So it appears the `Missing` item is correct, but it has been given the wrong SrcSpan. Originally it was a `noLoc`. I think I will just advance the column by one, then it is actually correct, and the annotation lookup will only succeed for the first one, so it will all just work. On Wed, Jun 17, 2015 at 5:14 PM, GHC wrote:

#10537: Parser: commas_tup_tail returns spurious "Missing" value -------------------------------------+------------------------------------- Reporter: alanz | Owner: alanz Type: bug | Status: new Priority: normal | Milestone: 7.10.2 Component: Compiler | Version: 7.10.1 (Parser) | Keywords: Resolution: | Architecture: Operating System: Unknown/Multiple | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Revisions: -------------------------------------+-------------------------------------

Comment (by mpickering):

I think you're right Simon. There is something wrong with the annotations though, firstly both the Missing elements have the same SrcSpan `L tests/examples/Tuple.hs:3:24`.

{{{ ((tests/examples/Tuple.hs:3:8, AnnComma), [tests/examples/Tuple.hs:3:9]), ((tests/examples/Tuple.hs:3:11-17, AnnComma), [tests/examples/Tuple.hs:3:18]), ((tests/examples/Tuple.hs:3:20-22, AnnComma), [tests/examples/Tuple.hs:3:23]), ((tests/examples/Tuple.hs:3:24, AnnComma), [tests/examples/Tuple.hs:3:24]), }}}

Here are the relevant annotations. Notice how the first three commas are associated with the preceding item (much like they are for lists) but the last one is associated directly with the `Missing`. As both `Missing`s have the same SrcSpan then this is causing duplicated output in `ghc- exactprint`.

-- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/10537#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler