anatoly/autocad-mcp-pro
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

fix(task-7): crossings test the exact arc, so a line through a jump's apex counts

Browse Source 
flatten_bulges chops a semicircle into an even number of chords (36 for
r=50, 12 for r=5), so a flattened vertex lands exactly on the apex — the
axis of symmetry, the one place a P&ID line most plausibly meets a
crossing jump. segments_cross treats a vertex hit as a junction, so both
neighbouring chords dropped it and the arc contributed 0: the (45,0)->(55,0)
bulge -1 jump reported crossings 0 at x=50 and 1 at x=50.3 or x=49, the
plan's own (0,0)->(100,0) semicircle 0 at x=50, and the mirrored frame
(extrusion -Z) 0 at x=-50.

count_crossings now takes bulges= (per-chain DXF bulge lists, zero-padded)
and tests every bulged edge as its exact arc through the new
segment_arc_crossings — circle/segment intersection restricted to the
sweep — with the same junction rule as the straight case: a hit within
_EPS of the segment's ends, of the arc's own two ends, or a tangent, is
not a crossing. Nothing along the sweep is a vertex any more. The arc
parametrisation is factored into _arc_of_bulge and shared with
flatten_bulges, which keeps its promised signature and its ezdxf pin but
no longer feeds the crossing count. draw_line passes the vertices and
bulges through unflattened.

Pinned in tests: the apex case on the semicircle (x=50, y 20->60), the
r=5 jump at x=50 / 50.3 / 49 on +Z and -Z frames, the junction cases
(T-touch on the arc, through the arc's endpoint, chord side, tangent
along the apex, in-and-out at y=4.9 counts 2), and 4x400 seeded random
segment/arc pairs against ezdxf's ConstructionArc.intersect_line (a
20 000-case run in the worktree found 0 mismatches). The eight new
assertions fail on the flattened rule and pass on the exact one.

