Commit Graph

28 Commits

Author SHA1 Message Date
Huon Wilson
d9c54f8387 librustc: use LLVM intrinsics for several floating point operations.
Achieves at least 5x speed up for some functions!

Also, reorganise the delegation code so that the delegated function wrappers
have the #[inline(always)] annotation, and reduce the repetition of
delegate!(..).
2013-04-21 01:40:48 +10:00
Brendan Zabarauskas
939a97f5cb Add #[inline(always)] to each operator method 2013-04-18 23:24:24 +10:00
Brendan Zabarauskas
c4685477e0 Consolidate tests of numeric operations 2013-04-14 02:53:00 +10:00
Brendan Zabarauskas
353ce872e2 Generate NumCast impls and tests using macros 2013-04-14 01:14:02 +10:00
Huon Wilson
41c6f67109 libcore: from_str_common: provide option to ignore underscores.
Implement the possible improvement listed in the comment on
from_str_bytes_common.
2013-04-08 16:35:39 +10:00
Patrick Walton
1e91595520 librustc: Remove fail_unless! 2013-03-29 16:39:08 -07:00
Patrick Walton
85c9fc6f8f librustc: Remove the const declaration form everywhere 2013-03-22 22:24:35 -07:00
Patrick Walton
be9bddd463 libcore: Remove pure from libcore. rs=depure 2013-03-22 10:29:17 -07:00
Patrick Walton
4634f7edae librustc: Remove all uses of static from functions. rs=destatic 2013-03-22 10:27:39 -07:00
Patrick Walton
d7e74b5e91 librustc: Convert all uses of assert over to fail_unless! 2013-03-07 22:37:57 -08:00
Alex Crichton
cb4ab76e4a Adding missing imports for tests, and gate off others 2013-03-04 12:27:01 -05:00
Brian Anderson
bcf626812b Rename core::private to core::unstable. #4743 2013-03-01 14:55:47 -08:00
Patrick Walton
07c3f5c0de librustc: Forbid pub or priv before trait implementations 2013-02-27 09:40:16 -08:00
Matthijs 'Thiez' Hofstra
675bdcf341 Created libcore/private/intrinsics.rs, which aims to contain every
rustc intrinsic. Several files in libcore have been changed to use
these intrinsics.
As of yet, none of the intrinsics are documented.
2013-02-20 20:39:49 +01:00
bors
3e97cce27d auto merge of #4957 : Kimundi/rust/incoming, r=catamorphism
Moved them into own module and made them not depend on an Round trait impl for integers and generic math functions that can fail on integers any more.
2013-02-15 17:29:57 -08:00
Patrick Walton
bb833ca0f0 librustc: Stop parsing impl Type : Trait and fix several declarations that slipped through. r=tjc 2013-02-15 16:59:56 -08:00
Marvin Löbel
bd93a36d73 Made num <-> str conversion functions use NumStrConv trait
Removed hacky dependency on Round trait and generic infinity functions
Removed generic-runtime-failure-depending-on-type behavior
2013-02-15 05:20:36 +01:00
Marvin Löbel
adac6cb5c6 Moved numeric string conversion functions into own module 2013-02-15 03:29:36 +01:00
Patrick Walton
9143688197 librustc: Replace impl Type : Trait with impl Trait for Type. rs=implflipping 2013-02-14 14:44:12 -08:00
Patrick Walton
216e85fadf libcore: Move the numeric operations out of Num. r=brson
Sadly I could not use trait inheritance due to a type parameter substitution
bug.
2013-02-14 08:14:01 -08:00
Nick Desaulniers
4445b38df2 Remove die!, raplace invocations with fail! Issue #4524 pt 3 2013-02-13 17:01:32 -08:00
Brendan Zabarauskas
48b2141b83 Add NumCast trait for generic numeric type casts 2013-02-11 12:33:05 +11:00
Niko Matsakis
a32498d846 Make ~fn non-copyable, make &fn copyable, split barefn/closure types,
correct handling of moves for struct-record update.

Part of #3678.  Fixes #2828, #3904, #4719.
2013-02-07 05:53:30 -08:00
Erick Tryzelaar
9adfa59d8e core: convert ToStr::to_str to take explicit &self 2013-02-03 20:47:26 -08:00
Marvin Löbel
eeb89c5012 Solved float, f32 and f64 to_str_radix() special value ambiguity.
Calling it on a special value now causes a failure, however `to_str_radix_special()` is provided which can be
used if those values are expected, and which returns a tupel to allow differentating them.
2013-02-03 15:37:25 -08:00
Marvin Löbel
a612e49a21 Converted the floating point types to the new string conversion functions.
Also fixed all conflicting calls of the old functions in the rest of the codebase.

The set of string conversion functions for each float type now consists of those items:
- to_str(), converts to number in base 10
- to_str_hex(), converts to number in base 16
- to_str_radix(), converts to number in given radix
- to_str_exact(), converts to number in base 10 with a exact number of trailing digits
- to_str_digits(), converts to number in base 10 with a maximum number of trailing digits
- implementations for to_str::ToStr and num::ToStrRadix
- from_str(), parses a string as number in base 10 including decimal exponent and special values
- from_str_hex(), parses a string as a number in base 16 including binary exponent and special values
- from_str_radix(), parses a string as a number in a given base excluding any exponent and special values
- implementations for from_str::FromStr and num::FromStrRadix
2013-02-03 15:37:24 -08:00
Marvin Löbel
96f0512a45 Added Round trait to core 2013-02-03 15:37:23 -08:00
Marvin Löbel
40f0b45f8e Moved all numeric modules in core into own directory
Reason: Better grouping of related modules, future-proving for a more extensive math library.
2013-02-03 15:37:23 -08:00