Commit Graph

866 Commits

Author SHA1 Message Date
Alex Crichton
c765a8e7ad Fixing remaining warnings and errors throughout 2014-02-03 10:39:23 -08:00
Brendan Zabarauskas
729060dbb9 Remove Times trait
`Times::times` was always a second-class loop because it did not support the `break` and `continue` operations. Its playful appeal was then lost after `do` was disabled for closures. It's time to let this one go.
2014-01-30 14:52:25 +11:00
bors
e1580f6d09 auto merge of #11868 : bytbox/rust/remove-do, r=alexcrichton
Fixes #10815.
2014-01-29 08:21:38 -08:00
Scott Lawrence
3dbc1c34e6 Remove do keyword from test/ 2014-01-29 09:15:42 -05:00
David Manescu
4d0d3da9e4 extra: move arena to libarena
In line with the dissolution of libextra - #8784 - moves arena to its own library libarena.
Changes based on PR #11787. Updates .gitignore to ignore doc/arena.
2014-01-29 13:54:38 +11:00
bors
d3f70f5a7d auto merge of #11817 : salemtalha/rust/master, r=brson
Fixes Issue #11815
2014-01-26 15:26:30 -08:00
Salem Talha
cc61fc0994 Removed all instances of XXX in preparation for relaxing of FIXME rule 2014-01-26 14:42:53 -05:00
Alex Crichton
4d6836f418 Fix privacy fallout from previous change 2014-01-26 11:03:13 -08:00
Corey Richardson
7499e2dd45 Add benchmarks 2014-01-24 22:30:00 -08:00
Corey Richardson
69afce64c7 Update task-perf-one-million 2014-01-24 22:30:00 -08:00
Sean Chalmers
292ed3e55c Update flip() to be rev().
Consensus leaned in favour of using rev instead of flip.
2014-01-23 22:18:18 +01:00
Sean Chalmers
55d6e0e1b7 Rename Invert to Flip - Issue 10632
Renamed the invert() function in iter.rs to flip().

Also renamed the Invert<T> type to Flip<T>.

Some related code comments changed. Documentation that I could find has
been updated, and all the instances I could locate where the
function/type were called have been updated as well.
2014-01-23 21:50:18 +01:00
Simon Sapin
05ae134ace [std::str] Rename from_utf8_owned_opt() to from_utf8_owned(), drop the old from_utf8_owned() behavior 2014-01-21 15:48:48 -08:00
Simon Sapin
b5e65731c0 [std::vec] Rename .shift_opt() to .shift(), drop the old .shift() behavior 2014-01-21 15:48:47 -08:00
Simon Sapin
bada25e425 [std::vec] Rename .pop_opt() to .pop(), drop the old .pop() behavior 2014-01-21 15:48:47 -08:00
Simon Sapin
aa66b91767 [std::vec] Rename .last_opt() to .last(), drop the old .last() behavior 2014-01-21 15:48:46 -08:00
Huon Wilson
39713b8295 Remove unnecessary parentheses. 2014-01-21 22:00:18 +11:00
bors
bf07c80838 auto merge of #11629 : brson/rust/whattayaknowitsmoreandroidfixes, r=cmr 2014-01-18 07:36:43 -08:00
Brian Anderson
4a3a61bd83 xfail shootout-reverse-complement on android 2014-01-17 22:34:58 -08:00
bors
c58d2bacb7 auto merge of #11503 : FlaPer87/rust/master, r=huonw
The patch adds the missing pow method for all the implementations of the
Integer trait. This is a small addition that will most likely be
improved by the work happening in #10387.

Fixes #11499
2014-01-17 20:36:47 -08:00
Alex Crichton
295b46fc08 Tweak the interface of std::io
* Reexport io::mem and io::buffered structs directly under io, make mem/buffered
  private modules
* Remove with_mem_writer
* Remove DEFAULT_CAPACITY and use DEFAULT_BUF_SIZE (in io::buffered)
2014-01-17 10:00:47 -08:00
Flavio Percoco
ed7e576d9c Add a generic power function
The patch adds a `pow` function for types implementing `One`, `Mul` and
`Clone` trait.

The patch also renames f32 and f64 pow into powf in order to still have
a way to easily have float powers. It uses llvms intrinsics.

The pow implementation for all num types uses the exponentiation by
square.