Headless finding, headless evidence; no new ActiveX call was written, so
the shared live AutoCAD was not touched.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
This commit is contained in:
Umut CAN
2026-09-22 11:06:04 +03:00
co-authored by Claude Opus 5
parent 9cb4bee587
commit efcc8bbba1
4 changed files with 201 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files
engineering/pid/drawlines.py
+3 -2
View File
@@ -13,7 +13,6 @@ from .lines import (
PID_LINE_LAYERS,
aim_points,
count_crossings,
flatten_bulges,
format_line_number,
label_placement,
marker_positions,
@@ -238,7 +237,9 @@ async def draw_line(
existing = await existing_pid_lines(backend)
crossings = count_crossings(
path, [flatten_bulges(ln["vertices"], ln["bulges"]) for ln in existing]
path,
[ln["vertices"] for ln in existing],
bulges=[ln["bulges"] for ln in existing],
)
used: set[tuple[str, str]] = set()
max_seq = 0
engineering/pid/lines.py
+98 -27
View File
@@ -238,23 +238,50 @@ def marker_positions(vertices, min_len: float = 15.0) -> list[tuple[float, float
return out
#: Max chord deviation (mm) when a bulged polyline edge is flattened for the
#: crossing test — a tenth of the thinnest ISO 128 lineweight.
#: Max chord deviation (mm) when a bulged polyline edge is flattened into
#: chords (``flatten_bulges``) — a tenth of the thinnest ISO 128 lineweight.
#: The crossing count does **not** flatten: it tests the exact arc.
FLATTEN_SAGITTA = 0.05
def _arc_of_bulge(a: Point, b: Point, bulge: float) -> tuple[float, float, float, float, float]:
"""``(cx, cy, radius, start, theta)`` of the arc a→b with the DXF bulge
``tan(sweep/4)``: ``theta`` is the signed sweep in radians (positive
counter-clockwise) and ``start`` the angle of ``a`` about the centre.
Same arithmetic as ``engineering/measure.py`` (sweep ``4·atan(bulge)``,
radius ``chord / 2·sin(|sweep|/2)``) and ``graph._point_segment_distance``
(centre on the chord's left normal, ``radius - sagitta`` carrying it across
the chord past a semicircle). The caller guarantees a non-zero bulge and a
non-degenerate chord.
"""
chord = _dist(a, b)
theta = 4.0 * math.atan(bulge) # signed sweep, radians
radius = chord / (2.0 * math.sin(abs(theta) / 2.0))
sag = abs(bulge) * chord / 2.0
nx, ny = -(b[1] - a[1]) / chord, (b[0] - a[0]) / chord # left of a→b
offset = (radius - sag) * (1.0 if bulge > 0 else -1.0)
cx = (a[0] + b[0]) / 2.0 + offset * nx
cy = (a[1] + b[1]) / 2.0 + offset * ny
start = math.atan2(a[1] - cy, a[0] - cx)
return cx, cy, radius, start, theta
def flatten_bulges(points, bulges, sagitta: float = FLATTEN_SAGITTA) -> list[Point]:
"""The polyline as straight segments, every bulged edge replaced by chords
that deviate from the arc by at most ``sagitta``.
Same arc arithmetic as ``engineering/measure.py`` (sweep ``4·atan(bulge)``,
radius ``chord / 2·sin(|sweep|/2)``) and ``graph._point_segment_distance``
(centre on the chord's left normal, ``radius - sagitta`` carrying it across
the chord past a semicircle). Checked against ``ezdxf.math.bulge_to_arc``
+ ``ConstructionArc.flattening`` in ``tests/test_pid_lines.py``: identical
vertex count and positions to 1e-12 for bulges 0.2 … 2.0 of either sign.
A straight edge (bulge 0) is passed through; ``bulges`` shorter than the
edge count is padded with zeros.
Arc arithmetic from ``_arc_of_bulge``. Checked against
``ezdxf.math.bulge_to_arc`` + ``ConstructionArc.flattening`` in
``tests/test_pid_lines.py``: identical vertex count and positions to 1e-12
for bulges 0.2 … 2.0 of either sign. A straight edge (bulge 0) is passed
through; ``bulges`` shorter than the edge count is padded with zeros.
Not used by the crossing count: a flattened chain has artificial vertices,
and ``segments_cross`` treats a hit on any vertex as a junction — a
semicircle flattens to an even chord count, so a line through its apex
(the axis of symmetry, exactly where a P&ID line meets a crossing jump)
would be dropped by both chords. ``count_crossings`` tests the exact arc.
"""
pts = [(float(x), float(y)) for x, y in points]
if len(pts) < 2:
@@ -263,17 +290,9 @@ def flatten_bulges(points, bulges, sagitta: float = FLATTEN_SAGITTA) -> list[Poi
for i, (a, b) in enumerate(zip(pts, pts[1:], strict=False)):
out.append(a)
bulge = float(bulges[i]) if i < len(bulges) and bulges[i] else 0.0
chord = _dist(a, b)
if abs(bulge) < 1e-12 or chord < 1e-12:
if abs(bulge) < 1e-12 or _dist(a, b) < 1e-12:
continue
theta = 4.0 * math.atan(bulge) # signed sweep, radians
radius = chord / (2.0 * math.sin(abs(theta) / 2.0))
sag = abs(bulge) * chord / 2.0
nx, ny = -(b[1] - a[1]) / chord, (b[0] - a[0]) / chord # left of a→b
offset = (radius - sag) * (1.0 if bulge > 0 else -1.0)
cx = (a[0] + b[0]) / 2.0 + offset * nx
cy = (a[1] + b[1]) / 2.0 + offset * ny
start = math.atan2(a[1] - cy, a[0] - cx)
cx, cy, radius, start, theta = _arc_of_bulge(a, b, bulge)
if sagitta >= radius:
segments = 1
else:
@@ -296,15 +315,67 @@ def segments_cross(p1, p2, q1, q2) -> bool:
return d1 * d2 < -_EPS and d3 * d4 < -_EPS
def count_crossings(vertices, others) -> int:
def segment_arc_crossings(p1, p2, a, b, bulge: float) -> int:
"""Proper crossings (0, 1 or 2) of the straight segment p1→p2 with the
exact arc a→b of DXF ``bulge``; a zero bulge is the chord test.
The same junction rule as ``segments_cross``: a hit within ``_EPS`` (mm)
of either end of the segment or of the arc is a T-touch or a shared
vertex, not a crossing, and a tangent (the line reaches no deeper than
``_EPS`` into the circle) touches without crossing. Only the arc's own
two ends are junction points — nothing along the sweep is a vertex.
"""
if abs(bulge) < 1e-12 or _dist(a, b) < 1e-12:
return 1 if segments_cross(p1, p2, a, b) else 0
cx, cy, radius, start, theta = _arc_of_bulge(a, b, bulge)
dx, dy = p2[0] - p1[0], p2[1] - p1[1]
length_sq = dx * dx + dy * dy
if length_sq < 1e-24:
return 0
length = math.sqrt(length_sq)
fx, fy = p1[0] - cx, p1[1] - cy
# |p1 + t·d − c|² = r² → A t² + B t + C = 0 with A = |d|²
coeff_b = 2.0 * (fx * dx + fy * dy)
coeff_c = fx * fx + fy * fy - radius * radius
disc = coeff_b * coeff_b - 4.0 * length_sq * coeff_c
if disc <= 0.0:
return 0
root = math.sqrt(disc)
if root / (2.0 * length) <= _EPS:
return 0 # tangent: the half-chord inside the circle is nothing
sweep = abs(theta)
sign = 1.0 if theta > 0 else -1.0
count = 0
for numerator in (-coeff_b - root, -coeff_b + root):
t = numerator / (2.0 * length_sq)
if t * length <= _EPS or (1.0 - t) * length <= _EPS:
continue # the segment ends on the arc: a T-touch
x, y = p1[0] + t * dx, p1[1] + t * dy
along = (sign * (math.atan2(y - cy, x - cx) - start)) % (2.0 * math.pi)
if along * radius <= _EPS or abs(sweep - along) * radius <= _EPS:
continue # the arc's own end: the polyline's vertex, a junction
if along > sweep:
continue # on the circle but outside the arc
count += 1
return count
def count_crossings(vertices, others, bulges=None) -> int:
"""Proper crossings of the straight chain ``vertices`` with every chain in
``others``. ``bulges[i]`` is the per-vertex DXF bulge list of ``others[i]``
(the arc leaving each vertex; short lists are zero-padded, ``None`` means
every chain is straight): a bulged edge is tested as its exact arc, never
as its chord or as flattened chords.
"""
pts = [(float(x), float(y)) for x, y in vertices]
count = 0
for a, b in zip(pts, pts[1:], strict=False):
for other in others:
opts = [(float(x), float(y)) for x, y in other]
for c, d in zip(opts, opts[1:], strict=False):
if segments_cross(a, b, c, d):
count += 1
for index, other in enumerate(others):
opts = [(float(x), float(y)) for x, y in other]
obulges = list(bulges[index]) if bulges is not None and index < len(bulges) else []
for a, b in zip(pts, pts[1:], strict=False):
for j, (c, d) in enumerate(zip(opts, opts[1:], strict=False)):
bulge = float(obulges[j]) if j < len(obulges) and obulges[j] else 0.0
count += segment_arc_crossings(a, b, c, d, bulge)
return count