Alex Crichton [Sat, 21 Mar 2015 01:52:32 +0000 (18:52 -0700)]
Fix configuring lib targets with profile options
Alex Crichton [Sat, 21 Mar 2015 01:52:15 +0000 (18:52 -0700)]
Fix cargo test --no-run
Alex Crichton [Sat, 21 Mar 2015 00:48:55 +0000 (17:48 -0700)]
Fix test for `--test` filtering
The flag now only applies to integration tests, not all targets with the
specified name.
Alex Crichton [Sat, 21 Mar 2015 00:48:25 +0000 (17:48 -0700)]
Remove an extraneous is_test condition
If a profile's `test` flag is set to true, then it will always generate a unit
test so the `is_test` check is redundant.
Alex Crichton [Sat, 21 Mar 2015 00:47:34 +0000 (17:47 -0700)]
Fix cargo test filtering for binaries
This adds a new `--bin` flag to `cargo test` to specifically say that a binary
should be tested. Additionally the dependencies are tweaked such that binaries
to not depend on themselves being available.
Alex Crichton [Thu, 19 Feb 2015 19:30:35 +0000 (11:30 -0800)]
Overhaul how cargo treats profiles
This commit is a complete overhaul of how Cargo handles compilation profiles
internally. The external interface of Cargo is not affected by this change.
Previously each Target had a Profile embedded within it. Conceptually a Target
is an entry in the manifest (a binary, benchmark, etc) and a Profile controlled
various flags (e.g. --test, -C opt-level, etc). Each Package then contained many
profiles for each possible compilation mode. For example a Package would have
one target for testing a library, benchmarking a library, documenting a library,
etc. When it came to building these targets, Cargo would filter out the targets
listed to determine what needed to be built.
This filtering was largely done based off an "environment" represented as a
string. Each mode of compilation got a separate environment string like `"test"`
or `"bench"`. Altogether, however, this approach had a number of drawbacks:
* Examples in release mode do not currently work. This is due to how examples
are classified and how release mode is handled (e.g. the "release" environment
where examples are meant to be built in the "test" environment).
* It is not trivial to implement `cargo test --release` today.
* It is not trivial to implement `cargo build --bin foo` where *only* the binary
`foo` is built. The same is true for examples.
* It is not trivial to add selective building of a number of
binaries/examples/etc.
* Filtering the list of targets to build has some pretty hokey logic that
involves pseudo-environments like "doc-all" vs "doc". This logic is duplicated
in a few places and in general is quite brittle.
* The TOML parser greatly affects compilation due to the time at which profiles
are generated, which seems somewhat backwards.
* Profiles must be overridden, but only partially, at compile time becuase only
the top-level package's profile is applied.
In general, this system just needed an overhaul. This commit made a single
change of separating `Profile` from `Target` and then basically hit `make` until
all the tests passed again. The other large architectural changes are:
* Environment strings are now entirely gone.
* Filters are implemented in a much more robust fashion.
* Release mode is now handled much more gracefully.
* The unit of compilation in the backend is no longer (package, target) but
rather (package, target, profile). This change had to be propagated many
location in the `cargo_rustc` module.
* The backend does not filter targets to build *at all*. All filtering now
happens entirely in the frontend.
I'll test issues after this change lands, but the motivation behind this is to
open the door to quickly fixing a number of outstanding issues against Cargo.
This change itself is not intended to close many bugs.
bors [Thu, 19 Mar 2015 02:16:10 +0000 (02:16 +0000)]
Auto merge of #1433 - alexcrichton:dylib, r=huonw
This may not end up being stable for 1.0, and we don't really need much
functionality from it anyway.
