Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

In general, the test suite will not have any servers configured
so this warning is not very useful.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

This flips error handling around, so that 'die' now can format
an error message with call stacks and markers before raising
it to the top handler.  The top handler detects "verbatim"
deaths and prints them without formatting.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

Now if you say cabal -v"info +callstacks", Cabal invocations
will also get call stacks.

There's a heinous hack to handle version of Cabal that don't
support the extended format.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

New output:

Error:
    Problem with module re-exports:
      - The module 'Asdf'
        is not exported by any suitable package.
        It occurs in neither the 'exposed-modules' of this package,
        nor any of its 'build-depends' dependencies.
    In the stanza 'library'
    In the inplace package 'Reexport2-1.0'

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

* This is only enabled with the lib flag.
* We tell users not to use the library in the package description.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

This reverts commit bacdf19b.

'hackage-security' has been updated. I tested that build works now.

hackage-security hasn't been updated with new bounds, so this
unbreaks the build.

This reverts commit eec15a35.

[ci skip]

Edward Z. Yang authored 8 years ago and Edward Z. Yang committed 8 years ago

- Report all reexport errors, not just the first one
- Give more information about how the modules are different
- Add quotes and prettify the foramt
- Preserve spaces when reporting errors

The new message looks like this:

Problem with module re-exports:
  - Ambiguous reexport 'Data.Map'
    It could refer to either:
         'containers-0.5.6.2-59326c33e30ec8f6afd574cbac625bbb:Data.Map'
         brought into scope by the build dependency on containers
      or 'containers-dupe-0.1.0.0-2AdbRP7BsOEKELRWSQejuE:Data.Map'
         brought into scope by the build dependency on containers-dupe
    The ambiguity can be resolved by qualifying the
    re-export with a package name.
    The syntax is 'packagename:ModuleName [as NewName]'.
  - The module 'Missing'
    is not exported by any suitable package.
    It occurs in neither the 'exposed-modules' of this package,
    nor any of its 'build-depends' dependencies.
  - The module 'Private'
    is not exported by any suitable package.
    It occurs in neither the 'exposed-modules' of this package,
    nor any of its 'build-depends' dependencies
setup: Configuration failed

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

kristenk authored 8 years ago and Mikhail Glushenkov committed 8 years ago

Old output:
dependencies not linked: stanza TestStanzas incompatible;
dependencies not linked: flag FlagName "parsec" incompatible

New output:
dependencies not linked: stanza "test" incompatible;
dependencies not linked: flag "parsec" incompatible;

It's really an error to try and build a graph where you have duplicate
node keys, so remove Graph.fromList and add Graph.fromDistinctList. This
check is always on, not just an assertion, becuase we get it for free
given the way Maps can be constructed.

All uses of Graph.fromList are ok to convert to fromDistinctList.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

kristenk authored 8 years ago and Edward Z. Yang committed 8 years ago

For example, "setup.Cabal installed" forces cabal to use the installed Cabal
library for all setup scripts.

kristenk authored 8 years ago and Edward Z. Yang committed 8 years ago

For example, --constraint="any.pkg == 5" applies to "pkg" whether it is a
top-level dependency, setup dependency, or build tool dependency.

I also modified the UserConstraint type so that it is more similar to the
PackageConstraint type, now that both types need to express similar
"constraint scopes".

kristenk authored 8 years ago and Edward Z. Yang committed 8 years ago

I used the new constraint scope to enforce the minimum Cabal version in setup
scripts that lack a 'custom-setup' stanza.  It isn't exposed to users yet.

kristenk authored 8 years ago and Edward Z. Yang committed 8 years ago

This commit changes the meaning of constraints like 'pkg > 5'. Previously,
'pkg > 5' only applied to the top-level 'pkg' goal without a 'Namespace', but
now it applies to any top-level goal for 'pkg'. However, the 'Namespace' field is
currently only used by --independent-goals, so this change has no effect on
behavior.

echo-0.1.3 introduces a dependency on the minimalistic mintty library, so we
need to add that to ./bootstrap.sh as well.

mapDeps gives us all components, not just all exes components, so we need
to handle that. I'm feeling a bit guilty for the prior partiality, so added
some more explanatory comments too :).

The executable is unused so it shouldn't need to be built and
cause problems.

Also, convert the test to use the testsuite

I'd like to instead change the solver to route the used
components through (baby steps towards per-component
solving), rather than reconstruct it post-hoc from the
Package Description, but this gets the job done for now.