Monotonic locations (#17632)

When GHC is parsing a file generated by a tool, e.g. by the C preprocessor, the
tool may insert #line pragmas to adjust the locations reported to the user.

As the result, the locations recorded in RealSrcLoc are not monotonic. Elements
that appear later in the StringBuffer are not guaranteed to have a higher
line/column number.

In fact, there are no guarantees whatsoever, as #line pragmas can arbitrarily
modify locations. This lack of guarantees makes ideas such as #17544
infeasible.

This patch adds an additional bit of information to every SrcLoc:

	newtype BufPos = BufPos { bufPos :: Int }

A BufPos represents the location in the StringBuffer, unaffected by any
pragmas.

Updates haddock submodule.

Metric Increase:
    haddock.Cabal
    haddock.base
    haddock.compiler
    MultiLayerModules
    Naperian
    parsing001
    T12150

1 files changed, 17 insertions, 2 deletions

diff --git a/compiler/utils/Binary.hs b/compiler/utils/Binary.hs
index 498c4924de..1c52cb56fb 100644
--- a/compiler/utils/Binary.hs
+++ b/compiler/utils/Binary.hs
@@ -1380,10 +1380,24 @@ instance Binary RealSrcSpan where
             return (mkRealSrcSpan (mkRealSrcLoc f sl sc)
                                   (mkRealSrcLoc f el ec))
 
+instance Binary BufPos where
+  put_ bh (BufPos i) = put_ bh i
+  get bh = BufPos <$> get bh
+
+instance Binary BufSpan where
+  put_ bh (BufSpan start end) = do
+    put_ bh start
+    put_ bh end
+  get bh = do
+    start <- get bh
+    end <- get bh
+    return (BufSpan start end)
+
 instance Binary SrcSpan where
-  put_ bh (RealSrcSpan ss) = do
+  put_ bh (RealSrcSpan ss sb) = do
           putByte bh 0
           put_ bh ss
+          put_ bh sb
 
   put_ bh (UnhelpfulSpan s) = do
           putByte bh 1
@@ -1393,7 +1407,8 @@ instance Binary SrcSpan where
           h <- getByte bh
           case h of
             0 -> do ss <- get bh
-                    return (RealSrcSpan ss)
+                    sb <- get bh
+                    return (RealSrcSpan ss sb)
             _ -> do s <- get bh
                     return (UnhelpfulSpan s)