Alex Crichton [Thu, 19 Mar 2015 01:47:10 +0000 (18:47 -0700)]
Remove dependency on DynamicLibrary
This may not end up being stable for 1.0, and we don't really need much
functionality from it anyway.
bors [Wed, 18 Mar 2015 18:07:53 +0000 (18:07 +0000)]
Auto merge of #1431 - IvanUkhov:newlines, r=alexcrichton
Hi,
It seems that people generally prefer not having newlines after section headers in their `Cargo.toml`s. However, when one creates a new package via `cargo new`, a newline gets automatically inserted after `[package]`. Then a typical `Cargo.toml` ends up being some what inconsistent: there is a newline after `[package]`, and there are no newlines after other headers.
On top of that, the examples given on crates.io do not seem to be consistent: [sometimes](http://doc.crates.io/index.html) there is a newline after `[package]`, and [sometimes](http://doc.crates.io/manifest.html) there is none. The same applies to other section names.
I am wondering if there should be some style conventions in this regard.
Regards,
Ivan
bors [Wed, 18 Mar 2015 17:43:25 +0000 (17:43 +0000)]
Auto merge of #1428 - alexcrichton:update, r=alexcrichton
Ivan Ukhov [Wed, 18 Mar 2015 08:12:06 +0000 (09:12 +0100)]
Make the usage of newlines in Cargo.toml consistent
bors [Tue, 17 Mar 2015 16:32:06 +0000 (16:32 +0000)]
Auto merge of #1429 - IvanUkhov:build-script, r=alexcrichton
Alex Crichton [Mon, 16 Mar 2015 13:27:58 +0000 (06:27 -0700)]
Update to rust master
Ivan Ukhov [Tue, 17 Mar 2015 07:22:44 +0000 (08:22 +0100)]
Fix a typo in the description of the links manifest key
bors [Mon, 16 Mar 2015 13:47:04 +0000 (13:47 +0000)]
Auto merge of #1425 - alexcrichton:issue-1421, r=huonw
This was accidentally broken in a recent refactoring.
Closes #1421
Alex Crichton [Mon, 16 Mar 2015 13:32:50 +0000 (06:32 -0700)]
Re-fix the `help` command's implementation
This was accidentally broken in a recent refactoring.
Closes #1421
bors [Sat, 14 Mar 2015 18:01:07 +0000 (18:01 +0000)]
Auto merge of #1402 - cmr:issue-1015, r=alexcrichton
Deprecates the rustc-args metadata keyword in favor of rustc-link-lib and
rustc-link-search, which are more precise and allows for easy, correct
handling of spaces in pathnames.
Closes #1015
Corey Richardson [Tue, 10 Mar 2015 19:58:18 +0000 (15:58 -0400)]
Add rustc_private feature to tests which need it.
Corey Richardson [Tue, 10 Mar 2015 17:12:47 +0000 (13:12 -0400)]
Add dedicated metadata for adding -l/-L flags
Deprecates the rustc-args metadata keyword in favor of rustc-link-lib and
rustc-link-search, which are more precise and allows for easy, correct
handling of spaces in pathnames.
Closes #1015
bors [Thu, 12 Mar 2015 17:21:14 +0000 (17:21 +0000)]
Auto merge of #1412 - jakub-:rust-nightly, r=alexcrichton
Jakub Bukaj [Thu, 12 Mar 2015 10:42:59 +0000 (11:42 +0100)]
Update to the latest nightly
bors [Thu, 12 Mar 2015 02:41:03 +0000 (02:41 +0000)]
Auto merge of #1409 - alexcrichton:less-compiles, r=huonw
This was previously required to ensure that all relevant metadata files would
remain in place and not get deleted, but this auto-clean behavior has since been
removed so this should no longer be necessary.
bors [Thu, 12 Mar 2015 02:29:59 +0000 (02:29 +0000)]
Auto merge of #1406 - alexcrichton:issue-1404, r=wycats
When loading targets to compile, be sure to use the targets from the package
that's being passed down to the compilation step, not the one that was passed in
which is overridden.
Closes #1404
bors [Thu, 12 Mar 2015 02:17:26 +0000 (02:17 +0000)]
Auto merge of #1410 - johnz133:patch-2, r=huonw
cargo build now outputs to /target/debug, changed doc to reflect that.
steve suggested to add description for cargo build --release
autogenerates to guide.html. Used pre and code tags instead of ``` by following previous examples.
