Add has_data_left() to BufRead

[YuhanLiin]

This is a continuation of #40747 and also addresses #40745. The problem with the previous PR was that it had "eof" in its method name. This PR uses a more descriptive method name, but I'm open to changing it.

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @m-ou-se (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

Please see the contribution instructions for more information.

[the8472]

has_remaining_data seems less... informal?

Is there a reason to check the negative case, i.e. it being not empty as opposed to the is_empty that can be found on many other things?

Also, what's the benefit of this method if it requires error handling? If you have to do error handling anyway you might as well do, that way you only need to do error handling once instead of doing it in the loop condition and when reading the line.

let mut line = String::new();
loop {
   match std.read_line(&mut line) {
      Ok(0) => break, // eof
      Ok(_) => {
          // do something with line
      },
      Err(e) => panic!("read error {}", e);
   }
   line.clear()
}

[YuhanLiin]

I just couldn't think of a good name for the positive case, but I'm open for suggestions.

As for the benefits, the main use case I have for this is for things like deserializing object directly from a file in a loop. The deserialization function will return an error if it encounters an EOF, and depending on the Serde library it's not always easy to tell whether the error is caused by EOF. It's easier to check for EOF separately in the loop condition.

while file.has_data_left()? {
    let obj = rmp_serde::decode::from_read(&mut file)?;
    // rest of the loop
}

As shown above, the additional error handling isn't really a big deal when using question mark syntax (or just unwrapping).

[the8472]

Does that really solve the stated problem though? serde could still encounter an EOF in the middle of parsing an input.

[YuhanLiin]

If serde encounters an EOF in the middle of deserializing an object then it's a syntax error and the input is invalid, at least for that object. If serde finishes parsing the last object in a valid input then all that's left in the reader would be EOF. In that case having EOF checking would prevent serde from attempting to deserialize another object and throwing a syntax error. Having EOF checking between deserialize calls allows the loop to end gracefully when encountering EOF in a valid input.

[m-ou-se]

Thanks for your PR!

This looks reasonable to me. We can think a bit more about the name (is_finished, reached_end, has_remaining_data, ...?), but we can add it now as unstable as is, and have that discussion before we stabilize it.

[m-ou-se]

Can you open a new tracking issue for this?

[YuhanLiin]

Tracking issue here (#86423)

[m-ou-se]

@bors r+

[bors]

📌 Commit 99939c4 has been approved by m-ou-se

[bors]

⌛ Testing commit 99939c4 with merge ce1d561...

[bors]

☀️ Test successful - checks-actions
Approved by: m-ou-se
Pushing ce1d561 to master...