The array of options that is passed to the spawned build process is a
combination of the current ARGV, options passed in by a dependent
formula, and an existing install receipt. The objects that are
interacting here each expect the resulting collection to have certain
properties, and the expectations are not consistent.
Clear up this confusing mess by only dealing with Options collections.
This keeps our representation of options uniform across the codebase.
We can remove BuildOptions dependency on HomebrewArgvExtension, which
allows us to pass any Array-like collection to Tab.create. The only
other site inside of FormulaInstaller that uses the array is the #exec
call, and there it is splatted and thus we can substitute our Options
collection there as well.
Formulae can now pass build options to dependencies. The following
syntax is supported:
depends_on 'foo' => 'with-bar'
depends_on 'foo' => ['with-bar', 'with-baz']
If a dependency is already installed but lacks the required build
options, an exception is raised. Eventually we may be able to just stash
the existing keg and reinstall it with the combined set of used_options
and passed options, but enabling that is left for another day.
Move Formula.expand_dependencies into the Dependency class, and extend
it to allow arbitrary filters to be applied when enumerating deps.
When supplied with a block, expand_dependencies will yield a [dependent,
dependency] pair for each dependency, allowing callers to filter out
dependencies that may not be applicable or useful in a given situation.
Deps can be skipped by simple calling Dependency.prune in the block,
e.g.:
Dependency.expand_dependencies do |f, dep|
Dependency.prune if dep.to_formula.installed?
end
The return value of the method is the filtered list.
If no block is supplied, a default filter that omits optional or
recommended deps based on what the dependent formula has requested is
applied.
Formula#recursive_dependencies is now implemented on top of this,
allowing FormulaInstaller to exact detailed control over what deps are
installed. `brew missing` and `brew upgrade` can learn to use this to
apply the installed options set when expanding dependencies.
Move Formula.expand_deps and Formula#recursive_deps into compat, because
these methods do not respect the new optional and recommended tags and
thus should no longer be used.
Optional deps are not installed by default but generate a corresponding
"with-foo" option for the formula. Recommended deps _are_ installed by
default, and generate a corresponding "without-foo" option.
FormulaInstaller now attempts to take a lock on a "foo.brewing" file for
the formula and all of its dependencies before attempting installation.
The lock is an advisory lock implemented using flock(), and as such it
only locks out other processes that attempt to take the lock. It also
means that it is never necessary to manually remove the lock file,
because the lock is not enforced by I/O.
The uninstall, link, and unlink commands all learn to respect this lock
as well, so that the installation cannot be corrupted by a concurrent
Homebrew process, and keg operations cannot occur simultaneously.
This behaves like recursive_deps, but the resulting list consists of
Dependency objects instead of Formula objects. The list maintains the
installable order property of recursive_deps.
While in the area, add some comments clarifying the purpose of related
methods.
Currently we handle options in several ways, and it is hard to remember
what code needs an option string ("--foo"), what needs only the name
("foo") and what needs an Option object.
Now that Option objects can act as strings and be converted to JSON, we
can start using them instead of passing around strings between Formula
objects, Tab objects, and ARGV-style arrays.
The Options class is a special collection that can be queried for the
inclusion of options in any form: '--foo', 'foo', or Option.new("foo").
We want to be able to use Option objects in place of strings and have
this be transparent. Defining to_str means that methods like
Kernel#system and Kernel#exec will be able to perform an implicit
conversion.
Superenv normally filters out "-m32" flag, preventing 32bit builds.
Some software, however, still only work in 32bit mode.
If ENV.m32 is called, superenv does not filter out the "-m32" flag.
Also note, superenv, does not explicitly add the -m32 flag and
expects the build system of the software to know when and where to
provide this flag.
ClosesHomebrew/homebrew#16350.
Signed-off-by: Adam Vandenberg <flangy@gmail.com>
We already (correctly) allow -Wl, style linker arguments to pass
through; extend this to -Wp, (preprocessor) and -Wa, (assembler).
FixesHomebrew/homebrew#17252.
When a requirement is specified like:
satisfy { which "foo" }
There is no reason that we should inject all of ENV.userpaths! into the
build environment. Instead, infer the directory to be added to PATH from
the Pathname that is returned.
This is another step towards condensing the "which program" requirements
down into a one-liner DSL element.
Instead of overriding #satisfied?, Requirement subclasses can specify
the condition in a block:
satisfy do
some_condition?
end
The contents of the block are evaluated in the context of the instance,
and so have access to instance variables and instance methods as before.
Additionally, it is wrapped in an ENV.with_build_environment block. This
can be disabled by passing :build_env => false to satisfy:
satisfy :build_env => false do
some_condition?
end
Not thread safe! But I don't think we care.
We want to evaluate the env DSL block in the context of ENV for asthetic
reasons, but we also want access to methods on the requirement instance.
We can use #instance_exec to pass the requirement itself into the block:
class Foo < Requirement
env do |req|
append 'PATH', req.some_path
end
def some_path
which 'something'
end
end
Also add a simplified version of Object#instance_exec for Ruby 1.8.6.