Fixes bug #11499
2014-01-17 15:41:26 +01:00
Brendan Zabarauskas
4fc0452ace Remove re-exports of std::io::stdio::{print, println} in the prelude.
The `print!` and `println!` macros are now the preferred method of printing, and so there is no reason to export the `stdio` functions in the prelude. The functions have also been replaced by their macro counterparts in the tutorial and other documentation so that newcomers don't get confused about what they should be using.
2014-01-11 10:46:00 +11:00
Patrick Walton
b7ff9c1a59 libextra: Introduce typed arenas.
A typed arena is a type of arena that can only allocate objects of one
type. It is 3x faster than the existing arena and 13x faster than malloc
on Mac.
2014-01-07 16:44:41 -08:00
Brian Anderson
3b1862a82f Don't allow newtype structs to be dereferenced. #6246 2014-01-04 14:44:12 -08:00
Patrick Walton
c3694d732e test: De-@mut the test suite 2014-01-03 14:02:01 -08:00
Alex Crichton
6cad8f4f14 Test fixes and rebase conflicts
* vec::raw::to_ptr is gone
* Pausible => Pausable
* Removing @
* Calling the main task "<main>"
* Removing unused imports
* Removing unused mut
* Bringing some libextra tests up to date
* Allowing compiletest to work at stage0
* Fixing the bootstrap-from-c rmake tests
* assert => rtassert in a few cases
* printing to stderr instead of stdout in fail!()
2013-12-25 23:10:46 -08:00
Huon Wilson
2e8c522c62 std::vec: make the sorting closure use Ordering rather than just being
(implicitly) less_eq.
2013-12-22 18:16:50 +11:00
Huon Wilson
1b1e4caa79 std::vec: add a sugary .sort() method for plain Ord sorting.
This moves the custom sorting to `.sort_by`.
2013-12-21 09:35:18 +11:00
Huon Wilson
48fedcb36f extra: remove sort in favour of the std method.
Fixes #9676.
2013-12-20 12:38:46 +11:00
Alex Crichton
39a6c9d637 Test fallout from std::comm rewrite 2013-12-16 22:55:49 -08:00
Patrick Walton
a87786e3e9 librustc: Remove identifiers named box, since it's about to become a keyword. 2013-12-15 10:41:15 -08:00
Huon Wilson
4f62c969f6 std::vec: move pointless raw::get and unsafe_get functions.
This can easily be written as `(*v.unsafe_ref(i)).clone()`, or just
`*v.unsafe_ref(i)` for primitive types like `i32` (the common case).
2013-12-15 23:05:00 +11:00
Huon Wilson
f97040a93b std::vec: remove unnecessary count parameter on {bytes,
raw}::copy_memory.

Slices carry their length with them, so we can just use that
information.
2013-12-15 22:23:11 +11:00
bors
b2750bca0c auto merge of #10933 : TeXitoi/rust/shootout-fasta-rewrite, r=alexcrichton
improvements:
 - no managed box
 - no virtual calls
 - no useless copy
 - optimizations (bisect is slower, limit tests, BufferedWriter...)
 - pass shootout test
 - 10 times faster
2013-12-14 05:26:28 -08:00
Guillaume Pinot
64ca0ba6e9 rewrite of shootout-fasta.rs
improvements:
 - no managed box
 - no virtual calls
 - no useless copy
 - optimizations (bisect is slower, limit tests, BufferedWriter...)
 - pass shootout test
 - should be as fast as the best official test

Thanks to @cmr and @eddyb for their help!
2013-12-12 16:06:18 +01:00
Erik Price
5731ca3078 Make 'self lifetime illegal.
Also remove all instances of 'self within the codebase.

This fixes #10889.
2013-12-11 10:54:06 -08:00
klutzy
5a93d12e01 std::io: Add Buffer.lines(), change .bytes() api
-   `Buffer.lines()` returns `LineIterator` which yields line using
    `.read_line()`.
-   `Reader.bytes()` now takes `&mut self` instead of `self`.
-   `Reader.read_until()` swallows `EndOfFile`. This also affects
    `.read_line()`.
2013-12-11 15:50:51 +09:00
Patrick Walton
fd7a513bef libstd: Remove Cell from the library. 2013-12-10 17:55:09 -08:00
Patrick Walton
ec5603bf13 librustpkg: Make io::ignore_io_error() use RAII; remove a few more
cells.
2013-12-10 15:13:12 -08:00
Patrick Walton
786dea207d libextra: Another round of de-Cell-ing.
34 uses of `Cell` remain.
2013-12-10 15:13:12 -08:00
Kevin Ballard
408dc5ad1b Revert "libstd: Change Path::new to Path::init."
This reverts commit c54427ddfb.