John Zhang [Thu, 12 Mar 2015 01:59:19 +0000 (18:59 -0700)]
Updated build in guide.md
autogenerates to guide.html
Alex Crichton [Thu, 12 Mar 2015 01:33:17 +0000 (18:33 -0700)]
Don't add non-compiled packages to the graph
This was previously required to ensure that all relevant metadata files would
remain in place and not get deleted, but this auto-clean behavior has since been
removed so this should no longer be necessary.
Alex Crichton [Wed, 11 Mar 2015 20:32:59 +0000 (13:32 -0700)]
Don't intermingle targets between packages
When loading targets to compile, be sure to use the targets from the package
that's being passed down to the compilation step, not the one that was passed in
which is overridden.
Closes #1404
bors [Tue, 10 Mar 2015 03:42:12 +0000 (03:42 +0000)]
Auto merge of #1396 - alexcrichton:issue-1390, r=huonw
When calculating the output directory filename, only use the "file name"
component of the target triple specified as otherwise the output could be placed
into an odd location.
Closes #1390
Alex Crichton [Mon, 9 Mar 2015 16:49:33 +0000 (09:49 -0700)]
Improve experience with the flexible target specification
When calculating the output directory filename, only use the "file name"
component of the target triple specified as otherwise the output could be placed
into an odd location.
Closes #1390
bors [Mon, 9 Mar 2015 22:54:10 +0000 (22:54 +0000)]
Auto merge of #1395 - alexcrichton:profile, r=brson
* Don't use `{:?}` in profiling output, it's too noisy
* Allow an integer to be specified with `CARGO_PROFILE` indicating how many
levels deep should be printed.
bors [Mon, 9 Mar 2015 20:51:30 +0000 (20:51 +0000)]
Auto merge of #1397 - alexcrichton:update, r=alexcrichton
Continuation of #1384
Alex Crichton [Mon, 9 Mar 2015 19:07:35 +0000 (12:07 -0700)]
Fix all tests from the update to rust master
Alex Crichton [Mon, 9 Mar 2015 18:38:46 +0000 (11:38 -0700)]
Reduce usage of unstable features
Alex Crichton [Mon, 9 Mar 2015 18:30:37 +0000 (11:30 -0700)]
Update to rust master
Alex Crichton [Mon, 9 Mar 2015 18:13:32 +0000 (11:13 -0700)]
Merge branch 'master' of https://github.com/Victor-Savu/cargo into update
Alex Crichton [Mon, 9 Mar 2015 16:45:40 +0000 (09:45 -0700)]
Improve the profiling output of Cargo
* Don't use `{:?}` in profiling output, it's too noisy
* Allow an integer to be specified with `CARGO_PROFILE` indicating how many
levels deep should be printed.
bors [Sat, 7 Mar 2015 19:35:09 +0000 (19:35 +0000)]
Auto merge of #1389 - jimmycuadra:patch-1, r=alexcrichton
Add documentation for the `harness` manifest option, introduced in
a6589681351dc4a87b6fc268d47b7d608004b455.
Jimmy Cuadra [Sat, 7 Mar 2015 12:53:12 +0000 (04:53 -0800)]
Document the `harness` manifest option.
Add documentation for the `harness` manifest option, introduced in
a6589681351dc4a87b6fc268d47b7d608004b455.
Victor-Nicolae Savu [Fri, 6 Mar 2015 23:54:30 +0000 (00:54 +0100)]
Reverting to rustc nightly 2015-03-04 because threadpool build fails with more recent versions.
