Decode into the remaining buffer after incomplete_utf8 bytes in console read

[Joel-Wwalker]

When reading from a Windows console, Stdin::read first copies any leftover bytes of a partially read code point (incomplete_utf8) into the start of buf. The large-buffer path then decodes into buf starting at position 0, overwriting those bytes, and returns bytes_copied + value, so the clobbered bytes still count toward the length. The result is corrupted output with a stray byte at the end. The small-buffer path already decodes into &mut buf[bytes_copied..]; this change makes the large path do the same. @Ramla-I diagnosed the root cause in #142847 (comment).

I asked in the issue last week whether a reader-side fix was welcome alongside #142872 and figured I'd open it since it was sitting ready. Verified with a pseudoconsole feeding real console input through read_to_end (the report's repro) across 13 input alignments: 7 corrupted before the change, including the exact string from the issue (ж loses its second byte and a stray 0 gets appended), all 13 clean after. Console input can't be driven from CI, so there is no test.

#142872 keeps read_to_end from handing the reader a too-small buffer and is complementary: even with it, this path stays reachable through other callers, e.g. read_vectored with a 1-byte first buffer.

Fixes #142847

[rustbot]

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: @ChrisDenton, libs
• @ChrisDenton, libs expanded to 13 candidates
• Random selection from 7 candidates

[Mark-Simulacrum]

r? @the8472 / @ChrisDenton