fix: buffer logic in log-fetcher

[NikolaMilosa]

Deserializing incoming chunks as utf8_lossy breaks the encoding of journald protocol because there are non-utf8 bytes for the length of the message.

[Copilot]

Pull Request Overview

This PR fixes the buffering logic in the log-fetcher by switching from a String to a Vec for accumulating chunks, ensuring that non‑UTF8 bytes used in the journald protocol are preserved. Key changes include:

• Replacing the String buffer with a Vec and updating related operations.
• Adjusting the logic to find journal entry boundaries using byte slices.
• Updating error logging to handle entries missing the __CURSOR field and adding tests for binary journal fields.

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
rs/log-fetcher/src/main.rs	Updated buffering logic to use Vec and modified error handling for missing __CURSOR.
rs/log-fetcher/src/journald_parser.rs	Added a new test ensuring that binary journal fields are correctly processed.

[Copilot]

Using unwrap here may cause a panic if leftover_buffer contains non-UTF8 bytes. Consider using std::str::from_utf8_lossy to safely display the data.

Suggested change

	std::str::from_utf8(&leftover_buffer).unwrap(),
	std::str::from_utf8_lossy(&leftover_buffer),

[Copilot]

Using unwrap here may cause a panic if chunk contains non-UTF8 bytes. Consider using std::str::from_utf8_lossy to safely convert the bytes for logging.

Suggested change

	std::str::from_utf8(&chunk).unwrap()
	std::str::from_utf8_lossy(&chunk)