Victor-Nicolae Savu [Fri, 6 Mar 2015 21:54:41 +0000 (22:54 +0100)]
Merge branch 'master' of https://github.com/rust-lang/cargo
Victor-Nicolae Savu [Fri, 6 Mar 2015 21:52:55 +0000 (22:52 +0100)]
Updated versions for rustc, curl, docopt and regex.
bors [Fri, 6 Mar 2015 18:53:53 +0000 (18:53 +0000)]
Auto merge of #1373 - alexcrichton:issue-785, r=wycats
This commit now funnels all output of Cargo by default to be in `target/debug`
instead of the bare `target` directory. This change is targeted at raising
awareness of whether a debug build is being used (as opposed to a release
build). It is also aimed at remedying a common scenario where `cargo build` is
followed by `cargo build --release` and then the debug binaries are run by
accident.
This does not yet explore the option of providing symlinks to the most recent
build, hence this commit is a breaking change due to the restructuring of the
layout of the output.
Note that this commit does **not** change the output location for documentation.
All output of `cargo doc` continues to be funneled into the `target/doc`
directory.
Closes #785
Alex Crichton [Tue, 3 Mar 2015 21:46:02 +0000 (13:46 -0800)]
Change the default output location to target/debug
This commit now funnels all output of Cargo by default to be in `target/debug`
instead of the bare `target` directory. This change is targeted at raising
awareness of whether a debug build is being used (as opposed to a release
build). It is also aimed at remedying a common scenario where `cargo build` is
followed by `cargo build --release` and then the debug binaries are run by
accident.
This does not yet explore the option of providing symlinks to the most recent
build, hence this commit is a breaking change due to the restructuring of the
layout of the output.
Note that this commit does **not** change the output location for documentation.
All output of `cargo doc` continues to be funneled into the `target/doc`
directory.
Closes #785
bors [Fri, 6 Mar 2015 18:34:43 +0000 (18:34 +0000)]
Auto merge of #1387 - alexcrichton:issue-1382, r=brson
A synthetic "Package" is being created as part of this step but its paths were
pointing at the original root, not the unpacked tarball for testing. This meant
that some assumptions about the relative-ness of paths didn't quite end up
working out.
Closes #1382
Alex Crichton [Thu, 5 Mar 2015 23:29:05 +0000 (15:29 -0800)]
Correctly set up paths when packaging
A synthetic "Package" is being created as part of this step but its paths were
pointing at the original root, not the unpacked tarball for testing. This meant
that some assumptions about the relative-ness of paths didn't quite end up
working out.
Closes #1382
bors [Fri, 6 Mar 2015 01:41:22 +0000 (01:41 +0000)]
Auto merge of #1365 - ches:docs, r=brson
While reading through the docs, this section broke my concentration even though the functionality's purpose is fairly obvious, because of a grammatical agreement problem and some redundant/wordy terms. I hope this is an improvement.
A couple of other minor changes:
1. "Limitations" sound like things Cargo is *incapable* of doing. The points here are conscious decisions, so calling them "restrictions" seemed more appropriate to me.
2. TOML's source is linked to on the FAQ page, but I'm pretty likely to land directly on the Configuration page from a search, so a direct link from there seemed helpful.
r? @steveklabnik for docs.
bors [Fri, 6 Mar 2015 00:47:35 +0000 (00:47 +0000)]
Auto merge of #1383 - alexcrichton:issue-1381, r=huonw
Ensure we don't get erroneous information about invalid log levels or such.
Closes #1381
Victor-Nicolae Savu [Thu, 5 Mar 2015 22:04:40 +0000 (23:04 +0100)]
fixed feature attribute
Victor-Nicolae Savu [Thu, 5 Mar 2015 21:41:59 +0000 (22:41 +0100)]
removed [feature(env)] attributes
Alex Crichton [Thu, 5 Mar 2015 20:11:08 +0000 (12:11 -0800)]
Explicitly remove RUST_LOG when crawling output
Ensure we don't get erroneous information about invalid log levels or such.