Leave the #[ignores] in that were added to rustpkg tests.

Conflicts:
	src/librustc/driver/driver.rs
	src/librustc/metadata/creader.rs
2013-12-04 22:33:53 -08:00
bors
bd5305fbc8 auto merge of #10799 : TeXitoi/rust/shootout-reverse-complement-resurected, r=alexcrichton
This version is inspired by the best version in C by Mr Ledrug,
but without the parallelisation.
2013-12-04 20:46:23 -08:00
Guillaume Pinot
e76e83cb10 rewrite of shootout-reverse-complement.rs
This version is inspired by the best version in C by Mr Ledrug,
but without the parallelisation.
2013-12-04 18:15:53 +01:00
Huon Wilson
9d64e46013 std::str: remove from_utf8.
This function had type &[u8] -> ~str, i.e. it allocates a string
internally, even though the non-allocating version that take &[u8] ->
&str and ~[u8] -> ~str are all that is necessary in most circumstances.
2013-12-04 22:35:53 +11:00
Patrick Walton
c54427ddfb libstd: Change Path::new to Path::init. 2013-11-29 10:55:13 -08:00
Guillaume Pinot
6bd22494e9 shootout-spectralnorm resurection with parallelization 2013-11-28 20:46:59 +01:00
Patrick Walton
9521551b47 librustc: Fix merge fallout. 2013-11-26 11:04:39 -08:00
Patrick Walton
749ee53c6d librustc: Make || lambdas not infer to procs 2013-11-26 08:25:27 -08:00
Patrick Walton
38efa17bb8 test: Remove all remaining non-procedure uses of do. 2013-11-26 08:25:27 -08:00
Patrick Walton
9e610573ba librustc: Remove remaining uses of &fn() in favor of ||. 2013-11-26 08:20:58 -08:00
Patrick Walton
406813957b test: Remove most uses of &fn() from the tests. 2013-11-26 08:19:00 -08:00
Marvin Löbel
24b316a3b9 Removed unneccessary _iter suffixes from various APIs 2013-11-26 10:02:26 +01:00
bors
2cc1e16ac0 auto merge of #10603 : alexcrichton/rust/no-linked-failure, r=brson
The reasons for doing this are:

* The model on which linked failure is based is inherently complex
* The implementation is also very complex, and there are few remaining who
  fully understand the implementation
* There are existing race conditions in the core context switching function of
  the scheduler, and possibly others.
* It's unclear whether this model of linked failure maps well to a 1:1 threading
  model

Linked failure is often a desired aspect of tasks, but we would like to take a
much more conservative approach in re-implementing linked failure if at all.

Closes #8674
Closes #8318
Closes #8863
2013-11-24 21:32:13 -08:00
Alex Crichton
acca9e3834 Remove linked failure from the runtime
The reasons for doing this are:

* The model on which linked failure is based is inherently complex
* The implementation is also very complex, and there are few remaining who
  fully understand the implementation
* There are existing race conditions in the core context switching function of
  the scheduler, and possibly others.
* It's unclear whether this model of linked failure maps well to a 1:1 threading
  model

Linked failure is often a desired aspect of tasks, but we would like to take a
much more conservative approach in re-implementing linked failure if at all.

