base: Expose GHC.Unicode.unicodeVersion

This exposes a Data.Version.Version representing the version of the
Unicode database used by `base`. This should clear up some confusion I
have seen in tickets regarding with which Unicode versions a given GHC
can be expected to work.

While in town I also regenerated (but did not update) the Unicode
database with database 12.0.0. Strangely, the file cited in the README
no longer existed. Consequently, I used
https://www.unicode.org/Public/12.0.0/ucd/UnicodeData.txt and was
slightly surprised to find that there were a few changes.

6 files changed, 65 insertions, 12 deletions

diff --git a/libraries/base/Data/Version.hs-boot b/libraries/base/Data/Version.hs-boot
new file mode 100644
index 0000000000..63726cf6af
--- /dev/null
+++ b/libraries/base/Data/Version.hs-boot
@@ -0,0 +1,12 @@
+{-# LANGUAGE NoImplicitPrelude #-}
+
+module Data.Version
+  ( Version
+  , makeVersion
+  ) where
+
+import GHC.Base
+
+data Version
+
+makeVersion :: [Int] -> Version
diff --git a/libraries/base/GHC/Unicode.hs b/libraries/base/GHC/Unicode.hs
index 9d11b37d0c..6fba91f0e2 100644
--- a/libraries/base/GHC/Unicode.hs
+++ b/libraries/base/GHC/Unicode.hs
@@ -19,6 +19,7 @@
 -----------------------------------------------------------------------------
 
 module GHC.Unicode (
+        unicodeVersion,
         GeneralCategory (..), generalCategory,
         isAscii, isLatin1, isControl,
         isAsciiUpper, isAsciiLower,
@@ -36,12 +37,18 @@ import GHC.Real
 import GHC.Enum ( Enum (..), Bounded (..) )
 import GHC.Ix ( Ix (..) )
 import GHC.Num
+import {-# SOURCE #-} Data.Version
 
 -- Data.Char.chr already imports this and we need to define a Show instance
 -- for GeneralCategory
 import GHC.Show ( Show )
 
 #include "HsBaseConfig.h"
+#include "UnicodeVersion.h"
+
+-- | Version of Unicode standard used by @base@.
+unicodeVersion :: Version
+unicodeVersion = makeVersion UNICODE_VERSION_NUMS
 
 -- | Unicode General Categories (column 2 of the UnicodeData table) in
 -- the order they are listed in the Unicode standard (the Unicode
diff --git a/libraries/base/cbits/README.Unicode b/libraries/base/cbits/README.Unicode
index 6cc18464cd..1eef278c96 100644
--- a/libraries/base/cbits/README.Unicode
+++ b/libraries/base/cbits/README.Unicode
@@ -1,8 +1,12 @@
+Generating GHC's Unicode table
+==============================
 
 WCsubst.c is generated with:
 
-    sh ubconfc < UnicodeData.txt > WCsubst.c
+    sh ubconfc 12.0.0 < UnicodeData.txt > WCsubst.c
 
 where UnicodeData.txt came from
 
-    https://www.unicode.org/Public/12.0.0/ucd/UnicodeData-12.0.0d4.txt
+    https://www.unicode.org/Public/12.0.0/ucd/UnicodeData.txt
+
+Don't forget to mention the update in the User's Guide.
diff --git a/libraries/base/cbits/WCsubst.c b/libraries/base/cbits/WCsubst.c
index 9940405a8e..aa58dc244c 100644
--- a/libraries/base/cbits/WCsubst.c
+++ b/libraries/base/cbits/WCsubst.c
@@ -1,6 +1,6 @@
 /*-------------------------------------------------------------------------
 This is an automatically generated file: do not edit
-Generated by ubconfc at Tue Aug 14 10:04:18 UTC 2018
+Generated by ubconfc at Mon Feb 10 11:42:08 EST 2020
 @generated
 -------------------------------------------------------------------------*/
 