Closes #1381
Victor-Nicolae Savu [Thu, 5 Mar 2015 06:20:58 +0000 (07:20 +0100)]
Updated to rust nightly 2015-03-04
bors [Wed, 4 Mar 2015 17:47:32 +0000 (17:47 +0000)]
Auto merge of #1376 - fenhl:patch-1, r=alexcrichton
`dependencie"` → `dependencies`
Fenhl [Wed, 4 Mar 2015 12:31:34 +0000 (12:31 +0000)]
Fix a typo in Zsh completion
`dependencie"` → `dependencies`
bors [Tue, 3 Mar 2015 04:33:24 +0000 (04:33 +0000)]
Auto merge of #1372 - alexcrichton:flaky, r=alexcrichton
Due to random hash map traversals this test could build either crate `a` or
crate `b` first, causing the first four output lines to possibly change. Two of
the first four lines are guaranteed to be the compilation of `b` and the other
two are the compilation of `a` as well as building its build script, but no
other lines about compiling `a` can happen until the dependency `b` has finished
compiling.
Alex Crichton [Tue, 3 Mar 2015 04:31:13 +0000 (20:31 -0800)]
Relax output restrictions to make a test less flaky
Due to random hash map traversals this test could build either crate `a` or
crate `b` first, causing the first four output lines to possibly change. Two of
the first four lines are guaranteed to be the compilation of `b` and the other
two are the compilation of `a` as well as building its build script, but no
other lines about compiling `a` can happen until the dependency `b` has finished
compiling.
bors [Tue, 3 Mar 2015 03:52:16 +0000 (03:52 +0000)]
Auto merge of #1370 - alexcrichton:issue-1299, r=huonw
Apparently git submodules are checked in as '/' not '\' so if a '\' leaks
through it'll end up not being found!
Closes #1299
bors [Tue, 3 Mar 2015 03:36:55 +0000 (03:36 +0000)]
Auto merge of #1371 - alexcrichton:issue-1289, r=huonw
Before this commit the `dep_targets` function of `Context` didn't actually
return all dependencies in the sense that unit tests (and likely examples) would
not have the same package's library listed as a dependency. This commit
rectifies the situation by ensuring that the package's library is included
whenever necessary in the dependency list.
Closes #1289
Alex Crichton [Tue, 3 Mar 2015 03:17:25 +0000 (19:17 -0800)]
Fix dep_targets for tests depending on libraries
Before this commit the `dep_targets` function of `Context` didn't actually
return all dependencies in the sense that unit tests (and likely examples) would
not have the same package's library listed as a dependency. This commit
rectifies the situation by ensuring that the package's library is included
whenever necessary in the dependency list.
Closes #1289
Alex Crichton [Mon, 16 Feb 2015 04:23:09 +0000 (20:23 -0800)]
Translate '\' to '/' for submodule names
Apparently git submodules are checked in as '/' not '\' so if a '\' leaks
through it'll end up not being found!
Closes #1299
bors [Mon, 2 Mar 2015 22:15:51 +0000 (22:15 +0000)]
Auto merge of #1368 - cyndis:lto-dyn, r=alexcrichton
When building host code, -C prefer-dynamic is passed, but that option
is mutually exclusive with -C lto. Thus when LTO is specified, rustc
errors out.
Mikko Perttunen [Mon, 2 Mar 2015 17:09:05 +0000 (19:09 +0200)]
Add test for having lto enabled while having a build script
Mikko Perttunen [Sun, 1 Mar 2015 22:53:46 +0000 (00:53 +0200)]
Disable LTO for host (plugin, build script) builds
When building host code, -C prefer-dynamic is passed, but that option
is mutually exclusive with -C lto. Thus when LTO is specified, rustc
errors out.
bors [Sun, 1 Mar 2015 19:10:13 +0000 (19:10 +0000)]
Auto merge of #1367 - alexcrichton:fix-tar, r=alexcrichton
Fixes a bug reported in alexcrichton/flate2-rs#15 where the header wasn't being
properly read.
Alex Crichton [Sun, 1 Mar 2015 19:09:40 +0000 (11:09 -0800)]
Update tar dependency
Fixes a bug reported in alexcrichton/flate2-rs#15 where the header wasn't being
properly read.
