Compile against Rust Stable

[kkawakam]

Currently we take advantage of some non-stabilized rust APIs but eventually we will want to be able to compile against Rust stable. The below links are tracking issues that need to be closed out before we can compile against rust stable.

Update June 6 2016: #38 introduced workaround changes that allowed rustyline to be compiled against rust stable (1.9.0). These changes should eventually replaced by Read::chars and encode_utf8 after they become stabilized in the standard library.

TO-DO

• Read::chars
  • rust-lang/rust Tracking Issue
  • replace workaround introduced in Replace use of unstable features #38 (remove char_iter.rs)
• encode_utf8
  • rust-lang/rust Tracking Issue
  • replace workaround introduced in Replace use of unstable features #38 (remove encode_unicode crate)
• str_char
  • rust-lang/rust Tracking Issue
  • stabilized in Rust-1.9
• path_relative_from
  • rust-lang/rust Tracking Issue
  • stabilized in Rust 1.7
  • change from path_relative_from to strip_prefix was addressed in Fix calculated width of prompt with ansi formatting #23
• vec::Drain
  • rust-lang/rust Tracking Issue
  • stabilized in Rust 1.6

[pwrdwnsys]

Is this now complete with the changes merged from #38 ?

[kkawakam]

@pastoraleman I'm going to leave this open because we introduced some work-arounds in #38 to compile against 1.9.0 stable. When Read::chars and encode_utf8 become stabilized, they should replace the work-arounds introduced in #38.

I'll update the issue to reflect that change

[Yamakaky]

Did you consider using https://crates.io/crates/unicode-segmentation?

[gwenn]

@Yamakaky Correct me if I'm wrong, but it seems to me that the only places where this crate can be used is in Completer and "word" actions where the boundaries are specific (not conforming to unicode standard). And we still need to iterate on chars when reading the standard input stream...

[Yamakaky]

Oh, ok.
Opened unicode-rs/unicode-segmentation#7

[jonas-schievink]

Looks like this can now be closed! :)

[gwenn]

Negative !
Read::chars is still unstable.
But, encode_unicode can be removed (encode_utf8 is stabilised).

[jonas-schievink]

Oh, I see now. It compiles against stable, but there's a few workarounds/reimplementations for features that aren't yet stable.

[gwenn]

Read::chars has not been stabilized but removed !

[gwenn]

2.0 released