audit full semantic extent and AOI visuals

This commit is contained in:
OpenAI Codex
2026-04-18 09:43:42 +08:00
parent 5cb0eead7c
commit 449d76a463
9 changed files with 5113 additions and 1 deletions

View File

@@ -0,0 +1,368 @@
#!/usr/bin/env python3
from __future__ import annotations
import argparse
import json
import os
from collections import Counter, defaultdict
from dataclasses import dataclass
from pathlib import Path
from typing import Any
from mapbox_vector_tile.Mapbox import vector_tile_pb2
DEFAULT_SOURCE_ROOT = Path(
"/home/wwwroot/newpec/exported_auto/"
"tile.mapple-on.jp__newpec-mvt-20260106__z___x___y_.pbf/tiles"
)
@dataclass
class LayerGeometryStats:
point_count: int = 0
min_x: int = 0
max_x: int = 0
min_y: int = 0
max_y: int = 0
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser(description="审计并按需修复 NavSea PBF 的几何 / extent / native 风险。")
parser.add_argument("--target-root", type=Path, required=True)
parser.add_argument("--label", required=True)
parser.add_argument("--output-dir", type=Path, required=True)
parser.add_argument("--source-root", type=Path, default=DEFAULT_SOURCE_ROOT)
parser.add_argument("--apply-fix", action="store_true", help="若检测到 extent 与源瓦片不一致,则按源瓦片 extent 原地修复。")
parser.add_argument("--sample-limit", type=int, default=30)
parser.add_argument("--zoom", action="append", type=int, help="只审指定 zoom可重复传入多次。")
return parser.parse_args()
def zigzag_decode(value: int) -> int:
return (value >> 1) ^ (-(value & 1))
def decode_feature_geometry(geometry: list[int]) -> LayerGeometryStats:
stats = LayerGeometryStats()
cursor_x = 0
cursor_y = 0
initialized = False
idx = 0
while idx < len(geometry):
command = geometry[idx]
idx += 1
command_id = command & 0x7
command_count = command >> 3
if command_id in (1, 2): # MoveTo, LineTo
for _ in range(command_count):
if idx + 1 >= len(geometry):
raise ValueError("geometry command truncated")
cursor_x += zigzag_decode(geometry[idx])
cursor_y += zigzag_decode(geometry[idx + 1])
idx += 2
if not initialized:
stats.min_x = stats.max_x = cursor_x
stats.min_y = stats.max_y = cursor_y
initialized = True
else:
stats.min_x = min(stats.min_x, cursor_x)
stats.max_x = max(stats.max_x, cursor_x)
stats.min_y = min(stats.min_y, cursor_y)
stats.max_y = max(stats.max_y, cursor_y)
stats.point_count += 1
elif command_id == 7: # ClosePath
continue
else:
raise ValueError(f"unsupported geometry command id={command_id}")
return stats
def merge_layer_stats(existing: LayerGeometryStats | None, incoming: LayerGeometryStats) -> LayerGeometryStats:
if existing is None or existing.point_count == 0:
return incoming
if incoming.point_count == 0:
return existing
return LayerGeometryStats(
point_count=existing.point_count + incoming.point_count,
min_x=min(existing.min_x, incoming.min_x),
max_x=max(existing.max_x, incoming.max_x),
min_y=min(existing.min_y, incoming.min_y),
max_y=max(existing.max_y, incoming.max_y),
)
def analyze_tile(tile_path: Path) -> dict[str, Any]:
tile = vector_tile_pb2.tile()
tile.ParseFromString(tile_path.read_bytes())
layer_summaries: list[dict[str, Any]] = []
extent_values: set[int] = set()
tile_outside_extent = False
tile_max_overflow = 0
for layer in tile.layers:
extent = int(layer.extent or 4096)
extent_values.add(extent)
aggregate: LayerGeometryStats | None = None
for feature in layer.features:
geom_stats = decode_feature_geometry(list(feature.geometry))
aggregate = merge_layer_stats(aggregate, geom_stats)
