Compare commits

...

4 Commits

Author SHA1 Message Date
Dai Ha 94ec77a1bc t358: guard MemberSession's 5 rebuild sites and Profile.withProfile() against the back-compat-arity trap
CI / contract (pull_request) Successful in 1m23s
CI / build (pull_request) Successful in 1m53s
Follow-up to #357 (FleetConfig.withDefaults()). Reflectively enumerate each
record's own components, resolve the canonical constructor by exact
component types, build a real non-null value per component, run each
rebuild site, and assert every component survives (except the one it is
documented to change). Exclusion lists pinned at 0 for both.

Re-counted the Profile back-compat ladder directly against the source:
8 constructors (arities 25, 24, 22, 20, 18, 15, 14, 12) against a
canonical arity of 26 — the ticket's own number was explicitly untrusted.
2026-09-09 07:19:12 +07:00
Dai Ha 49f285cfda charter: a measured fact in an addendum must carry its own deletion trigger
CI / contract (push) Successful in 49s
CI / build (push) Successful in 1m36s
The operator chose this and its size; the reasoning below is the fleet01 lead's.

Placed in the canonical block's boundary paragraph rather than the orchestration
body. That paragraph already talks about the addendum layer instead of protocol,
every project that mounts the bridge inherits it, and it sits about 3800 characters
before the primary's step list, so it does not dilute the steps a lead reads while
working. Perishability is structurally an addendum problem: the block is
byte-identical across projects by construction, so a dated local measurement in the
block body would already be a layering violation.

What happened. The fleet01 lead's kb addendum held a dated merge-refusal section
that carried an instruction to delete itself once it stopped reproducing. On
2026-09-08 UTC the operator granted merge rights on akb/kb, the lead re-ran the
probe, got 409 'head out of date' where the identical request had returned 405
'User not allowed to merge PR' on 2026-09-06, and deleted the section as instructed.

Why four parts and not one. The lead's finding is that the banner did not work
because it was emphatic. It worked because the falsification condition was
executable: it carried the exact probe, the reason for the all-zeroes
head_commit_id, and what each response code meant. The lead did not have to
reconstruct the experiment or decide what would count as refutation, and just ran
it. A banner saying 'this may be out of date, verify before relying on it' costs
the same space and does nothing, because deciding what would falsify a claim is the
expensive step and a reader in the middle of another task will not pay it. So: the
date, the command, what each outcome means, and the instruction to delete. The
fourth without the second is decoration.

The closing clause is the justification for the machinery. Most stale notes are
merely wrong. This one went stale in the dangerous direction: it would have told a
future lead it could not merge at the exact moment merging became its job, silently
and with confidence. A note that goes harmlessly stale does not need this.

Note what is NOT centralized here. The banner text itself cannot be. What fired for
the lead was a specific instruction sitting on top of the specific stale fact, which
it could not read past on its way to acting. A rule elsewhere saying 'date your
measurements' would not have fired, because nobody reads that rule at the moment
they re-measure. This sentence sets the convention; the trigger still has to live
next to the fact it governs.

Propagated to the wiki template in the same turn, wiki 8c4f152 on main; the sync
check in this file's addendum reports 'in sync: True'.
2026-09-09 04:24:17 +07:00
Dai Ha 5a12ae7930 charter: test a refusal, and do not count a transport failure as one
CI / contract (push) Successful in 1m11s
CI / build (push) Successful in 1m38s
Step 8 gained a refusal paragraph in 2f71a30, which said what a lead does once the
forge refuses a merge. It did not say how a lead establishes that it was refused.
Both halves of this amendment come from the fleet01 lead, measured on akb/kb on
2026-09-08 UTC, and both are ways to be wrong about a permission you never tested.

Do not read a refusal off a permissions field. After the operator granted merge
rights, the lead re-ran its probe: POST .../pulls/53/merge with an all-zeroes
head_commit_id, chosen so the request cannot succeed on its merits and a rejection
can only mean the refusal. It returned 409 'head out of date' where the identical
request returned 405 'User not allowed to merge PR' on 2026-09-06. A 409 is payload
validation and sits after the permission gate, so the grant took. The lead reports
the repository permissions object did not change across that flip -- still
admin:false, push:true, pull:true. I did not read that object myself; my forge token
is a different identity and would return a different one, so this stays the lead's
measurement and not mine. Merge rights on a protected branch live in branch
protection, so a permissions field can be wrong in both directions.

