mikros/rust
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
12,118 Commits 2 Branches 0 Tags
Commit Graph

364 Commits

Author SHA1 Message Date
Patrick Walton
afd9a75c9e rustc: Fix cross-crate max/min-class-style constructors 2012-07-26 15:30:15 -07:00
Eric Holk
5805616aad Make parsing about 0.3 seconds faster. 2012-07-26 11:37:57 -07:00
Eric Holk
0a8b794a3a Add #[inline(never)], and also fixed inlining on vec::push 2012-07-26 11:37:57 -07:00
Eric Holk
fec749df16 Comments Only: Remove outdated FIXMEs. Fixes #2886. 2012-07-26 10:20:05 -07:00
Paul Stansifer
f4c093c4af Change #macro to macro_rules! in some cases. 2012-07-26 10:14:01 -07:00
Paul Stansifer
7f5fbd4f9d Allow old-style syntax extensions to be called with new syntax. 2012-07-26 09:52:21 -07:00
Eric Holk
ab1defb1ce Make parsing about 0.3 seconds faster. 2012-07-26 09:47:32 -07:00
Eric Holk
3aee39a6ec Add #[inline(never)], and also fixed inlining on vec::push 2012-07-25 17:30:13 -07:00
Eric Holk
6748f78cb1 Polymorphic protocols work well enough to do MapReduce. 
I did some horrible things with type variable naming here. It should do the right thing in most cases, but we'll need to go through and make it correct someday.
 2012-07-25 12:12:26 -07:00
Eric Holk
14adb98b64 Bounded protocols work well enough to compile core, but map reduce has too many type parameters, so we have to get fancier. 2012-07-25 12:12:25 -07:00
Eric Holk
1dde5e7fc2 Thread spans through the pipe compiler. They aren't perfect, but they make debugging far easier than core.rc:0:0. 
Changed the is_bounded check, so we fail compiling core right now due to not supporting type parameters.
 2012-07-25 12:12:25 -07:00
Eric Holk
7f5f1f90a0 Compiled a bounded version of pingpong. 
There are some failures in the other pipe tests, but these seem to just be a matter of generalizing the library code.

Updating pipes library so all tests pass again
 2012-07-25 12:12:25 -07:00
Eric Holk
b97fe98354 Generate buffer type for bounded protocols 2012-07-25 12:12:25 -07:00
Eric Holk
c28af26258 Refactor the bounded pingpong example to avoid needing to generate unsafe code. 
Took some steps towards bounded codegen.
 2012-07-25 12:12:25 -07:00
Eric Holk
d74fb9875b Refactoring pipes to allow implementing bounded protocols. 2012-07-25 12:12:25 -07:00
Patrick Walton
728d16cfca rustc: Parse new-style impl declarations 2012-07-24 16:38:24 -07:00
Patrick Walton
587b0edbbf rustc: Don't require that structs have constructors 2012-07-24 15:29:51 -07:00
Paul Stansifer
9103e43909 Bugfix: enable transcription to deal with zero-repetition cases. 2012-07-24 11:44:58 -07:00
Paul Stansifer
e20c5e7596 Improve an error message a little. 2012-07-24 11:44:58 -07:00
Paul Stansifer
f785cccf0f Bugfix: make the parser handle the case where zero repetitions occur, by handling parse results on the basis of what names the matcher expects to bind, not on what names are actually bound. 2012-07-24 11:44:58 -07:00
Patrick Walton
0d581bdafe syntax: Expect the closing brace after struct literals 2012-07-23 18:58:56 -07:00
Patrick Walton
df4db83ed8 rustc: Max/min classes: Add struct literal syntax 2012-07-23 16:40:05 -07:00
Patrick Walton
a57087e032 syntax: Remove reference to typestate in a comment 2012-07-23 13:24:58 -07:00
Patrick Walton
978ca03cb2 Revert "accept naked exprs with commas in pattern arms" due to pretty-printing failures 
This reverts commit f712b2d76b1077a2241916cc3269aa1d83ce3088.

In alt arms, the parser needs to do a little lookahead to determine
whether it's looking at a record literal or a block.

