Per-connector / per-cable tweak with name placeholder (port of upstream PR #357)

[SomethingNew71]

Summary

Ports upstream PR #357 — connectors and cables can now carry their own `tweak:` block (with the same `override` / `append` shape as the global one), and a placeholder substring gets rewritten to the node's actual designator at instantiation. Makes it practical to author one tweak template and apply it to many components.

Example

```yaml
tweak:
placeholder: "@@"

connectors:
X1:
pinlabels: [A, B]
tweak:
append:
- "@@_extra [color=red, style=dashed];"
```

renders as `X1_extra [color=red, style=dashed];` in the .gv. Same connector definition reused as X2 → `X2_extra ...`.

Placeholder semantics

Per-node `placeholder`	Effect
`None` (default)	Falls back to global `tweak.placeholder`
Empty string `""`	Explicitly opts out of substitution for this node
Any string	Used as-is for substitution

Implementation

• DataClasses.py — `Tweak` gains `placeholder: Optional[str] = None`. `Connector` and `Cable` gain `tweak: Optional[Tweak] = None`, with `post_init` coercing dict literals to `Tweak` instances.
• Harness.py — new `Harness._extend_tweak(node)` method, called from `add_connector()` and `add_cable()`. Performs placeholder substitution and merges into `self.tweak`. Raises `ValueError` if two nodes contribute conflicting overrides.
• docs/syntax.md — documents new fields.

Differences from upstream PR wireviz#357

Renamed `extend_tweak` to `_extend_tweak` to mark it private. Otherwise faithful to the original logic. `make_list` already exists in master's `wv_bom` so no helper backport needed.

Verification

• ✅ Smoke test: per-connector + per-cable `append:` blocks produce `X1_extra`, `W1_label`, `cable W1` in rendered .gv (placeholders substituted correctly)
• ✅ `build_examples.py` deterministic outputs byte-identical (no existing example uses the new syntax)

🤖 Generated with Claude Code

[SomethingNew71]

@kvid — heads up, your upstream #357 (per-connector/per-cable tweak with name placeholder) is being incorporated here.

Context: We use WireViz to document Classic Mini Cooper wiring harnesses at Classic Mini DIY and are building a GUI on top of it. Upstream wireviz/WireViz looks abandoned, so we're forking and pulling in the useful open PRs.

Adapted faithfully — only renamed the public method extend_tweak to private _extend_tweak since callers shouldn't reach into it directly. The placeholder fallback semantics (None falls through to global, empty string opts out) and the conflict-detection in override merging are preserved as-is. make_list is already in current master's wv_bom, so no helper backport needed.

Commit attribution links back to your PR. Thanks for the work — clean design.

[gemini-code-assist]

Code Review

This pull request introduces per-connector and per-cable tweak overrides with placeholder substitution, allowing users to define templates that are automatically customized with the specific node's name. The changes include documentation updates, new fields in the Tweak, Connector, and Cable data classes, and the implementation of the _extend_tweak method in Harness.py to handle the merging and substitution logic. Review feedback identified two potential crash scenarios in the new logic: an AttributeError when the substitution lambda processes non-string values (such as null in YAML), and a potential TypeError in graph generation caused by setting override entries to None instead of maintaining them as dictionaries.

[gemini-code-assist]

The rph lambda will raise an AttributeError if it is called with a None value. This can happen in the override dictionary when an attribute is being deleted (set to null in YAML). The lambda should be updated to handle non-string values gracefully.

Suggested change

	rph = (lambda s: s.replace(ph, node.name)) if ph else (lambda s: s)
	rph = (lambda s: s.replace(ph, node.name) if isinstance(s, str) else s) if ph else (lambda s: s)

[gemini-code-assist]

Setting s_override[ident] to None when s_dict is empty will cause a crash in Harness.create_graph() (specifically at line 629) because it expects the values in the override dictionary to be dictionaries, not None. It is safer to keep it as an empty dictionary.

Suggested change

	s_override[ident] = s_dict or None
	s_override[ident] = s_dict