Do not count a transport failure as a refusal. The lead's first attempt returned
HTTP 000, because GITEA_HOST already carries a scheme and a trailing slash and the
URL came out as https://https://git.ltms.dev//api/... Under a 'not 200' test that is
indistinguishable from being refused. A probe exists to separate a refusal from
everything else, so an error that never reached the gate has to be a third answer
that concludes nothing.

Propagated to the wiki template in the same turn, wiki 8c2ef96 on main; the sync
check in this file's addendum reports 'in sync: True'.
2026-09-09 04:09:08 +07:00
Dai Ha 127e6832a9 Merge #374: fleetd holds off idle sleep while any member is live
CI / contract (push) Successful in 1m18s
CI / build (push) Successful in 2m3s
Lands PR #355 (fleetd #354's sibling), rebased onto current main by a worker
after 39 commits of drift left it unmergeable.

The problem, measured on the original branch: a fleetd host idle-slept after as
little as one minute (pmset -g custom reported 'sleep 1' on battery). Overnight
the daemon's AMQP link dropped 13 times, and every drop minute had a sleep or
wake event in pmset -g log in the same minute or the one before. The AMQP churn
is the visible symptom; the real cost is a member mid-turn freezing with the
host, and a long turn with nobody typing is exactly the case that goes idle.

IdleSleepGuard holds an OS-level assertion for as long as at least one member is
live. It is driven by SessionManager's existing onAcquire/onRelease hooks rather
than a second member count kept in parallel, so it reads the same registry
fleet_list's numbers come from, and only a real 0->1 or 1->0 crossing touches the
OS. It fails safe: a mechanism that cannot acquire means nothing is ever held,
and it never throws, never blocks a spawn, a release, or shutdown.

Conflict resolution was the whole job, and all three were in config plumbing:
ConfigRef, FleetConfig and ConfigRefTopLevelReportingCoverageTest. The power
package is byte-identical to the original branch commit.

Verified on this merge, not taken from the worker's report:
  mvn clean install -> Tests run: 1439, Failures: 0, Errors: 0, BUILD SUCCESS
  (1425 on main + 14 new: 4 caffeinate, 5 guard, 1 wiring, 4 config)

The denominator recount, which the worker flagged as its own weakest number
because this file's count has drifted three times before (#330/#333/#337). I
counted it mechanically rather than reading it: FleetConfig has 24 canonical
record components; COLD_KEYS 5, DEFERRED_KEYS 13, SPLIT_KEYS 3, plus the 3 the
javadoc names as hot-excluded (placement, memberCredentials, memberLoginShell).
5+13+3+3 = 24. The javadoc's '24 components: 5 cold, 13 deferred, 3 split, 3
hot-excluded' is correct. The worker's prose called idleSleepGuard the 25th
constructor argument; it is the 24th. The code is right, the report was off by
one.

