Fix S-101 multipoint sounding glyphs collapsing onto one anchor

[philliphoff]

Summary

Multi-digit depth soundings in S-101 ENCs were rendering as overlapping black blobs instead of clean numeric labels. The root cause is in the shared drawing-instruction parser: S-101 multipoint sounding features emit one AugmentedPoint:GeographicCRS,lon,lat per sounding, followed by several PointInstructions, one per digit glyph (e.g. SOUNDG21;SOUNDG13;SOUNDG08 for a three-digit depth). The parser reset the augmented anchor to null after the first PointInstruction, so every glyph but the first fell back to the feature's primary geometry point and piled up on the same spot.

The fix treats AugmentedPoint as a spatial-geometry context (S-100 Part 9 §11.5) that persists across all following point/text instructions until the next AugmentedPoint or ClearGeometry, rather than as per-instruction state. Per-instruction style state (rotation, scale, offset, text alignment) is still reset as before.

This PR also includes the earlier companion fix for the translucent stacked squares around soundings (dedupe of hit-test rectangles to one per feature/anchor) and removes temporary pivot debug logging used during investigation. Confirmed visually in the viewer: the affected soundings now render as a single value.

Spec alignment

• S-101 ENC (s101-enc)
• S-100 framework (s100-framework)
• S-102 bathymetry (s102-bathymetry)
• S-104 water level (s104-water-level)
• S-111 surface currents (s111-surface-currents)
• S-124 navigational warnings (s124-nav-warnings)
• S-129 under keel clearance (s129-ukc)
• N/A - change is purely infrastructural (build, CI, docs, tooling)

Spec section references cited in code/docs: S-100 Part 9 §11.5 (AugmentedPoint spatial-geometry context persistence), cited in the parser comments.

Note: the parser lives in EncDotNet.S100.Core and is shared by all vector products, but the bug only manifests for products that emit multiple point instructions under a single AugmentedPoint (S-101 multipoint soundings). Behaviour for single-point features is unchanged, since the anchor stays null when no AugmentedPoint is emitted.

Tests

• Added/updated xunit tests under tests/
• Tests requiring real data files use SkippableFact
• dotnet test --configuration Release passes locally

The parser regression test was updated to assert that the anchor persists across all glyphs of a multipoint sounding and switches on the next AugmentedPoint.

Documentation

• Updated the affected project's src/<project>/README.md
• Updated conceptual docs under docs/ if user-facing behaviour changed
• New public APIs have XML doc comments

N/A - no public API surface changed; behaviour-only fix with in-code comments citing the spec.

Dependencies

• No new NuGet dependencies, OR versions added to Directory.Packages.props (not in the .csproj)
• gh-advisory-database security check run for any new dependency

Breaking changes

None.

[github-actions]

Performance Gate

✅ PASSED — no regressions.

Threshold: 10.0%, MAD multiplier (k): 3.0, retry-zone mult: 2.0×

Scenario summary

Scenario	Status	Δ median (%)	z (Δ/MAD)	Base median (ms)	Samples (b/c)
exchange-set-open	✅ pass	-1.1	-1.27	1.30	5/5
s101-portray-cold	✅ pass	+2.5	+0.90	0.26	5/5
s101-portray-warm	✅ pass	+17.3	+1.64	0.21	5/5
s101-real-cold	✅ pass	+12.6	+5.45	0.27	5/5
s101-real-warm	✅ pass	+5.3	+1.97	0.29	5/5
s101-render-warm	✅ pass	+18.7	+5.33	0.21	5/5
s102-coverage	✅ pass	+10.2	+1.26	0.17	5/5
s102-coverage-open	✅ pass	+2.0	+1.27	2.81	20/20
s102-coverage-render-large	✅ pass	+7.6	+1.61	0.19	5/5
s102-real-warm	✅ pass	+1.3	+0.37	0.31	5/5
s111-real-warm	✅ pass	+2.6	+0.73	0.29	5/5
s124-vector	✅ pass	-15.1	-3.15	0.25	5/5
s201-vector	✅ pass	+33.0	+22.12	0.17	5/5

exchange-set-open

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	1.30	1.28
Baseline MAD (ms)	0.01	—
Δ median	—	-1.1%
z (Δ/MAD)	—	-1.27

Spans (sum of all iterations)

Span	Baseline (ms)	Candidate (ms)	Delta	Status
s100.asset.read	2.50	2.47	-1.1%	▫️
s100.exchangeset.parse	44.94	44.30	-1.4%	▫️

Metrics

Metric	Baseline	Candidate	Delta	Status
s100.asset.read.duration	0.19	0.23	+26.3%	❌

s101-portray-cold

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.26	0.27
Baseline MAD (ms)	0.01	—
Δ median	—	+2.5%
z (Δ/MAD)	—	+0.90

s101-portray-warm

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.21	0.24
Baseline MAD (ms)	0.02	—
Δ median	—	+17.3%
z (Δ/MAD)	—	+1.64

s101-real-cold

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.27	0.30
Baseline MAD (ms)	0.01	—
Δ median	—	+12.6%
z (Δ/MAD)	—	+5.45

s101-real-warm

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.29	0.30
Baseline MAD (ms)	0.01	—
Δ median	—	+5.3%
z (Δ/MAD)	—	+1.97

s101-render-warm

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.21	0.25
Baseline MAD (ms)	0.01	—
Δ median	—	+18.7%
z (Δ/MAD)	—	+5.33

s102-coverage

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.17	0.18
Baseline MAD (ms)	0.01	—
Δ median	—	+10.2%
z (Δ/MAD)	—	+1.26

s102-coverage-open

Iteration statistics

Stat	Baseline	Candidate
Samples	20	20
Median (ms)	2.81	2.87
Baseline MAD (ms)	0.05	—
Δ median	—	+2.0%
z (Δ/MAD)	—	+1.27

Spans (sum of all iterations)

Span	Baseline (ms)	Candidate (ms)	Delta	Status
s100.dataset.open	452.51	453.48	+0.2%	▫️
s100.hdf5.dataset.read	164.21	165.24	+0.6%	▫️
s100.hdf5.file.open	18.38	19.10	+3.9%	▫️
s100.hdf5.open	18.58	18.68	+0.5%	▫️

Metrics

Metric	Baseline	Candidate	Delta	Status
s100.hdf5.read.bytes	36456.00	36456.00	+0.0%	▫️
s100.hdf5.read.duration	27.58	26.77	-2.9%	▫️
s100.hdf5.read.duration	32.36	31.75	-1.9%	▫️
s100.hdf5.read.duration	11.11	11.05	-0.6%	▫️

s102-coverage-render-large

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.19	0.21
Baseline MAD (ms)	0.01	—
Δ median	—	+7.6%
z (Δ/MAD)	—	+1.61

s102-real-warm

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.31	0.32
Baseline MAD (ms)	0.01	—
Δ median	—	+1.3%
z (Δ/MAD)	—	+0.37

s111-real-warm

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.29	0.30
Baseline MAD (ms)	0.01	—
Δ median	—	+2.6%
z (Δ/MAD)	—	+0.73

s124-vector

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.25	0.21
Baseline MAD (ms)	0.01	—
Δ median	—	-15.1%
z (Δ/MAD)	—	-3.15

s201-vector

Iteration statistics

Stat	Baseline	Candidate
Samples	5	5
Median (ms)	0.17	0.23
Baseline MAD (ms)	0.00	—
Δ median	—	+33.0%
z (Δ/MAD)	—	+22.12

---

Generated by EncDotNet.S100.PerfReport gate command