Ches Martin [Sat, 28 Feb 2015 16:57:42 +0000 (11:57 -0500)]
docs: Improve grammar/wording for `cargo owner`
bors [Fri, 27 Feb 2015 23:53:09 +0000 (23:53 +0000)]
Auto merge of #1360 - alexcrichton:new-io, r=alexcrichton
cc @aturon, this is one giant commit.
Alex Crichton [Fri, 27 Feb 2015 01:04:25 +0000 (17:04 -0800)]
Update to rust master and std::{io, path}
bors [Wed, 25 Feb 2015 18:27:54 +0000 (18:27 +0000)]
Auto merge of #1354 - mbrubeck:gh-pages-title, r=alexcrichton
This should make them easier to find via search engines, or in browser history/bookmarks.
See #1302 for an old version of this PR.
Matt Brubeck [Wed, 25 Feb 2015 17:09:11 +0000 (09:09 -0800)]
Add "Cargo" to doc titles
bors [Wed, 25 Feb 2015 04:33:45 +0000 (04:33 +0000)]
Auto merge of #1347 - alexcrichton:no-more-urllib, r=huonw
We've had too many problems using urllib to make it worth it.
Closes #1346
Closes #1088
cc #1319 (using a proxy through curl)
Alex Crichton [Wed, 25 Feb 2015 04:32:06 +0000 (20:32 -0800)]
Merge pull request #1350 from sfackler/master
Don't time out after 0.006 seconds
Steven Fackler [Wed, 25 Feb 2015 04:20:59 +0000 (20:20 -0800)]
Don't time out after 0.006 seconds
Alex Crichton [Tue, 24 Feb 2015 23:44:37 +0000 (15:44 -0800)]
Fixup some python style which was previously reverted
Alex Crichton [Tue, 24 Feb 2015 18:07:36 +0000 (10:07 -0800)]
Probe for `curl` when configuring
Alex Crichton [Tue, 24 Feb 2015 18:06:04 +0000 (10:06 -0800)]
Revert "Directly use python’s urllib instead of curl in etc/dl-snapshot.py"
This reverts commit
aceba880fdf83cbd694fa1a6e572c600b17d1c37.
Conflicts:
src/etc/dl-snapshot.py
bors [Tue, 24 Feb 2015 20:14:59 +0000 (20:14 +0000)]
Auto merge of #1345 - badboy:pass-cargo-name, r=alexcrichton
This reverts a change from
ee5e24ff8b5.
Before it passed "foo" as `argv[0]`, which does not help much (consider the case there is one script symlinked to different names to handle different things).
This patch changes that to `cargo-{command}`.
It also reverts the passing of "foo" on the help command. Before it used `os::arg`, a different (and ugly) solution would be:
let args = &[env::args().next().unwrap().clone(), "-h".to_string()];
Jan-Erik Rediger [Tue, 24 Feb 2015 11:00:58 +0000 (12:00 +0100)]
Strip first unused argument when passing to subcommand
bors [Tue, 24 Feb 2015 19:57:32 +0000 (19:57 +0000)]
Auto merge of #1348 - alexcrichton:rustup, r=alexcrichton
Alex Crichton [Tue, 24 Feb 2015 19:14:38 +0000 (11:14 -0800)]
Fixup warnings and the travis script
Alex Crichton [Tue, 24 Feb 2015 18:58:25 +0000 (10:58 -0800)]
Merge branch 'rustup' of https://github.com/globin/cargo into rustup
bors [Tue, 24 Feb 2015 17:36:11 +0000 (17:36 +0000)]
Auto merge of #1344 - lfairy:dehyphenator, r=alexcrichton
Closes #1342.