if aggregate is None:
aggregate = LayerGeometryStats()
outside_extent = False
overflow = 0
if aggregate.point_count:
overflow = max(
0,
-aggregate.min_x,
-aggregate.min_y,
aggregate.max_x - extent,
aggregate.max_y - extent,
)
outside_extent = overflow > 0
tile_outside_extent = tile_outside_extent or outside_extent
tile_max_overflow = max(tile_max_overflow, overflow)
layer_summaries.append(
{
"name": layer.name,
"extent": extent,
"point_count": aggregate.point_count,
"min_x": aggregate.min_x,
"max_x": aggregate.max_x,
"min_y": aggregate.min_y,
"max_y": aggregate.max_y,
"outside_extent": outside_extent,
"overflow": overflow,
}
)
return {
"extent_values": sorted(extent_values),
"outside_extent": tile_outside_extent,
"max_overflow": tile_max_overflow,
"layers": layer_summaries,
}
def desired_extent_from_source(source_path: Path) -> int | None:
if not source_path.exists():
return None
tile = vector_tile_pb2.tile()
tile.ParseFromString(source_path.read_bytes())
extents = sorted({int(layer.extent or 4096) for layer in tile.layers})
if not extents:
return None
if len(extents) == 1:
return extents[0]
return max(extents)
def repair_tile_extents(tile_path: Path, desired_extent: int) -> bool:
tile = vector_tile_pb2.tile()
raw = tile_path.read_bytes()
tile.ParseFromString(raw)
changed = False
for layer in tile.layers:
current = int(layer.extent or 4096)
if current != desired_extent:
layer.extent = desired_extent
changed = True
if changed:
tile_path.write_bytes(tile.SerializeToString())
return changed
def hardlink_copy_tree(src_root: Path, dst_root: Path) -> None:
if dst_root.exists():
raise SystemExit(f"backup root already exists: {dst_root}")
dst_root.mkdir(parents=True, exist_ok=True)
os.system(f"cp -al {src_root}/. {dst_root}/")
def write_report(
output_dir: Path,
label: str,
target_root: Path,
source_root: Path,
summary: dict[str, Any],
) -> tuple[Path, Path]:
output_dir.mkdir(parents=True, exist_ok=True)
report_json = output_dir / "geometry_native_audit.json"
report_md = output_dir / "geometry_native_audit.md"
report_json.write_text(json.dumps(summary, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
lines = [
f"# {label} 几何 / Extent / Native 风险审计",
"",
"## 范围",
"",
f"- target root: `{target_root}`",
f"- source root: `{source_root}`",
f"- tiles: `{summary['tile_count']}`",
f"- repaired tiles: `{summary['repaired_tiles']}`",
"",
"## 总体结论",
"",
f"- 坐标超当前 extent 的瓦片数:`{summary['outside_extent_tiles']}`",
f"- 与源瓦片 extent 不一致的瓦片数:`{summary['source_extent_mismatch_tiles']}`",
f"- 可直接按源 extent 修复的瓦片数:`{summary['fixable_from_source_tiles']}`",
"",
"## Zoom 汇总",
"",
]
for zoom in sorted(summary["zoom_summary"], key=int):
item = summary["zoom_summary"][zoom]
lines.extend(
[
f"### z{zoom}",
"",
f"- tiles: `{item['tile_count']}`",
f"- outside extent: `{item['outside_extent_tiles']}`",
f"- source mismatch: `{item['source_extent_mismatch_tiles']}`",
f"- extent counts: `{item['extent_counts']}`",
"",
]
)
lines.extend(["## 代表问题样本", ""])
for item in summary["sample_tiles"]:
lines.extend(
[
f"### `{item['tile']}`",
"",
f"- extent values: `{item['extent_values']}`",
f"- source extent: `{item['source_extent']}`",
f"- outside extent: `{item['outside_extent']}`",
f"- max overflow: `{item['max_overflow']}`",
f"- fixable from source: `{item['fixable_from_source']}`",
"",
]
)
report_md.write_text("\n".join(lines) + "\n", encoding="utf-8")
return report_md, report_json
