Cover the line reader's escapes, comments and quoted switch

[dmccoystephenson]

Stage B test expansion: the server-wide config line reader and the installId builder precondition. Only TraceClientTest.java is changed. Production code is untouched.

Summary

• serverWideTags_doubleQuotedValuesUnescapeNewlineTabCarriageReturnAndBackslash: covers the \n, \t, \r, \\ and unknown-escape branches of scalar(), and confirms that single-quoted values do not unescape. Before this, only \" had a test.
• serverWideTags_dropValuesThatOpenAnAnchorAliasTagOrFoldedScalar: covers the &, *, !, %, @, backtick and > indicators. Before this, only [, { and | had a test.
• serverWideTags_aHashIsACommentOnlyAfterWhitespace: issue#13 stays whole, a tab before # starts a comment, a tab after the colon separates the value, and a tab-indented line under a space-indented block is skipped.
• serverWideConfig_theSwitchIgnoresATrailingCommentAndLeadingIndent: enabled: false # ..., an indented enabled: outside tags:, and enabled :\tFALSE all disable.
• serverWideConfig_aQuotedFalseDoesNotDisableYet: characterization. enabled: "false" currently leaves reporting on. This is filed as A quoted enabled: "false" in the server-wide config does not turn reporting off #15 (opt-out contract, needs a maintainer decision). The test pins today's behaviour and should be flipped by whatever PR resolves A quoted enabled: "false" in the server-wide config does not turn reporting off #15.
• builder_rejectsAnOverlongInstallIdAndAcceptsOneAtTheLimit: the IllegalArgumentException message, and a 255-character ID with surrounding whitespace that is accepted after trimming.

Test plan

• mvn -B verify locally (Java 21): Tests run: 72, Failures: 0, Errors: 0, Skipped: 0, and the 6 new tests appear in the surefire report
• CI matrix build (8), build (17), build (21) green (the new tests use only Java 8 APIs)

Deferred issues

• withVersion pushes an event already at MAX_TAGS over the server's tag limit #13 (withVersion exceeds MAX_TAGS): skipped because it changes what is sent on the wire and is listed as a maintainer decision.
• A quoted enabled: "false" in the server-wide config does not turn reporting off #15 (quoted enabled:): filed this cycle and skipped for the same reason, since it changes the opt-out contract.

No tracking issue — gap found during triage.

This PR description was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

🤖 Generated with Claude Code

---

drafted by Claude on behalf of Daniel Stephenson

[dmccoystephenson]

Self-review rubric (CI anchor green on the PR head: build (8), build (17), build (21) all pass, run 37757109192):

• Scope: PASS. git diff --stat origin/main shows one file, TraceClientTest.java (+81, test code only).
• Tests-new: PASS (n/a). No new public members. All six tests target existing, previously undriven branches of scalar(), parseServerWideConfig and Builder.installId.
• Tests-fix: n/a. No production change. This is a characterization-only Stage B cycle.
• Sibling structure: PASS. The new tests follow the subject_behaviourInPlainWords naming, sit next to their siblings (serverWideConfig_*, serverWideTags_*, installId_*), and use the tagsOf / parseServerWideConfig helpers the neighbouring tests use.
• Sibling renames: n/a. Nothing renamed.
• Docs: PASS. No behaviour changed, so no docs row moves. The README "Server-wide tags" bullets (quoting, # comments, rejected constructs) are consistent with what the new tests pin.
• Issue resolution: n/a. No Closes. A quoted enabled: "false" in the server-wide config does not turn reporting off #15 was filed for the gap found and is referenced, not closed.
• CI: PASS. All three matrix legs are green. Locally mvn -B verify reported Tests run: 72, Failures: 0, and the surefire report lists all 6 new tests.
• No dependencies / Single file / Java 8: PASS. pom.xml is unchanged, git ls-files src/main still lists only TraceClient.java, and the build (8) leg is green.
• Never throws / Non-blocking / Environment seam / Version agreement / Opt-out contract: PASS (unchanged). No production code was touched.

Judgment calls flagged for the reviewer:

• src/test/java/software/stephenson/trace/TraceClientTest.java serverWideConfig_aQuotedFalseDoesNotDisableYet: the Yet suffix and the comment presume A quoted enabled: "false" in the server-wide config does not turn reporting off #15 will flip this assertion. If the maintainer decides to keep quoted values meaningful-as-strings, the test should be renamed and the comment dropped rather than flipped.
• serverWideTags_aHashIsACommentOnlyAfterWhitespace pins that a tab-indented entry under a space-indented block is silently skipped. That is current, documented-by-Javadoc behaviour ("lines indented differently from its first entry"), but an operator mixing tabs and spaces could find it surprising.

Summary: test-only PR, CI green on all legs. No do-not-auto-merge path matched. A human merge is still required, because this dispatch is not authorized to merge.

This review was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

---

drafted by Claude on behalf of Daniel Stephenson