Also there are some indentation issues in the expected source.
 2012-07-19 07:53:55 -07:00
Niko Matsakis
f712b2d76b accept naked exprs with commas in pattern arms 
pretty printing will use them, but indentation is slightly off
if the expr is long
 2012-07-18 20:16:41 -07:00
Tim Chevalier
de5d5e6eeb Remove non-existent imports 2012-07-18 18:02:07 -07:00
Patrick Walton
3ac5b4a86f syntax: Parse multiple trait refs in a single implementation 2012-07-18 16:05:17 -07:00
Niko Matsakis
e0ea67a2a6 prevent regions from escaping in ifaces; remove &r.T syntax 2012-07-18 11:48:58 -07:00
Eric Holk
7b8171ef2d Added liveness analysis for protocols, and removed warnings about empty states. 2012-07-17 17:46:31 -07:00
Eric Holk
ee855caa5d Don't ICE when protocol steps to invalid mesasge. 2012-07-17 17:46:31 -07:00
Patrick Walton
b71a8827e3 rustc: Fix coherence errors in the build 2012-07-17 16:49:54 -07:00
Patrick Walton
db020ab63c rustc: Implement and enforce instance coherence 2012-07-17 15:46:43 -07:00
Tim Chevalier
b5729bd600 Support attributes on class ctors and dtors 
Closes #2660
 2012-07-17 12:40:59 -07:00
Eric Holk
c8739cb0bc Error checking for protocols. We'll need spans though. 2012-07-17 10:35:59 -07:00
Niko Matsakis
0e42004bab introduce an owned kind for data that contains no borrowed ptrs 2012-07-16 20:18:18 -07:00
Eric Holk
d39f6fc676 Add support for terminal states (issue #2862) 2012-07-16 14:50:47 -07:00
Niko Matsakis
41a21f053c remove typestate from code, tests, and docs 2012-07-14 17:37:20 -07:00
Michael Sullivan
7b2f4755f3 Get rid of ast::ty_vstore, which was only used for fixed length. 2012-07-14 12:45:52 -07:00
Michael Sullivan
e2af785606 Make the new world order normative. Closes #2908. 2012-07-14 01:03:44 -07:00
Michael Sullivan
92743dc2a6 Move the world over to using the new style string literals and types. Closes #2907. 2012-07-14 01:03:43 -07:00
Michael Sullivan
6247a529e3 Handle prefix notations for strings in patterns. This is kind of gross. 2012-07-13 19:08:18 -07:00
Michael Sullivan
628d3e9d38 Change (hopefully) all of the code that generates strs asts to produce ~strs. 2012-07-13 17:03:54 -07:00
Michael Sullivan
f5e69d611e Change the pretty printer to print vstores for strs in prefix notation. 2012-07-13 17:03:54 -07:00
Michael Sullivan
985b52be6d Support prefix notation for vstore strings. Closes #2906. 2012-07-13 17:03:49 -07:00
Lindsey Kuper
78d11b899b Clean up various bugs with trait parsing. 2012-07-13 15:31:39 -07:00
Lindsey Kuper
5a63b2100e More consistent use of backticks and "expected" in error messages. 
Got some of the debug messages, here, too.  I figure it doesn't hurt
to get used to doing this even in places where users won't ever see
it.
 2012-07-13 15:31:39 -07:00
Tim Chevalier
07a81ad12e Refactor how impl self types are stored 
In order to avoid a confusing use of the tcache, I added an extra
node ID field to trait refs. Now trait refs have a "ref ID" (the one
that resolve3 resolves) and an "impl ID" (the one that you look up
in the tcache to get the self type).

Closes #2434
 2012-07-13 14:47:04 -07:00
Lindsey Kuper
fc9c4c3245 Front-end support for default impls in traits. 2012-07-13 11:16:07 -07:00
Niko Matsakis
90e435e808 change region syntax to &r/T in place of &r.T 2012-07-13 10:20:50 -07:00
Tim Chevalier
78ec6fe30c Obliterate the callee_id hack 
Exprs that could be applications of overloaded operators
(expr_unary, expr_binary, expr_index) relied on the previous node ID
being "reserved" to carry extra typechecking info. This was
incredibly error-prone. Fixed it; now all exprs have two node IDs
(which will be wasted in some cases; future work could make this
an option instead if the extra int field ends up being a performance
problem).

Closes #2804
 2012-07-12 19:02:07 -07:00