Closes #8674
Closes #8318
Closes #8863
2013-11-24 21:21:12 -08:00
Steven Fackler
5759f2fc57 Strip down Cell functionality 2013-11-22 21:19:53 -08:00
Patrick Walton
ba739b2135 librustc: Convert ~fn() to proc() everywhere. 2013-11-18 18:27:31 -08:00
Guillaume Pinot
ec27c09bdf pidigits resurected
Changes:
* default value when no args
* license
* removed libc printing
* use extra::bigint instead of handmade gmp binding
2013-11-18 22:23:38 +01:00
Guillaume Pinot
93bb99ea09 Prettier long string breaking. 2013-11-17 13:09:56 +01:00
Guillaume Pinot
0c3b04fa0f Clarrify the message for test mode and use u8 instead of i8 for storing bits 2013-11-17 12:47:47 +01:00
Guillaume Pinot
6258857beb resurect shootout-mandelbrot.rs
Changes:
* add licence;
* remove usage of libc and unsafe;
* use BufferedWriter to improve performance;
* use a DummyWriter to cancel binary output in test.
2013-11-17 12:04:36 +01:00
bors
90754ae9c9 auto merge of #10478 : TeXitoi/rust/shootout-meteor, r=brson
This implementation of the meteor contest implements:
 - insertion check with bit trick;
 - pregenetation of every feasible placement of the pieces on the
   board;
 - filtering of placement that implies unfeasible board
 - central symetry breaking

related to #2776
2013-11-14 22:01:26 -08:00
Guillaume Pinot
d2bcc7b621 Improve variable naming. 2013-11-14 23:22:44 +01:00
Guillaume Pinot
74d27311a7 remove useless Iterate impl, and permute the order of the argument of iterate
Writing iterate(x, f) is more logical as rust iterator pipeline is left to right.
2013-11-14 23:05:33 +01:00
Guillaume Pinot
89a9ce0cc6 Add shootout meteor contest benchmark.
This implementation of the meteor contest implements:
 - insertion check with bit trick;
 - pregenetation of every feasible placement of the pieces on the
   board;
 - filtering of placement that implies unfeasible board
 - central symetry breaking
2013-11-14 09:11:33 +01:00
Alex Crichton
9bcf557589 Implement native::IoFactory
This commit re-organizes the io::native module slightly in order to have a
working implementation of rtio::IoFactory which uses native implementations. The
goal is to seamlessly multiplex among libuv/native implementations wherever
necessary.

Right now most of the native I/O is unimplemented, but we have existing bindings
for file descriptors and processes which have been hooked up. What this means is
that you can now invoke println!() from libstd with no local task, no local
scheduler, and even without libuv.

There's still plenty of work to do on the native I/O factory, but this is the
first steps into making it an official portion of the standard library. I don't
expect anyone to reach into io::native directly, but rather only std::io
primitives will be used. Each std::io interface seamlessly falls back onto the
native I/O implementation if the local scheduler doesn't have a libuv one
(hurray trait ojects!)
2013-11-13 18:34:59 -08:00
Carol W
33ee4335b9 Fixed xfail for nbody shootout benchmark by correcting command line parse.
Cleaned up unneeded imports and type changes to resolve compiler warnings.
2013-11-12 15:14:07 -08:00
Alex Crichton
49ee49296b Move std::rt::io to std::io 2013-11-11 20:44:07 -08:00
Alex Crichton
7755ffd013 Remove #[fixed_stack_segment] and #[rust_stack]
These two attributes are no longer useful now that Rust has decided to leave
segmented stacks behind. It is assumed that the rust task's stack is always
large enough to make an FFI call (due to the stack being very large).

There's always the case of stack overflow, however, to consider. This does not
change the behavior of stack overflow in Rust. This is still normally triggered
by the __morestack function and aborts the whole process.

C stack overflow will continue to corrupt the stack, however (as it did before
this commit as well). The future improvement of a guard page at the end of every
rust stack is still unimplemented and is intended to be the mechanism through
which we attempt to detect C stack overflow.

Closes #8822
Closes #10155
2013-11-11 10:40:34 -08:00
Young-il Choi
9bb1e256d7 temporarily disable tests on android and tagging issue number #10393 2013-11-10 12:02:23 +09:00
Huon Wilson
b95a8c63fd std::ascii: Provide a copyless [Ascii] -> str method.
This renames to_str_ascii to as_str_ascii and makes it non-copying,
which is possible now that strings no longer have a hidden extra
byte/null terminator.

Fixes #6120.
2013-11-08 10:20:06 +11:00
Alex Crichton
f19d083362 Fill out the remaining functionality in io::file
This adds bindings to the remaining functions provided by libuv, all of which
are useful operations on files which need to get exposed somehow.

Some highlights:

* Dropped `FileReader` and `FileWriter` and `FileStream` for one `File` type
* Moved all file-related methods to be static methods under `File`
* All directory related methods are still top-level functions
* Created `io::FilePermission` types (backed by u32) that are what you'd expect
* Created `io::FileType` and refactored `FileStat` to use FileType and
  FilePermission
* Removed the expanding matrix of `FileMode` operations. The mode of reading a
  file will not have the O_CREAT flag, but a write mode will always have the
  O_CREAT flag.

