Rename internal `canonicalise` identifiers to Oxford `-ize` spelling

[lodyai]

Summary

Several of netsuke's own identifiers, localization key constants, and Fluent
message IDs spell "canonicalise" with the British -ise ending. Per the project
convention, identifiers should use en-GB-oxendict (Oxford -ize) spelling unless
they reference an external API. These are internal names, so they should be
canonicalize.

The rest of the codebase already follows this — the cycle module, hasher, and
runner::process::paths all use canonicalize, and the external
std::fs::canonicalize call is (correctly) spelled -ize. The stdlib::which
and stdlib::path modules are the outliers.

This was surfaced while adding the typos spelling gate (#407): references to
canonicalise in the docs had to be wrapped in backticks precisely because they
name these real identifiers. Renaming the identifiers is the proper fix; the doc
backticks can then be removed.

Affected identifiers (internal — should become -ize)

Enum variants and helpers in the which resolver:

• src/stdlib/which/resolve_error.rs: ResolveError::Canonicalise,
  ResolveError::CanonicaliseNonUtf8, and their string discriminants
  "canonicalise" / "canonicalise_non_utf8"
• src/stdlib/which/lookup/mod.rs: pub(super) fn canonicalise(...) and its
  call sites (lines 65, 96, 113, 234)
• src/stdlib/which/mod.rs: fn canonicalise_message(...) and the match arms
  over the renamed variants

Localization key constants in src/localization/keys.rs:

• STDLIB_PATH_ACTION_CANONICALISE
• STDLIB_WHICH_CANONICALISE_FAILED
• STDLIB_WHICH_CANONICALISE_NON_UTF8

Fluent message IDs (the string values of those constants), which must be renamed
in lockstep across every locale or the build-time l10n audit
(build_l10n_audit.rs) will fail:

• stdlib.path.action.canonicalise
• stdlib.which.canonicalise_failed
• stdlib.which.canonicalise_non_utf8
• Files: locales/en-US/messages.ftl, locales/es-ES/messages.ftl

Tests and doc comments:

• tests/std_filter_tests/which_filter_tests.rs: test name
  which_filter_all_with_duplicates_deduplicates_canonicalised_paths
• Doc comments in src/runner/process/paths.rs,
  src/stdlib/which/lookup/mod.rs, and src/stdlib/path/path_utils.rs
• The context string "canonicalise stdlib cache URL" in
  tests/bdd/steps/stdlib/assertions.rs

Explicitly out of scope (keep as-is)

• std::fs::canonicalize (src/stdlib/which/lookup/mod.rs:264) — external API,
  already -ize.

Suggested acceptance criteria

• All internal canonicalise/Canonicalise identifiers renamed to the
  Oxford -ize spelling.
• The three localization key constants and their Fluent message IDs renamed
  in lockstep across keys.rs, locales/en-US/messages.ftl, and
  locales/es-ES/messages.ftl, with make test / the build-time l10n audit
  green.
• Doc-comment and test-name occurrences updated.
• Doc references to canonicalise that were backticked in Enforce en-GB-oxendict spelling with typos #407 (for
  example in
  docs/execplans/3-14-4-command-available-non-throwing-executable-probe.md)
  updated to the new identifier where they name the code.
• make check-fmt, make lint, make test, and make markdownlint pass.

[coderabbitai]

🔗 Related PRs

leynos/lag-complexity#11 - docs: tidy design doc headings and spelling [closed]
leynos/femtologging#350 - Standardize API spelling to oxendict (-ize) across codebase [closed]
#311 - Record Verus and Stateright phase-1 boundary (4.1.3) [closed]
#335 - Promote command_available to non-throwing executable probe (3.14.4) [closed]
leynos/ortho-config#348 - Promote LocalizeCmd to a public extension trait (11.1.1) [closed]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!