Mutation run on merge, on the half the worker verified by READING rather than by
proving -- it said it had checked that withDefaults()'s final call binds the true
canonical constructor. I dropped the trailing idleSleepGuard argument so the call
silently binds the 23-arg back-compat overload. It compiles, which is the whole
hazard. Caught: 1 failure, 3 errors, BUILD FAILURE, and
FleetConfigWithDefaultsPreservesEveryComponentTest names the dropped component
and prints its own denominator -- '24 components, 24 checked, 0 excluded, 23
survived'. That test was added on main after this exact defect happened live when
idleSleepGuard was added on a sibling branch; the worker had to add the missing
entry to it, and doing so is what makes the guard cover this component at all.
2026-09-07 20:31:41 +07:00
3 changed files with 376 additions and 0 deletions
+14
View File
@@ -7,6 +7,14 @@
> wiki ([Use Cases](https://git.ltms.dev/fleet/fleetd/wiki/7-Use-Cases) → *The portable
> CLAUDE.md block*); improvements go to the template first, then out to each project. Anything
> specific to *this* repo lives under §Project addendum below, never inline above it.
>
> **Anything you measure in an addendum is perishable.** Date it, give the command that
> re-measures it and what each outcome means, and tell the reader to delete the section once
> it stops reproducing. The four parts work together: deciding what would falsify a claim is
> the expensive step, and a reader in the middle of another task will not pay it, so a bare
> "verify before relying on this" costs the same space and does nothing. The case this is for
> is a note that goes stale as a live restriction — it will tell a future session it cannot do
> the thing at the moment doing it becomes the job.
If no `fleet_*` MCP tools are mounted in this session, this section does not apply — skip it.
@@ -100,6 +108,12 @@ below are the procedure — run them in order, every task, not only the big ones
without having read the diff yourself. A refusal is exactly when that shortcut is tempting,
because no action is left that forces you to look, and taking it turns this step into
forwarding a reviewer's verdict — which is delegating the merge by proxy, two lines above.
**Test a refusal; do not read it off a permissions field.** A protected branch holds its merge
rights separately from the repository permissions, so that field can say yes while the merge is
refused, and still say no after a grant makes it work. Probe instead, with a request that cannot
succeed on its merits, so a rejection can only mean the refusal. Treat a transport failure as a
third answer that proves nothing: a timeout, a DNS error or a bad URL is not a refusal, and
counting it as one makes you sure of something you never measured.
**Steps 3 and 4 are separate on purpose** — spawning and sending in one loop is how parallel work
silently becomes serial, and it is the most common way this layer is wasted. For the same reason,
@@ -0,0 +1,172 @@
package dev.ltms.fleet.config;
import org.junit.jupiter.api.Test;
import java.lang.reflect.Constructor;
import java.lang.reflect.RecordComponent;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Locale;
import java.util.Map;
import java.util.Objects;
import java.util.Set;
import java.util.TreeSet;
import static org.junit.jupiter.api.Assertions.assertEquals;
/**
* Fleetd #358, the same "defect factory" #357 guarded on {@code FleetConfig.withDefaults()}
* (see {@code FleetConfigWithDefaultsPreservesEveryComponentTest}), reproduced here on
* {@link FleetConfig.Profile}. {@code Profile} carries a long back-compat constructor ladder — 8
* constructors, re-counted directly against the source rather than trusted from the ticket, at
* arities 25, 24, 22, 20, 18, 15, 14 and 12, against a canonical arity of 26 — and exactly ONE
* rebuild site, {@link FleetConfig.Profile#withProfile(String)}, whose own
* {@code return new Profile(...)} call is written at a literal 26-arg count. Add a 27th component
* and its established back-compat constructor at the old (26-arg) arity, and {@code withProfile}'s
* own call becomes a legal match for that new overload — silently dropping the new component every
* time a profile's name is defaulted from its {@code workers:} key.
*
* <p>Builds one {@link FleetConfig.Profile} through the TRUE canonical constructor — resolved by
* the record's own component types via {@code getDeclaredConstructor}, never by argument count —
* with a real, distinctive, non-null value in every component, calls {@link
* FleetConfig.Profile#withProfile(String)}, and asserts every component except {@code profile}
* itself survives unchanged, while {@code profile} comes back as the new name it was given.
*
* <p>Every value here is chosen so {@code Profile}'s own compact constructor (which normalizes
