fix(constellation): assemble multipart MT SMS before matching

[HumbleDrummer]

Summary

• assemble the ordered PDU parts from one SMS_RECEIVED broadcast into a single logical message before matching
• reject empty broadcasts and missing body parts instead of accepting a partial message
• preserve single-part behavior and use the first available originating address

Why

MtSmsInbox currently converts each PDU into a separate ReceivedSms. For a multipart SMS, a challenge split across segments can never match as one body, and the segments are buffered separately. This change assembles the ordered parts delivered in the broadcast before buffering and matching.

Verification

• :play-services-constellation-core:testDebugUnitTest — 14 tests, 0 failures/errors
• :play-services-constellation-core:testReleaseUnitTest — 14 tests, 0 failures/errors
• Debug and Release assemble/lint passed for play-services-constellation and play-services-constellation-core
• git diff --check passed
• remote branch content was compared with the locally tested patch and matches exactly

The full monorepo build could not complete in the offline audit environment because com.android.volley:volley:1.2.1 was not present in the local cache. Dependency resolution stopped in firebase-auth-core, outside the changed modules.

Scope

This is a narrow correctness follow-up on top of microg#3784. It does not claim to solve all of microg#2994.

No physical SIM/locked-bootloader RCS end-to-end test was performed. The existing microg#3784 hardware-validation gap remains.

Implementation and tests were prepared with Codex assistance; both commits include explicit co-author attribution.

[qodo-code-review]

ⓘ Qodo reviews are paused because the subscription is no longer active. Ask your workspace admin to reactivate the subscription to resume reviews. Manage billing

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-14T10:58:33.062075Z	6af8c09	PR opened
🔒 Security Review	✅ Completed 2026-09-14T10:58:29.107157Z	6af8c09	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[HumbleDrummer]

@naormeit @opstic @unpluggederan — we are looking for one experienced ROM builder/tester for a narrow hardware check of microg#3784 together with this follow-up at commit 6af8c095.

The main blocker is deployment: on a locked-bootloader device, the combined GmsCore build needs signing compatible with the tester's microG ROM. Is there already a prepared build/ROM path for this branch?

@opstic, you previously reported testing the MT-SMS flow; @unpluggederan, you coordinated the wider carrier testing. Would either of you be able to test this head, or point us to a ROM builder/tester who already has:

• a physical microG device with a locked bootloader and no root/Magisk;
• an active SIM and current Google Messages;
• a second RCS endpoint for send/receive?

We can provide a short checklist and log filters. Please share only manually redacted evidence; raw logcat may contain phone numbers or tokens. With the tester's consent, we will credit the public report as Hardware-tested-by.

This is a volunteer verification request. No payment or bounty share is promised, and this is not a full microg#2994 bounty claim.