, 120000, 40000], sheet_size="A0", sheet_scale=100.0)
|
|
)
|
|
labels = [i["label"] for i in ev["images"]]
|
|
assert labels == ["overall", "tile-1", "tile-2", "tile-3", "tile-4"] # 上限 5 张
|
|
|
|
|
|
def test_capture_images_falls_back_without_scale(tiling_env):
|
|
"""缺 sheet_size/sheet_scale → 退回固定 2x2 象限，并按内容过滤空白象限。"""
|
|
pytest.importorskip("PIL")
|
|
client = _tiling_client(_png_b64())
|
|
# 实体完全落在**左下**象限内（不贴边界）
|
|
client.responses["query_entities"] = ok_text(
|
|
{"total": 1, "items": [{"handle": "A", "type": "LINE",
|
|
"bbox": [0, 0, 50000, 15000]}]}
|
|
)
|
|
ctx = make_ctx(client=client)
|
|
ev = collect_frame(ctx, 1, frame([0, 0, 120000, 40000], sheet_size="ZZ"))
|
|
assert [i["label"] for i in ev["images"]] == ["overall", "tile-1"]
|
|
|
|
|
|
def test_capture_images_fallback_keeps_all_when_no_boxes(tiling_env):
|
|
"""拿不到实体 bbox（无法判断内容）→ 回退路径不过滤，四象限全拍。"""
|
|
pytest.importorskip("PIL")
|
|
client = _tiling_client(_png_b64())
|
|
client.responses["query_entities"] = ok_text({"total": 0, "items": []})
|
|
ctx = make_ctx(client=client)
|
|
ev = collect_frame(ctx, 1, frame([0, 0, 120000, 40000], sheet_size="ZZ"))
|
|
assert [i["label"] for i in ev["images"]] == [
|
|
"overall", "tile-1", "tile-2", "tile-3", "tile-4",
|
|
]
|
|
|
|
|
|
def test_capture_images_falls_back_when_truncated(tiling_env):
|
|
"""实体被截断 → 禁用内容过滤，退回固定策略。"""
|
|
pytest.importorskip("PIL")
|
|
client = _tiling_client(_png_b64())
|
|
client.responses["query_entities"] = ok_text(
|
|
{"total": 500, "items": [{"handle": "A", "type": "LINE",
|
|
"bbox": [0, 0, 60000, 20000]}]}
|
|
)
|
|
ctx = make_ctx(client=client)
|
|
ev = collect_frame(
|
|
ctx, 1, frame([0, 0, 120000, 40000], sheet_size="A0", sheet_scale=100.0)
|
|
)
|
|
assert [i["label"] for i in ev["images"]] == [
|
|
"overall", "tile-1", "tile-2", "tile-3", "tile-4",
|
|
]
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# 内容过滤 / 绘图区最大化
|
|
# ---------------------------------------------------------------------------
|
|
def test_apply_keep_filters_and_falls_back():
|
|
regions = [
|
|
{"min_x": 0.0, "min_y": 0.0, "max_x": 1.0, "max_y": 1.0},
|
|
{"min_x": 1.0, "min_y": 1.0, "max_x": 2.0, "max_y": 2.0},
|
|
]
|
|
keep_first = lambda w: w["min_x"] == 0.0 # noqa: E731
|
|
assert _apply_keep(regions, keep_first) == regions[:1]
|
|
assert _apply_keep(regions, None) == regions # 无谓词 → 原样
|
|
assert _apply_keep(regions, lambda w: False) == regions # 全被滤掉 → 退回原集合
|
|
assert _apply_keep([], keep_first) == []
|
|
|
|
|
|
def test_content_predicate_none_without_boxes():
|
|
assert _content_predicate([]) is None
|
|
assert _content_predicate(None) is None
|
|
|
|
|
|
def test_content_predicate_detects_intersection():
|
|
keep = _content_predicate([(0.0, 0.0, 5.0, 5.0)])
|
|
assert keep({"min_x": 4.0, "min_y": 4.0, "max_x": 9.0, "max_y": 9.0}) is True
|
|
assert keep({"min_x": 6.0, "min_y": 6.0, "max_x": 9.0, "max_y": 9.0}) is False
|
|
|
|
|
|
