search

ucsd-progsys / liquidhaskell

Liquid Types For Haskell
BSD 3-Clause "New" or "Revised" License
1.19k stars 137 forks source link

Slight differences in parser output #981

Closed alanz closed 7 years ago

alanz commented 7 years ago

In my parser-az branch I now have all the original tests passing.

The parse results are slightly different from before in that more places are getting dummy symbols.

e.g. x :: a -> a used to parse as

"Asrts ([\"x\" (dummyLoc)],(a -> a (dummyLoc),Nothing))"

but now does as

"Asrts ([\"x\" (dummyLoc)],(lq_tmp$db##0:a -> a (dummyLoc),Nothing))"

Is this a problem?

See https://github.com/alanz/liquidhaskell/commit/6a938f23ae258d74b8a7e4e2825b9716768278a5 for the changes to the tests, showing the old and new output for various cases.

ranjitjhala commented 7 years ago

Hi @alanz,

Thanks, this is awesome!

I think this is fine - in fact IIRC we should have put in a dummy binder in this case!

Should we merge now or?

Thanks a ton!

Ranjit.

On Sat, Apr 1, 2017 at 1:21 AM Alan Zimmerman notifications@github.com wrote:

In my parser-az https://github.com/alanz/liquidhaskell/tree/parser-az branch I now have all the original tests passing.

The parse results are slightly different from before in that more places are getting dummy symbols.

e.g. x :: a -> a used to parse as

"Asrts ([\"x\" (dummyLoc)],(a -> a (dummyLoc),Nothing))"

but now does as

"Asrts ([\"x\" (dummyLoc)],(lq_tmp$db##0:a -> a (dummyLoc),Nothing))"

Is this a problem?

See alanz@6a938f2 https://github.com/alanz/liquidhaskell/commit/6a938f23ae258d74b8a7e4e2825b9716768278a5 for the changes to the tests, showing the old and new output for various cases.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ucsd-progsys/liquidhaskell/issues/981, or mute the thread https://github.com/notifications/unsubscribe-auth/ABkuOPq41fBAUG9LZ4STgpYoN_oNGovOks5rrgkXgaJpZM4Mwahs .

alanz commented 7 years ago

You can merge now.

I will probably tweak a bit more in future, for specific error messages and to clean up the junk that is still there.

Alan

On 1 Apr 2017 17:09, "Ranjit Jhala" notifications@github.com wrote:

Hi @alanz,

Thanks, this is awesome!

I think this is fine - in fact IIRC we should have put in a dummy binder in this case!

Should we merge now or?

Thanks a ton!

Ranjit.

On Sat, Apr 1, 2017 at 1:21 AM Alan Zimmerman notifications@github.com wrote:

In my parser-az https://github.com/alanz/liquidhaskell/tree/parser-az branch I now have all the original tests passing.

The parse results are slightly different from before in that more places are getting dummy symbols.

e.g. x :: a -> a used to parse as

"Asrts ([\"x\" (dummyLoc)],(a -> a (dummyLoc),Nothing))"

but now does as

"Asrts ([\"x\" (dummyLoc)],(lq_tmp$db##0:a -> a (dummyLoc),Nothing))"

Is this a problem?

See alanz@6a938f2 https://github.com/alanz/liquidhaskell/commit/ 6a938f23ae258d74b8a7e4e2825b9716768278a5 for the changes to the tests, showing the old and new output for various cases.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ucsd-progsys/liquidhaskell/issues/981, or mute the thread https://github.com/notifications/unsubscribe-auth/ ABkuOPq41fBAUG9LZ4STgpYoN_oNGovOks5rrgkXgaJpZM4Mwahs .

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ucsd-progsys/liquidhaskell/issues/981#issuecomment-290925878, or mute the thread https://github.com/notifications/unsubscribe-auth/AAZAB0KWgmsxMW9-sKJ985L27PW9blOVks5rrmiVgaJpZM4Mwahs .