* several components — defaults {@code argv}/{@code kind}/{@code placement}/{@code workspace}/
* {@code gitHostEnv}, nulls a handful of blank-checked strings, clamps {@code weight}, coerces
* {@code subscription}) leaves it unchanged: every String is non-blank and already in the shape the
* compact constructor would otherwise coerce it to (e.g. {@code placement} is already lowercase),
* and every collection is non-empty. That is what makes "must survive unchanged" a valid assertion
* for every component below, the same reasoning {@code FleetConfigWithDefaultsPreservesEveryComponentTest}
* documents for {@code withDefaults()}.
*
* <p>{@link #EXCLUDED_FROM_SURVIVAL_CHECK} is kept deliberately empty and size-pinned by
* {@link #exclusionListSizeIsPinned()} — a checker whose escape hatch can grow to silence a failure
* is not a checker. Every one of {@code Profile}'s 26 current components has a real, non-null,
* non-blank value here and none is excluded.
*/
class FleetConfigProfileWithProfilePreservesEveryComponentTest {
private static final RecordComponent[] COMPONENTS = FleetConfig.Profile.class.getRecordComponents();
/** Deliberately empty today; grow it only with a matching justification, and re-pin the size. */
private static final Set<String> EXCLUDED_FROM_SURVIVAL_CHECK = Set.of();
/** One real, distinctive, non-null value per component, chosen to survive the compact ctor. */
private static Map<String, Object> baseValues() {
Map<String, Object> v = new LinkedHashMap<>();
v.put("profile", "profile-guard");
v.put("baseUrl", "https://guard.example/base");
v.put("model", "model-guard");
v.put("configDir", "/config/guard");
v.put("tokenEnv", "GUARD_TOKEN");
v.put("argv", List.of("guard-cmd"));
v.put("placement", "guard-placement");
v.put("workspace", "workspace-guard");
v.put("tabLabel", "tab-guard");
v.put("mcpUrl", "https://mcp.guard/");
v.put("cwd", "/cwd/guard");
v.put("parityOverlay", List.of(".guardrc"));
v.put("gitTokenEnv", "GUARD_GIT_TOKEN");
v.put("gitHostEnv", "GUARD_GIT_HOST");
v.put("kind", "claude-code");
v.put("env", Map.of("GUARD_ENV", "1"));
v.put("weight", 2.5f);
v.put("maxLoad", 4);
v.put("subscription", Boolean.TRUE);
v.put("exhaustedPattern", "pattern-guard");
v.put("credentialId", "cred-guard");
v.put("ideMcpUrl", "https://ide.guard/");
v.put("ideProjectDir", "ide-project-guard");
v.put("ideOpenCommand", "open-guard {dir}");
v.put("autoCompactWindow", 150_000);
v.put("errorPattern", "error-pattern-guard");
assertNamesMatchComponents(v);
return v;
}
/**
* Guards {@link #baseValues()} itself against drifting from the record's real shape — forgetting
* to add a new component here fails this assertion by name, rather than silently checking one
* component fewer than the record has.
*/
private static void assertNamesMatchComponents(Map<String, Object> values) {
Set<String> names = new TreeSet<>();
for (RecordComponent rc : COMPONENTS) {
names.add(rc.getName());
}
assertEquals(names, new TreeSet<>(values.keySet()),
"this test's value map has drifted from FleetConfig.Profile's actual components — "
+ "update baseValues() alongside the record");
}
/**
* Builds a {@link FleetConfig.Profile} through the TRUE canonical constructor — resolved by the
* record's own component types, not by argument count — so this never accidentally exercises a
* back-compat overload the way a literal {@code new Profile(...)} call risks doing.
*/
private static FleetConfig.Profile profileOf(Map<String, Object> values) throws ReflectiveOperationException {
Class<?>[] types = Arrays.stream(COMPONENTS).map(RecordComponent::getType).toArray(Class<?>[]::new);
Object[] args = Arrays.stream(COMPONENTS).map(rc -> values.get(rc.getName())).toArray();
Constructor<FleetConfig.Profile> ctor = FleetConfig.Profile.class.getDeclaredConstructor(types);
return ctor.newInstance(args);
}
@Test
void exclusionListSizeIsPinned() {
assertEquals(0, EXCLUDED_FROM_SURVIVAL_CHECK.size(),
"EXCLUDED_FROM_SURVIVAL_CHECK grew from 0 — every entry needs a justification in "
+ "this test class's javadoc AND this assertion re-pinned to the new size; a "
+ "growing exclusion list that silences failures on its own is not a guard");
}
/**
* The mutation this is built to catch: make {@code withProfile(String)}'s final constructor call
* literal at some arg count, add one more component to the record with a new back-compat
* constructor at the old arity, and the stale call silently rebinds. Every component here is real
* and non-null/non-blank, so none of it should be replaced by {@code withProfile}, except
* {@code profile} itself, which the method is documented to replace.
*/
@Test
void withProfilePreservesEveryOtherComponent() throws ReflectiveOperationException {
Map<String, Object> base = baseValues();
FleetConfig.Profile profile = profileOf(base);
FleetConfig.Profile renamed = profile.withProfile("renamed-profile-guard");
