fix: validate Baggage key by W3C standards

[gruebel]

Relates to #2717

Changes

• Also did some small performance improvements, which makes it faster while still adding the extra burden of kay validation

Merge requirement checklist

• CONTRIBUTING guidelines followed
• Unit tests added/updated (if applicable)
• Appropriate CHANGELOG.md files updated for non-trivial, user-facing changes
• Changes in public API reviewed (if applicable)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 80.5%. Comparing base (31b494b) to head (bbffd71).
Report is 1 commits behind head on main.

Additional details and impacted files

@@          Coverage Diff          @@
##            main   #2804   +/-   ##
=====================================
  Coverage   80.4%   80.5%           
=====================================
  Files        124     124           
  Lines      23390   23411   +21     
=====================================
+ Hits       18828   18849   +21     
  Misses      4562    4562           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[lalitb]

Each character in the key is sequentially validated against the INVALID_ASCII_KEY_CHARS array. Using a bitmask for INVALID_ASCII_KEY_CHARS can optimize this check by enabling constant-time lookups instead of an iterative scan.

[scottgerring]

ooh nice idea

[cijothomas]

@fraillt Please see if you have some perf suggestions.

[fraillt]

I'm not sure how to do it using bitmasks (as these characters are quite scattered and there's no easy binary pattern to match/mask), but usually when there's not so many values to check, it's very efficient to do it with switch/match, like this:

fn is_key_valid2(key: &[u8]) -> bool {
    !key.is_empty()
        && key.iter().all(|b| match b {
            b'!' => true,
            b'"' => false,
            b'#' => true,
            b'$' => true,
            b'%' => true,
            b'&' => true,
            b'\'' => true,
            b'(' => false,
            b')' => false,
            b'*' => true,
            b'+' => true,
            b',' => false,
            b'-' => true,
            b'.' => true,
            b'/' => false,
            b'0' => true,
            b'1' => true,
            b'2' => true,
            b'3' => true,
            b'4' => true,
            b'5' => true,
            b'6' => true,
            b'7' => true,
            b'8' => true,
            b'9' => true,
            b':' => false,
            b';' => false,
            b'<' => false,
            b'=' => false,
            b'>' => false,
            b'?' => false,
            b'@' => false,
            b'A' => true,
            b'B' => true,
            b'C' => true,
            b'D' => true,
            b'E' => true,
            b'F' => true,
            b'G' => true,
            b'H' => true,
            b'I' => true,
            b'J' => true,
            b'K' => true,
            b'L' => true,
            b'M' => true,
            b'N' => true,
            b'O' => true,
            b'P' => true,
            b'Q' => true,
            b'R' => true,
            b'S' => true,
            b'T' => true,
            b'U' => true,
            b'V' => true,
            b'W' => true,
            b'X' => true,
            b'Y' => true,
            b'Z' => true,
            b'[' => false,
            b'\\' => false,
            b']' => false,
            b'^' => true,
            b'_' => true,
            b'`' => true,
            b'a' => true,
            b'b' => true,
            b'c' => true,
            b'd' => true,
            b'e' => true,
            b'f' => true,
            b'g' => true,
            b'h' => true,
            b'i' => true,
            b'j' => true,
            b'k' => true,
            b'l' => true,
            b'm' => true,
            b'n' => true,
            b'o' => true,
            b'p' => true,
            b'q' => true,
            b'r' => true,
            b's' => true,
            b't' => true,
            b'u' => true,
            b'v' => true,
            b'w' => true,
            b'x' => true,
            b'y' => true,
            b'z' => true,
            b'{' => false,
            b'|' => true,
            b'}' => false,
            b'~' => true,
            _ => false,
        })
}

This is equivalent to b.is_ascii_graphic() && !INVALID_ASCII_KEY_CHARS.contains(b), notice that i also ignore control characters (is_ascii_graphic() not is_ascii).

I have tested this is_key_valid2 function, locally on my machine, and I get +3x performance improvement.

[cijothomas]

Thanks. Would be nice to incorporate the perf optimizations in this or in a follow up.

[gruebel]

Thanks. Would be nice to incorporate the perf optimizations in this or in a follow up.

@cijothomas I created a new ticket for it #2835. I tested the proposed idea/solution, but it had locally no measurable effect for me 😢