File size: 6,025 Bytes
15d1f1d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 | #!/usr/bin/env python3
"""Release candidate: fragment-safe views and hidden-only Curve bounds.
This add-only final supersedes ``eval_geometric052_rc5j``. The host cannot
update an existing file, so the conservative visible-first Curve contract is
published under an explicit final entry point.
Two scoped repairs are stacked on RC5I:
* uniquely contained disconnected prediction fragments may be reassociated
with one transformed semantic reference box; outside or ambiguous fragments
remain explicit extras;
* Curve-CD/F1 retains its legacy visible-only bounds bit-for-bit whenever
those bounds exist. Only a hidden-only target/prediction falls back to
analytic bounds of the non-degenerate LINE/CIRCLE/ARC primitives that Curve
actually samples.
There is no path/hash identity shortcut and no change to global alignment,
sampling density, Curve thresholds, or downstream reward formulae.
"""
from __future__ import annotations
import math
import threading
from pathlib import Path
from typing import Any, Mapping, Sequence
import numpy as np
try:
from . import eval_geometric052_rc5i as RC5I
from . import view_entity_matching_v52_rc5j_fragment_candidate as MATCH
except ImportError: # direct evaluator import
import eval_geometric052_rc5i as RC5I
import view_entity_matching_v52_rc5j_fragment_candidate as MATCH
METRIC_VERSION = "v5.2-rc5j-final"
SCHEMA_VERSION = "edp-geometric-evaluator-v5.2-rc5j-final"
ENGINEERING_THRESHOLDS = dict(RC5I.ENGINEERING_THRESHOLDS)
PHYSICAL_SUPPORTED_SEMANTICS = RC5I.PHYSICAL_SUPPORTED_SEMANTICS
MAX_MATCHED_OFFSET_DISPERSION_FRACTION = (
RC5I.MAX_MATCHED_OFFSET_DISPERSION_FRACTION
)
RC3 = RC5I.RC3
BASE = RC5I.BASE
CURVE = BASE.CURVE
_PATCH_LOCK = threading.RLock()
def _hidden_fallback_entity_bounds(
entities: Mapping[str, Any],
) -> tuple[np.ndarray, np.ndarray] | None:
"""Return legacy visible bounds, else sampled all-channel analytic bounds."""
visible_bounds = BASE.EV5._safe_bbox(entities)
if visible_bounds is not None:
return visible_bounds
records = BASE.EV5._primitive_records(entities)
sampled: list[Mapping[str, Any]] = []
for record in records:
family = str(record.get("family") or "")
primitive = record.get("primitive")
if not isinstance(primitive, Sequence) or isinstance(primitive, (str, bytes)):
continue
try:
if family == "lines":
length = math.hypot(
float(primitive[2]) - float(primitive[0]),
float(primitive[3]) - float(primitive[1]),
)
elif family == "circles":
length = 2.0 * math.pi * abs(float(primitive[2]))
elif family == "arcs":
radius = abs(float(primitive[2]))
sweep_degrees = (
float(primitive[4]) - float(primitive[3])
) % 360.0
if sweep_degrees <= 1.0e-12:
sweep_degrees = 360.0
length = radius * math.radians(sweep_degrees)
else:
continue
except (IndexError, TypeError, ValueError):
continue
if math.isfinite(length) and length > 1.0e-12:
sampled.append(record)
if not sampled:
return None
lows = np.stack(
[np.asarray(record["low"], dtype=float) for record in sampled]
)
highs = np.stack(
[np.asarray(record["high"], dtype=float) for record in sampled]
)
if not np.isfinite(lows).all() or not np.isfinite(highs).all():
return None
return np.min(lows, axis=0), np.max(highs, axis=0)
def score_dxf_pair(
pred: str | Path,
gt: str | Path,
sidecar: Mapping[str, Any] | None = None,
category: str | None = None,
**kwargs: Any,
) -> dict[str, Any]:
with _PATCH_LOCK:
original_match = RC3.MATCH
original_curve_bounds = CURVE._entity_bounds
RC3.MATCH = MATCH
CURVE._entity_bounds = _hidden_fallback_entity_bounds
try:
result = RC5I.score_dxf_pair(
pred,
gt,
sidecar=sidecar,
category=category,
**kwargs,
)
finally:
CURVE._entity_bounds = original_curve_bounds
RC3.MATCH = original_match
result["schema_version"] = SCHEMA_VERSION
result["metric_version"] = METRIC_VERSION
result.setdefault("metric_definition", {}).update(
{
"fragment_reassociation_policy": {
"partition": (
"unique_complete_bbox_containment_after_coarse_shared_transform"
),
"outside_or_ambiguous_fragments": "retained_as_explicit_extras",
"activation": "strict_shape_layout_objective_improvement",
"path_or_hash_identity_used": False,
},
"curve_bounds_policy": {
"families": ["LINE", "CIRCLE", "ARC"],
"visibility": "legacy_visible_first_hidden_only_fallback",
"bounds": (
"legacy_visible_bounds_else_analytic_extrema_of_all_"
"nonzero_arc_length_primitives"
),
"global_alignment_changed": False,
"curve_sampling_or_thresholds_changed": False,
},
}
)
result.setdefault("diagnostics", {})["rc5j_repairs"] = {
"fragment_context_builder": MATCH.SCHEMA_VERSION,
"curve_bounds_hidden_fallback_only": True,
"path_or_hash_identity_used": False,
}
release = result.get("diagnostics", {}).get("release_status")
if isinstance(release, dict):
release["candidate"] = METRIC_VERSION
return result
__all__ = [
"ENGINEERING_THRESHOLDS",
"MAX_MATCHED_OFFSET_DISPERSION_FRACTION",
"METRIC_VERSION",
"PHYSICAL_SUPPORTED_SEMANTICS",
"SCHEMA_VERSION",
"_hidden_fallback_entity_bounds",
"score_dxf_pair",
]
|