List<String> dropped = new ArrayList<>();
int checked = 0;
for (RecordComponent rc : COMPONENTS) {
String name = rc.getName();
if (EXCLUDED_FROM_SURVIVAL_CHECK.contains(name)) {
continue;
}
checked++;
Object expected = "profile".equals(name) ? "renamed-profile-guard" : base.get(name);
Object actual;
try {
actual = rc.getAccessor().invoke(renamed);
} catch (ReflectiveOperationException e) {
throw new RuntimeException("failed to read FleetConfig.Profile." + name + "()", e);
}
if (!Objects.equals(expected, actual)) {
dropped.add(String.format(Locale.ROOT,
"%s: withProfile() was expected to carry (%s) for '%s' but returned %s — a "
+ "component silently dropped by withProfile(), the shape of the "
+ "defect this test exists to catch (its final \"return new "
+ "Profile(...)\" call binding to a back-compat constructor instead "
+ "of the true canonical one)",
name, expected, name, actual));
}
}
System.out.printf(Locale.ROOT,
"FleetConfig.Profile.withProfile() component-survival coverage — %d components, %d "
+ "checked, %d excluded, %d survived%n",
COMPONENTS.length, checked, EXCLUDED_FROM_SURVIVAL_CHECK.size(), checked - dropped.size());
assertEquals(List.of(), dropped,
"withProfile() silently dropped these components: " + dropped);
}
}
@@ -0,0 +1,190 @@
package dev.ltms.fleet.session;
import dev.ltms.fleet.peer.CharterReceipt;
import dev.ltms.fleet.peer.MemberRole;
import org.junit.jupiter.api.Test;
import java.lang.reflect.Constructor;
import java.lang.reflect.RecordComponent;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Locale;
import java.util.Map;
import java.util.Objects;
import java.util.Set;
import java.util.TreeSet;
import java.util.function.Function;
import static org.junit.jupiter.api.Assertions.assertEquals;
/**
* Fleetd #358, the same "defect factory" #357 guarded on {@code FleetConfig.withDefaults()}
* (see {@code FleetConfigWithDefaultsPreservesEveryComponentTest}), reproduced here on
* {@link MemberSession} — the worse of the two sibling cases named in #358, because this record
* has FIVE independent rebuild sites instead of one: {@link MemberSession#withState},
* {@link MemberSession#withActivity}, {@link MemberSession#bumpTurn},
* {@link MemberSession#withAgentSessionId} and {@link MemberSession#withFailureReason} each end in
* their own literal {@code new MemberSession(...)} call. Add a 16th component and add the
* established back-compat constructor at the old (15-arg) arity, and every one of those five
* literal calls becomes a legal match for that new overload — silently dropping the new component,
* independently, on whichever of the five paths a missed update leaves behind. That is harder to
* spot than #357's single call site: the field would survive through some transitions and vanish
* through others.
*
* <p>Each check below builds one {@link MemberSession} through the TRUE canonical constructor —
* resolved by the record's own component types via {@code getDeclaredConstructor}, never by
* argument count, so it can never itself land on a back-compat overload — with a real, distinctive,
* non-null value in every component, calls the real rebuild method under test, and asserts every
* component the method is not documented to change survives unchanged, while the component(s) it IS
* documented to change come back as the new value it was given. A component that comes back
* anything else was silently dropped or lost — the shape of the defect this test exists to catch.
*
* <p>{@link #EXCLUDED_FROM_SURVIVAL_CHECK} is kept deliberately empty and size-pinned by
* {@link #exclusionListSizeIsPinned()}, for the same reason {@code FleetConfig}'s guard pins its own
* exclusion list at zero: a checker whose escape hatch can grow to silence a failure is not a
* checker. Every one of {@link MemberSession}'s 15 current components has a real, non-null,
* non-blank value here and none is excluded.
*/
class MemberSessionRebuildPreservesEveryComponentTest {
private static final RecordComponent[] COMPONENTS = MemberSession.class.getRecordComponents();
/** Deliberately empty today; grow it only with a matching justification, and re-pin the size. */
private static final Set<String> EXCLUDED_FROM_SURVIVAL_CHECK = Set.of();
/** One real, distinctive, non-null value per component — none of the 15 is excluded. */
private static Map<String, Object> baseValues() {
Map<String, Object> v = new LinkedHashMap<>();
v.put("paneId", "pane-guard");
v.put("terminalId", "term-guard");
v.put("profile", "profile-guard");
v.put("role", MemberRole.REVIEWER);
v.put("cwd", "/wt/guard");
v.put("ownerTerminal", "owner-guard");
v.put("spawnedAtNanos", 111_111L);
v.put("lastActivityAtNanos", 222_222L);
v.put("turnCount", 7);
v.put("state", MemberSession.State.BUSY);
v.put("worktree", "/wt/guard-tree");
v.put("branch", "worker/guard-branch");