Closes #10130
Closes #10131
Closes #10121
2013-11-03 15:15:42 -08:00
Alex Crichton
9c1851019f Remove all blocking std::os blocking functions
This commit moves all thread-blocking I/O functions from the std::os module.
Their replacements can be found in either std::rt::io::file or in a hidden
"old_os" module inside of native::file. I didn't want to outright delete these
functions because they have a lot of special casing learned over time for each
OS/platform, and I imagine that these will someday get integrated into a
blocking implementation of IoFactory. For now, they're moved to a private module
to prevent bitrot and still have tests to ensure that they work.

I've also expanded the extensions to a few more methods defined on Path, most of
which were previously defined in std::os but now have non-thread-blocking
implementations as part of using the current IoFactory.

The api of io::file is in flux, but I plan on changing it in the next commit as
well.

Closes #10057
2013-11-03 15:15:42 -08:00
bors
c7853c9ad2 auto merge of #10145 : TeXitoi/rust/issues-2913, r=bstrie
The old version didn't follow shootout's directives: no memory were
deallocated.  At the same time, parallelisation of the computation.

fix #2913
2013-10-29 16:26:25 -07:00
Guillaume Pinot
4ca32e1bfc rewrite shootout-binarytrees.rs to match shootout directives
The old version didn't follow shootout's directives: no memory were
deallocated.  At the same time, parallelisation of the computation.

fix #2913
2013-10-29 08:56:16 +01:00
Ziad Hatahet
3797f2bfe6 Capitalize statics in f32 and f64 mods
Fixes #10077
2013-10-28 19:35:56 -07:00
Marvin Löbel
fa8e71a825 Allow fail messages to be caught, and introduce the Any trait
Some code cleanup, sorting of import blocks

Removed std::unstable::UnsafeArc's use of Either

Added run-fail tests for the new FailWithCause impls

Changed future_result and try to return Result<(), ~Any>.

- Internally, there is an enum of possible fail messages passend around.
- In case of linked failure or a string message, the ~Any gets
  lazyly allocated in future_results recv method.
- For that, future result now returns a wrapper around a Port.
- Moved and renamed task::TaskResult into rt::task::UnwindResult
  and made it an internal enum.
- Introduced a replacement typedef `type TaskResult = Result<(), ~Any>`.
2013-10-28 08:50:32 +01:00
bors
950add4e49 auto merge of #10072 : brson/rust/modelines, r=thestinger
These are relics that serve no purpose.
2013-10-26 02:16:08 -07:00
Brian Anderson
9a773da681 Remove ancient emacs mode lines from test cases
These are relics that serve no purpose.
2013-10-25 17:41:54 -07:00
Alex Crichton
620ab3853a Test fixes and merge conflicts 2013-10-24 14:21:58 -07:00
Alex Crichton
61ed2cfb55 Remove even more of std::io
Big fish fried here:

    extra::json
    most of the compiler
    extra::io_util removed
    extra::fileinput removed

Fish left to fry

    extra::ebml
2013-10-24 14:21:57 -07:00
bors
e2428b791c auto merge of #10032 : thestinger/rust/snapshot, r=huonw 2013-10-23 15:11:07 -07:00
Daniel Micay
142672dca4 register snapshots 2013-10-23 18:06:12 -04:00
Huon Wilson
148f737c19 std::rand: add distributions::Range for generating [lo, hi).
This reifies the computations required for uniformity done by
(the old) `Rng.gen_integer_range` (now Rng.gen_range), so that they can
be amortised over many invocations, if it is called in a loop.

