Fix $STANDARD_INFORMATION parsing for 48-byte values

[omerbenamram]

Summary

• Fix $STANDARD_INFORMATION parsing to respect the resident value length (48 vs 72 bytes).
• Avoid reading past 48-byte values into the next attribute record header (previously produced garbage owner_id/security_id/quota/usn).
• Add unit + integration regression tests (covers samples/MFT entry 5 with a 48-byte $STANDARD_INFORMATION).

Test plan

• cargo fmt --all -- --check
• cargo clippy --all-targets --all-features -- -D warnings
• cargo test -p mft --all-features

---

Note

Ensures $STANDARD_INFORMATION is parsed according to its resident value length, avoiding reads into subsequent attribute headers.

• Read exact resident.data_size for StandardInformation and parse via new StandardInfoAttr::from_slice (supports 48- and 72-byte layouts)
• Return zeroed extended fields when value is 48 bytes; reject intermediate sizes as invalid
• Add unit tests for 48/72-byte parsing and an integration test against samples/MFT entry 5 to prevent leakage into next attribute header

^{Written by Cursor Bugbot for commit 8b1668a. This will update automatically on new commits. Configure here.}