feat: draft main-line centreline stations from plan sheets (Task 3, HITL)

extract_alignment_plan.py takes the midpoint of each station's two side
ticks (CA-ALGN-STSM) as the centreline point. Sheet 001 gives 4+700..5+500
with consecutive residual <= 0.02 m. Sheet 002 gives identical output, so
the duplicate question is left to HITL. Profile EL is not yet extracted.

Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A5Cun3RHQKna9qxE6ybZ4A

scripts/extract_alignment_plan.py

@@ -0,0 +1,81 @@
+#!/usr/bin/env python3
+"""Task 3 (HITL draft): main-line centreline station table from a plan sheet.
+
+Method (layers CA-ALGN-TEXT labels and CA-ALGN-STSM station ticks):
+  1. Station labels: 'N+mmm' is explicit; bare numbers take the km of the next
+     explicit label when they precede it (700 before 5+000 -> 4+700), otherwise
+     the km of the previous explicit label.
+  2. Each station has two labels (one per side of the road). Each label is
+     paired with its nearest tick; the centreline point is the midpoint of the
+     two tick points.
+  3. Checks reported (not forced): side offset per station, and consecutive
+     centreline distance vs station difference (residual).
+DRAFT for HITL review. No value is adjusted to fit."""
+import csv
+import math
+import re
+import sys
+from collections import defaultdict
+
+import ezdxf
+
+SRC, OUT = sys.argv[1], sys.argv[2]
+msp = ezdxf.readfile(SRC).modelspace()
+explicit_re = re.compile(r"^(\d+)\+(\d{3})$")
+
+labels = [(e.dxf.insert.x, e.dxf.insert.y, e.dxf.text.strip())
+          for e in msp.query("TEXT") if e.dxf.layer == "CA-ALGN-TEXT"]
+ticks = []
+for e in msp.query("LINE"):
+    if e.dxf.layer == "CA-ALGN-STSM":
+        a, b = e.dxf.start, e.dxf.end
+        ticks.append(((a.x + b.x) / 2, (a.y + b.y) / 2))
+
+labels = [l for l in labels if explicit_re.match(l[2]) or l[2].isdigit()]
+labels.sort(key=lambda t: -t[1])  # along the sheet, NNE -> SSW
+explicit_idx = [i for i, l in enumerate(labels) if explicit_re.match(l[2])]
+station_of = {}
+for i, (x, y, s) in enumerate(labels):
+    if explicit_re.match(s):
+        m = explicit_re.match(s)
+        station_of[i] = int(m.group(1)) * 1000 + int(m.group(2))
+        continue
+    v = int(s)
+    nxt = next((j for j in explicit_idx if j > i), None)
+    prv = next((j for j in reversed(explicit_idx) if j < i), None)
+    if prv is not None:
+        km = int(explicit_re.match(labels[prv][2]).group(1))
+    else:
+        km = int(explicit_re.match(labels[nxt][2]).group(1)) - 1
+    station_of[i] = km * 1000 + v
+
+groups = defaultdict(list)
+for i, (x, y, s) in enumerate(labels):
+    groups[station_of[i]].append((x, y))
+
+
+def nearest_tick(pt):
+    return min(ticks, key=lambda t: math.hypot(t[0] - pt[0], t[1] - pt[1]))
+
+centre = []
+for sta in sorted(groups):
+    pts = groups[sta]
+    if len(pts) != 2:
+        continue
+    t1, t2 = nearest_tick(pts[0]), nearest_tick(pts[1])
+    side_gap = math.dist(t1, t2)
+    mid = ((t1[0] + t2[0]) / 2, (t1[1] + t2[1]) / 2)
+    centre.append((sta, mid[0], mid[1], side_gap))
+
+with open(OUT, "w", newline="", encoding="utf-8") as fh:
+    w = csv.writer(fh)
+    w.writerow(["station_m", "centre_x", "centre_y", "side_gap_m"])
+    for sta, x, y, g in centre:
+        w.writerow([sta, round(x, 3), round(y, 3), round(g, 3)])
+
+print(f"stations={len(centre)} (labels={len(labels)}, ticks={len(ticks)})")
+prev = None
+for sta, x, y, g in centre:
+    step = "" if prev is None else f"  d={math.dist(prev[1:3], (x, y)):8.3f} dSTA={sta - prev[0]:5d} res={math.dist(prev[1:3], (x, y)) - (sta - prev[0]):+.3f}"
+    print(f"  STA {sta:>5}  ({x:12.3f}, {y:12.3f})  side_gap={g:6.2f}{step}")
+    prev = (sta, x, y)