bors [Tue, 24 Feb 2015 17:13:58 +0000 (17:13 +0000)]
Auto merge of #1291 - alexcrichton:issue-1083, r=huonw
Hopefully this will help diagnose #1083
Robin Gloster [Tue, 24 Feb 2015 12:12:05 +0000 (13:12 +0100)]
fix style
Robin Gloster [Tue, 24 Feb 2015 12:00:58 +0000 (13:00 +0100)]
fix travis script
Robin Gloster [Sun, 22 Feb 2015 23:06:12 +0000 (00:06 +0100)]
rustup
Chris Wong [Tue, 24 Feb 2015 07:44:33 +0000 (20:44 +1300)]
Use an underscore instead of a hyphen as a dummy crate name
Closes #1342.
bors [Tue, 24 Feb 2015 00:28:31 +0000 (00:28 +0000)]
Auto merge of #1340 - alexcrichton:issue-1329, r=huonw
Closes #1329
bors [Tue, 24 Feb 2015 00:17:22 +0000 (00:17 +0000)]
Auto merge of #1341 - alexcrichton:issue-1333, r=huonw
Closes #1333
Alex Crichton [Tue, 24 Feb 2015 00:16:17 +0000 (16:16 -0800)]
Bump default timeout to 60s
Closes #1333
Alex Crichton [Tue, 24 Feb 2015 00:14:34 +0000 (16:14 -0800)]
Download registry packages on `cargo fetch`
Closes #1329
bors [Mon, 23 Feb 2015 17:45:20 +0000 (17:45 +0000)]
Auto merge of #1339 - jakub-:patch-1, r=alexcrichton
Jakub [Mon, 23 Feb 2015 13:12:44 +0000 (14:12 +0100)]
Add `Third-party cargo subcommands` to the README
bors [Thu, 19 Feb 2015 15:44:24 +0000 (15:44 +0000)]
Auto merge of #1321 - dlevy47:bugfix/no-proxy, r=alexcrichton
Currently, cargo uses `curl_easy_setopt` to explicitly set the http proxy on a curl handle if any of the following are present:
* cargo config `http.proxy`
* git config `http.proxy`
* `HTTP_PROXY` environment variable
The act of explicitly setting the http proxy disables curl's logic for determining which proxy to use for a URL, and breaks the `no_proxy` environment variable. This PR privatizes `cargo::ops::registry::http_proxy` and adds an additional function `cargo::ops::registry::http_proxy_exists` to determine whether or not to use the `git2-curl` subtransport.
calc0000 [Wed, 18 Feb 2015 01:39:27 +0000 (20:39 -0500)]
Change cargo to explicitly set curl proxy settings only when necessary
(when it is specified in either the cargo config or the git config).
Otherwise, use curl's logic to pick a proxy (or not).
bors [Wed, 18 Feb 2015 16:20:53 +0000 (16:20 +0000)]
Auto merge of #1288 - alexcrichton:issue-1058, r=huonw
Previously the build script's profile was used instead of the target's build
script, causing the wrong environment variables to get set.
Closes #1058
Alex Crichton [Wed, 11 Feb 2015 05:03:31 +0000 (21:03 -0800)]
Set build script env vars correctly
Previously the build script's profile was used instead of the target's build
script, causing the wrong environment variables to get set.
Closes #1058
bors [Tue, 17 Feb 2015 17:32:25 +0000 (17:32 +0000)]
Auto merge of #1314 - AnthIste:build-script-updates, r=alexcrichton
The examples given for `build.rs` do not compile against the latest rust. Instead of adapting to use modules such as `std::old_io`, I attempted to port the examples to use the new `io` and `path` modules.
Ruben Bakker [Sun, 15 Feb 2015 20:34:51 +0000 (22:34 +0200)]
Update example build.rs scripts to reflect latest Rust changes
bors [Tue, 17 Feb 2015 05:00:30 +0000 (05:00 +0000)]
Auto merge of #1317 - alexcrichton:issue-1297, r=huonw
Not sure why we were doing this, it's quite useful!
Closes #1297
Alex Crichton [Mon, 16 Feb 2015 04:45:59 +0000 (20:45 -0800)]
Don't discard docopt error information
Not sure why we were doing this, it's quite useful!
Closes #1297