test(map): assert iteration Name in 9-13 so item namer is actually verified

[wangyb-A]

Problem

Requirement 9-13 (map with a custom item namer) identified its two MapIteration
events by placeholder id (${ITER0}, ${ITER1}) and never asserted Name. The
runner only checks keys present in the expected event and ignores extra actual keys,
so an SDK that accepted an item namer and then silently ignored it still passed
9-13. The requirement only proved that setting the option did not break results or
history shape.

Change

Name is an assertable top-level field on ContextStarted / ContextSucceeded.
9-13 now asserts it literally on all six context events:

• Map context: named-items
• iteration 0: item-1
• iteration 1: item-2

Iteration ids stay as placeholders — the name is what the feature controls, the id
is not.

Validation

Ran the full map suite against us-west-2 for all three SDKs with a runner built
from this branch:

SDK	map suite	9-13
JS	20 PASSED	PASSED
Python	20 PASSED	PASSED
Java	15 PASSED / 5 NOT_IMPLEMENTED / 0 FAILED	PASSED

The recorded histories confirm the assertion is genuinely exercised rather than
skipped — all three SDKs emit identical names:

ContextStarted   Map           Name='named-items'
ContextStarted   MapIteration  Name='item-1'
ContextSucceeded MapIteration  Name='item-1'
ContextStarted   MapIteration  Name='item-2'
ContextSucceeded MapIteration  Name='item-2'
ContextSucceeded Map           Name='named-items'

Source-level confirmation that each SDK forwards the namer's return value verbatim
rather than decorating it:

• JS — map-handler.ts sets name: config.itemNamer(item, index); concurrent-execution-handler.ts resolves item.name || item.id
• Python — operation/map.py sets name=config.item_namer(item, i); concurrency/executor.py returns the bound name when present, else {prefix}{index}
• Java — MapOperation passes itemNamer.apply(item, i) straight to enqueueItem

111 runner unit tests pass.

Also

Dropped the stale note claiming the Java MapConfig has no item-namer field and
that 9-13 is therefore expected to be permanently uncovered for Java. Java now
covers it via aws/aws-durable-execution-sdk-java#576.

[github-actions]

Claude AI review

Review — PR #55: assert iteration Name in map/9-13

No actionable findings. The change is a single data-only edit to test-requirements/map/9-13.yaml and is correct, schema-compatible, and internally consistent.

Verification performed:

• Matcher premise confirmed (src/aws_durable_execution_conformance_tests/history.py:194-200): matching recurses over expected keys only and errors only when an expected key is missing from the actual event. So before this PR, Name was never in the expected dict and an SDK that accepted the item namer but ignored it still matched. Adding Name as a literal makes it enforced.
• Name is an established assertable top-level field on lifecycle events — used literally in wait/2-2.yaml, wait/2-3.yaml, wait_for_condition/6-3.yaml and as placeholders in invoke/5-2.yaml. Asserting it on ContextStarted/ContextSucceeded fits the schema; the PR's recorded histories confirm the field is present on actual context events.
• Internal consistency: the six added values (named-items, item-1×2, item-2×2, named-items) match the invocations prose and the map(name="named-items", …) call. Handler text was correctly tightened from "from the item and its index" to "from the item": with items [1,2], names item-1/item-2 are item-value-derived (index-derived would be item-0/item-1), so the edit removes a real ambiguity.
• Iteration ids correctly remain placeholders (${ITER0}/${ITER1}); only Name is asserted literally, matching the stated intent.

Residual test risk (informational):

• Positive assertion only — no dedicated negative fixture; the "an ignoring SDK must fail" guarantee rests on matcher behavior, not its own regression test.
• Removing the "Java permanently uncovered" note makes 9-13 a hard Java requirement coupled to cross-repo aws/aws-durable-execution-sdk-java#576. Where that SDK change is absent, Java 9-13 will now FAIL rather than be treated as unimplemented. Intended per the PR, but not verifiable from this repository.

Reviewed commit 6399cda335a19718d7a7e7f52e2e4146f7178359. Workflow run

[github-actions]

Codex AI review

No actionable findings.

Residual risk: cross-SDK integration coverage is required to confirm successful context events consistently emit the asserted Name fields; repository code was not executed during this review.

Reviewed commit 6399cda335a19718d7a7e7f52e2e4146f7178359. Workflow run