def main() -> None:
args = parse_args()
target_root = args.target_root
source_root = args.source_root
output_dir = args.output_dir
if not target_root.exists():
raise SystemExit(f"target root not found: {target_root}")
if not source_root.exists():
raise SystemExit(f"source root not found: {source_root}")
zoom_summary: dict[str, dict[str, Any]] = defaultdict(
lambda: {
"tile_count": 0,
"outside_extent_tiles": 0,
"source_extent_mismatch_tiles": 0,
"extent_counts": Counter(),
}
)
sample_tiles: list[dict[str, Any]] = []
repaired_tiles = 0
outside_extent_tiles = 0
source_extent_mismatch_tiles = 0
fixable_from_source_tiles = 0
tile_paths = sorted(target_root.glob("*/*/*.pbf"))
if args.zoom:
zoom_filter = {str(item) for item in args.zoom}
tile_paths = [path for path in tile_paths if path.parent.parent.name in zoom_filter]
for tile_path in tile_paths:
rel = tile_path.relative_to(target_root)
zoom = rel.parts[0]
analysis = analyze_tile(tile_path)
source_extent = desired_extent_from_source(source_root / rel)
current_extent = max(analysis["extent_values"]) if analysis["extent_values"] else None
mismatch = source_extent is not None and current_extent is not None and source_extent != current_extent
fixable = bool(
mismatch
and source_extent is not None
and analysis["outside_extent"]
and analysis["max_overflow"] <= max(0, source_extent - (current_extent or 0))
)
zoom_item = zoom_summary[zoom]
zoom_item["tile_count"] += 1
for extent in analysis["extent_values"]:
zoom_item["extent_counts"][extent] += 1
if analysis["outside_extent"]:
outside_extent_tiles += 1
zoom_item["outside_extent_tiles"] += 1
if mismatch:
source_extent_mismatch_tiles += 1
zoom_item["source_extent_mismatch_tiles"] += 1
if fixable:
fixable_from_source_tiles += 1
if args.apply_fix and fixable and source_extent is not None:
if repair_tile_extents(tile_path, source_extent):
repaired_tiles += 1
analysis = analyze_tile(tile_path)
if len(sample_tiles) < args.sample_limit and (analysis["outside_extent"] or mismatch):
sample_tiles.append(
{
"tile": str(rel),
"extent_values": analysis["extent_values"],
"source_extent": source_extent,
"outside_extent": analysis["outside_extent"],
"max_overflow": analysis["max_overflow"],
"fixable_from_source": fixable,
}
)
normalized_zoom_summary = {
zoom: {
"tile_count": item["tile_count"],
"outside_extent_tiles": item["outside_extent_tiles"],
"source_extent_mismatch_tiles": item["source_extent_mismatch_tiles"],
"extent_counts": dict(sorted(item["extent_counts"].items())),
}
for zoom, item in sorted(zoom_summary.items(), key=lambda kv: int(kv[0]))
}
summary = {
"label": args.label,
"target_root": str(target_root),
"source_root": str(source_root),
"tile_count": len(tile_paths),
"outside_extent_tiles": outside_extent_tiles,
"source_extent_mismatch_tiles": source_extent_mismatch_tiles,
"fixable_from_source_tiles": fixable_from_source_tiles,
"repaired_tiles": repaired_tiles,
"zoom_summary": normalized_zoom_summary,
"sample_tiles": sample_tiles,
}
report_md, report_json = write_report(output_dir, args.label, target_root, source_root, summary)
print(
json.dumps(
{
"report_md": str(report_md),
"report_json": str(report_json),
"outside_extent_tiles": outside_extent_tiles,
"source_extent_mismatch_tiles": source_extent_mismatch_tiles,
"fixable_from_source_tiles": fixable_from_source_tiles,
"repaired_tiles": repaired_tiles,
},
ensure_ascii=False,
indent=2,
)
)
if __name__ == "__main__":
main()