@@ -90,7 +90,7 @@ struct _charblock_
 #define GENCAT_MN 2097152
 #define GENCAT_LO 16384
 #define MAX_UNI_CHAR 1114109
-#define NUM_BLOCKS 3349
+#define NUM_BLOCKS 3352
 #define NUM_CONVBLOCKS 1326
 #define NUM_SPACEBLOCKS 7
 #define NUM_LAT1BLOCKS 63
@@ -1485,7 +1485,8 @@ static const struct _charblock_ allchars[]={
 	{5112, 6, &rule110},
 	{5120, 1, &rule7},
 	{5121, 620, &rule14},
-	{5741, 2, &rule2},
+	{5741, 1, &rule13},
+	{5742, 1, &rule2},
 	{5743, 17, &rule14},
 	{5760, 1, &rule1},
 	{5761, 26, &rule14},
@@ -2799,8 +2800,8 @@ static const struct _charblock_ allchars[]={
 	{43444, 2, &rule124},
 	{43446, 4, &rule92},
 	{43450, 2, &rule124},
-	{43452, 1, &rule92},
-	{43453, 4, &rule124},
+	{43452, 2, &rule92},
+	{43454, 3, &rule124},
 	{43457, 13, &rule2},
 	{43471, 1, &rule91},
 	{43472, 10, &rule8},
@@ -3302,14 +3303,14 @@ static const struct _charblock_ allchars[]={
 	{71935, 1, &rule14},
 	{72096, 8, &rule14},
 	{72106, 39, &rule14},
-	{72145, 1, &rule124},
-	{72146, 1, &rule92},
-	{72147, 1, &rule124},
+	{72145, 3, &rule124},
 	{72148, 4, &rule92},
 	{72154, 2, &rule92},
 	{72156, 4, &rule124},
 	{72160, 1, &rule92},
-	{72161, 3, &rule14},
+	{72161, 1, &rule14},
+	{72162, 1, &rule2},
+	{72163, 1, &rule14},
 	{72164, 1, &rule124},
 	{72192, 1, &rule14},
 	{72193, 10, &rule92},
@@ -3545,7 +3546,8 @@ static const struct _charblock_ allchars[]={
 	{123184, 7, &rule92},
 	{123191, 7, &rule91},
 	{123200, 10, &rule8},
-	{123214, 2, &rule14},
+	{123214, 1, &rule14},
+	{123215, 1, &rule13},
 	{123584, 44, &rule14},
 	{123628, 4, &rule92},
 	{123632, 10, &rule8},
@@ -3556,6 +3558,7 @@ static const struct _charblock_ allchars[]={
 	{125184, 34, &rule203},
 	{125218, 34, &rule204},
 	{125252, 7, &rule92},
+	{125259, 1, &rule91},
 	{125264, 10, &rule8},
 	{125278, 2, &rule2},
 	{126065, 59, &rule17},
diff --git a/libraries/base/cbits/ubconfc b/libraries/base/cbits/ubconfc
index 4d325866bb..cd29641c58 100644..100755
--- a/libraries/base/cbits/ubconfc
+++ b/libraries/base/cbits/ubconfc
@@ -17,6 +17,26 @@
 
 #	Output the file header
 
+VERSION="$1"
+if [ -z "$VERSION" ]; then
+  echo "Usage: $0 [unicode version]"
+  exit 1
+fi
+
+# This file is #included from GHC.Unicode and is used to define
+# GHC.Unicode.unicodeVersion.
+cat > $(dirname $0)/../include/UnicodeVersion.h <<EOF
+#if 0
+This is an automatically generated file: do not edit
+Generated by `basename $0` at `date`
+@generated
+#endif
+
+#define UNICODE_VERSION_NUMS [$(echo $VERSION | sed 's/\./,/g')]
+EOF
+
+exec > $(dirname $0)/WCsubst.c
+
 echo "/*-------------------------------------------------------------------------"
 echo "This is an automatically generated file: do not edit"
 echo "Generated by `basename $0` at `date`"
diff --git a/libraries/base/include/UnicodeVersion.h b/libraries/base/include/UnicodeVersion.h
new file mode 100644
index 0000000000..14852682ac
--- /dev/null
+++ b/libraries/base/include/UnicodeVersion.h
@@ -0,0 +1,7 @@
+#if 0
+This is an automatically generated file: do not edit
+Generated by ubconfc at Mon Feb 10 11:42:08 EST 2020
+@generated
+#endif
+
+#define UNICODE_VERSION_NUMS [12,0,0]