@pytest.fixture
|
|
def drawing_area_env(monkeypatch):
|
|
"""打开绘图区最大化，并把等待时间归零（用例不真的 sleep）。"""
|
|
from dwg_review import config
|
|
|
|
monkeypatch.setattr(config, "DRAWING_AREA_MAXIMIZE", True)
|
|
monkeypatch.setattr(config, "DRAWING_AREA_ENTER_WAIT", 0.0)
|
|
monkeypatch.setattr(config, "DRAWING_AREA_RESTORE_WAIT", 0.0)
|
|
return config
|
|
|
|
|
|
def _area_client(responses=None):
|
|
"""带 set_drawing_area 工具的假客户端。"""
|
|
return FakeMcpClient(
|
|
tools=STANDARD_TOOLS + [tool_schema("set_drawing_area", "绘图区最大化")],
|
|
responses=responses,
|
|
)
|
|
|
|
|
|
def test_maximized_drawing_area_calls_max_then_restore(drawing_area_env):
|
|
client = _area_client()
|
|
ctx = make_ctx(client=client)
|
|
with maximized_drawing_area(ctx):
|
|
assert client.called_names() == ["set_drawing_area"] # 进入时只 max
|
|
assert [c["arguments"] for c in client.called("set_drawing_area")] == [
|
|
{"mode": "max"}, {"mode": "restore"},
|
|
]
|
|
|
|
|
|
def test_maximized_drawing_area_degrades_without_tool(drawing_area_env):
|
|
"""旧版插件没有该工具 → 静默降级，一次调用都不发。"""
|
|
client = FakeMcpClient()
|
|
ctx = make_ctx(client=client)
|
|
with maximized_drawing_area(ctx):
|
|
pass
|
|
assert client.calls == []
|
|
|
|
|
|
def test_maximized_drawing_area_disabled_by_config(drawing_area_env):
|
|
drawing_area_env.DRAWING_AREA_MAXIMIZE = False
|
|
client = _area_client()
|
|
ctx = make_ctx(client=client)
|
|
with maximized_drawing_area(ctx):
|
|
pass
|
|
assert client.calls == []
|
|
|
|
|
|
def test_maximized_drawing_area_offline_passthrough(drawing_area_env):
|
|
"""离线源（tools=None）即使开启最大化也直接放行，不碰 CAD。"""
|
|
ctx = make_ctx(connect_cad=False)
|
|
assert ctx.tools is None
|
|
with maximized_drawing_area(ctx):
|
|
pass # 不应抛 AttributeError
|
|
|
|
|
|
def test_maximized_drawing_area_restores_even_if_body_raises(drawing_area_env):
|
|
client = _area_client()
|
|
ctx = make_ctx(client=client)
|
|
with pytest.raises(RuntimeError):
|
|
with maximized_drawing_area(ctx):
|
|
raise RuntimeError("boom")
|
|
assert client.arguments_of("set_drawing_area") == {"mode": "restore"}
|
|
|
|
|
|
def test_maximized_drawing_area_degrades_when_max_fails(drawing_area_env):
|
|
"""max 调用失败 → 视为没进入最大化，不再发 restore，也不抛异常。"""
|
|
client = _area_client(responses={"set_drawing_area": err("no window")})
|
|
ctx = make_ctx(client=client)
|
|
with maximized_drawing_area(ctx):
|
|
pass
|
|
assert [c["arguments"] for c in client.called("set_drawing_area")] == [{"mode": "max"}]
|
|
|
|
|
|
def test_maximized_drawing_area_restore_failure_does_not_mask_body(drawing_area_env):
|
|
"""restore 失败不应掩盖业务异常。"""
|
|
client = _area_client()
|
|
ctx = make_ctx(client=client)
|
|
|
|
def boom(args):
|
|
if args.get("mode") == "restore":
|
|
raise RuntimeError("restore failed")
|
|
return ok_text({})
|
|
|
|
client.responses["set_drawing_area"] = boom
|
|
with pytest.raises(ValueError):
|
|
with maximized_drawing_area(ctx):
|
|
raise ValueError("business error")
|