v.put("charterReceipt", new CharterReceipt(
MemberRole.DEV, "profile-guard", "fleet.charters.dev", "deadbeefguard", 42));
v.put("agentSessionId", "agent-guard");
v.put("failureReason", "reason-guard");
assertNamesMatchComponents(v);
return v;
}
/**
* Guards {@link #baseValues()} itself against drifting from the record's real shape — forgetting
* to add a new component here fails this assertion by name, rather than silently checking one
* component fewer than the record has.
*/
private static void assertNamesMatchComponents(Map<String, Object> values) {
Set<String> names = new TreeSet<>();
for (RecordComponent rc : COMPONENTS) {
names.add(rc.getName());
}
assertEquals(names, new TreeSet<>(values.keySet()),
"this test's value map has drifted from MemberSession's actual components — "
+ "update baseValues() alongside the record");
}
/**
* Builds a {@link MemberSession} through the TRUE canonical constructor — resolved by the
* record's own component types, not by argument count — so this never accidentally exercises a
* back-compat overload the way a literal {@code new MemberSession(...)} call risks doing.
*/
private static MemberSession sessionOf(Map<String, Object> values) throws ReflectiveOperationException {
Class<?>[] types = Arrays.stream(COMPONENTS).map(RecordComponent::getType).toArray(Class<?>[]::new);
Object[] args = Arrays.stream(COMPONENTS).map(rc -> values.get(rc.getName())).toArray();
Constructor<MemberSession> ctor = MemberSession.class.getDeclaredConstructor(types);
return ctor.newInstance(args);
}
@Test
void exclusionListSizeIsPinned() {
assertEquals(0, EXCLUDED_FROM_SURVIVAL_CHECK.size(),
"EXCLUDED_FROM_SURVIVAL_CHECK grew from 0 — every entry needs a justification in "
+ "this test class's javadoc AND this assertion re-pinned to the new size; a "
+ "growing exclusion list that silences failures on its own is not a guard");
}
/**
* Shared check for one rebuild site: build a base session with a real value in every component,
* call {@code rebuild}, and assert every component comes back equal to {@code expectedOverrides}
* when named there, or equal to the base value otherwise. Prints the same denominator style as
* {@code FleetConfigWithDefaultsPreservesEveryComponentTest}.
*/
private void checkRebuildSite(String siteName, Function<MemberSession, MemberSession> rebuild,
Map<String, Object> expectedOverrides) throws ReflectiveOperationException {
Map<String, Object> base = baseValues();
MemberSession session = sessionOf(base);
MemberSession result = rebuild.apply(session);
List<String> dropped = new ArrayList<>();
int checked = 0;
for (RecordComponent rc : COMPONENTS) {
String name = rc.getName();
if (EXCLUDED_FROM_SURVIVAL_CHECK.contains(name)) {
continue;
}
checked++;
Object expected = expectedOverrides.containsKey(name) ? expectedOverrides.get(name) : base.get(name);
Object actual;
try {
actual = rc.getAccessor().invoke(result);
} catch (ReflectiveOperationException e) {
throw new RuntimeException("failed to read MemberSession." + name + "()", e);
}
if (!Objects.equals(expected, actual)) {
dropped.add(String.format(Locale.ROOT,
"%s: %s() was expected to carry (%s) for '%s' but returned %s — a component "
+ "silently dropped by %s(), the shape of the defect this test exists "
+ "to catch (its final \"return new MemberSession(...)\" call binding "
+ "to a back-compat constructor instead of the true canonical one)",
name, siteName, expected, name, actual, siteName));
}
}
System.out.printf(Locale.ROOT,
"MemberSession.%s() component-survival coverage — %d components, %d checked, %d "
+ "excluded, %d survived%n",
siteName, COMPONENTS.length, checked, EXCLUDED_FROM_SURVIVAL_CHECK.size(),
checked - dropped.size());
assertEquals(List.of(), dropped,
siteName + "() silently dropped these components: " + dropped);
}
@Test
void withStatePreservesEveryOtherComponent() throws ReflectiveOperationException {
checkRebuildSite("withState", s -> s.withState(MemberSession.State.FAILED),
Map.of("state", MemberSession.State.FAILED));
}
@Test
void withActivityPreservesEveryOtherComponent() throws ReflectiveOperationException {
checkRebuildSite("withActivity", s -> s.withActivity(999_999L),
Map.of("lastActivityAtNanos", 999_999L));
}
@Test
void bumpTurnPreservesEveryOtherComponent() throws ReflectiveOperationException {
checkRebuildSite("bumpTurn", s -> s.bumpTurn(999_999L),
Map.of("lastActivityAtNanos", 999_999L, "turnCount", 8));
}
@Test
void withAgentSessionIdPreservesEveryOtherComponent() throws ReflectiveOperationException {
checkRebuildSite("withAgentSessionId", s -> s.withAgentSessionId("agent-updated"),
Map.of("agentSessionId", "agent-updated"));
}
@Test
void withFailureReasonPreservesEveryOtherComponent() throws ReflectiveOperationException {
checkRebuildSite("withFailureReason", s -> s.withFailureReason("reason-updated"),
Map.of("failureReason", "reason-updated"));
}
}