Also, it makes it correct, but using a trait + impls for each type,
rather than trying to coerce `Int` + `u64` to do the right thing. This
also makes it more extensible, e.g. big integers could & should
implement SampleRange.
2013-10-23 10:40:06 +11:00
Alex Crichton
daf5f5a4d1 Drop the '2' suffix from logging macros
Who doesn't like a massive renaming?
2013-10-22 08:09:56 -07:00
bors
d052912297 auto merge of #9851 : alexcrichton/rust/include_bin, r=huonw
Previously an ExprLit was created *per byte* causing a huge increase in memory
bloat. This adds a new `lit_binary` to contain a literal of binary data, which
is currently only used by the include_bin! syntax extension. This massively
speeds up compilation times of the shootout-k-nucleotide-pipes test

    before:
        time: 469s
        memory: 6GB
        assertion failure in LLVM (section too large)

    after:
        time: 2.50s
        memory: 124MB

Closes #2598
2013-10-18 09:41:33 -07:00
Alex Crichton
273784e9bf Optimize include_bin! for large inputs
Previously an ExprLit was created *per byte* causing a huge increase in memory
bloat. This adds a new `lit_binary` to contain a literal of binary data, which
is currently only used by the include_bin! syntax extension. This massively
speeds up compilation times of the shootout-k-nucleotide-pipes test

    before:
        time: 469s
        memory: 6GB
        assertion failure in LLVM (section too large)

    after:
        time: 2.50s
        memory: 124MB

Closes #2598
2013-10-18 09:20:08 -07:00
Marvin Löbel
3011801256 Made std::task::TaskBuilder::future_result() easier to use 2013-10-18 10:43:41 +02:00
Kevin Ballard
d6d9b92683 path2: Adjust the API to remove all the _str mutation methods
Add a new trait BytesContainer that is implemented for both byte vectors
and strings.

Convert Path::from_vec and ::from_str to one function, Path::new().

Remove all the _str-suffixed mutation methods (push, join, with_*,
set_*) and modify the non-suffixed versions to use BytesContainer.
2013-10-15 22:18:30 -07:00
Kevin Ballard
73d3d00ec4 path2: Replace the path module outright
Remove the old path.
Rename path2 to path.
Update all clients for the new path.

Also make some miscellaneous changes to the Path APIs to help the
adoption process.
2013-10-15 21:56:54 -07:00
Alex Crichton
2290ce14f2 Remove some users of io::file_reader 2013-10-10 03:38:51 -07:00
Huon Wilson
92725ae765 std::rand: Add a trait for seeding RNGs: SeedableRng.
This provides 2 methods: .reseed() and ::from_seed that modify and
create respecitively.

Implement this trait for the RNGs in the stdlib for which this makes
sense.
2013-10-09 22:22:42 +11:00
Huon Wilson
a2b509656a std::rand: Add an implementation of ISAAC64.
This is 2x faster on 64-bit computers at generating anything larger
than 32-bits.

It has been verified against the canonical C implementation from the
website of the creator of ISAAC64.

Also, move `Rng.next` to `Rng.next_u32` and add `Rng.next_u64` to
take full advantage of the wider word width; otherwise Isaac64 will
always be squeezed down into a u32 wasting half the entropy and
offering no advantage over the 32-bit variant.
2013-10-09 22:22:42 +11:00
Daniel Micay
313052aeb2 rm useless fast_ffi attributes
this is no longer used by the compiler
2013-10-08 09:03:43 -04:00
Alex Crichton
3396365cab Add appropriate #[feature] directives to tests 2013-10-06 14:39:25 -07:00
bors
d00c9269dc auto merge of #9665 : alexcrichton/rust/snapshot, r=brson
Uses the new snapshots to kill the old `loop` and introduce the new `continue`.
2013-10-02 02:31:29 -07:00
Alex Crichton
4f67dcb24a Migrate users of 'loop' to 'continue'
Closes #9467
2013-10-01 15:53:13 -07:00
Daniel Micay
c9d4ad07c4 remove the float type
It is simply defined as `f64` across every platform right now.

A use case hasn't been presented for a `float` type defined as the
highest precision floating point type implemented in hardware on the
platform. Performance-wise, using the smallest precision correct for the
use case greatly saves on cache space and allows for fitting more
numbers into SSE/AVX registers.

If there was a use case, this could be implemented as simply a type
alias or a struct thanks to `#[cfg(...)]`.

Closes #6592

The mailing list thread, for reference:

https://mail.mozilla.org/pipermail/rust-dev/2013-July/004632.html
2013-10-01 14:54:10 -04:00
Alex Crichton
73c6c9109f bench: Remove usage of fmt! 2013-09-30 23:21:20 -07:00
Alex Crichton
409182de6d Update the compiler to not use printf/printfln 2013-09-26 17:05:59 -07:00