Proper fix for #380 -- use different algorithm to generate codepages

source code which allows to take gaps into account thus making unneccessary
extended [index] = value, syntax in to_ucs2 array
(This used to be commit 4ab08a4bb8dbd023eb85d16a594d3c4ed066f9ac)

2 files changed, 19 insertions, 3 deletions

diff --git a/source3/script/gen-8bit-gap.awk b/source3/script/gen-8bit-gap.awk
new file mode 100644
index 00000000000..59a1a23be07
--- /dev/null
+++ b/source3/script/gen-8bit-gap.awk
@@ -0,0 +1,18 @@
+BEGIN {
+	for (i=0; i<256; i++) {
+		tbl[sprintf("%02x",i)] = "0x0000";
+	}
+}
+
+/^<U([[:xdigit:]][[:xdigit:]][[:xdigit:]][[:xdigit:]])>[[:space:]]*.x([[:xdigit:]][[:xdigit:]])[:space:]*.*$/ {
+	tbl[substr($2,3,2)]=sprintf("0x%s",substr($1,3,4));
+}
+
+END {
+	for(i=0; i<32; i++) {
+		for(j=0; j<8; j++) {
+			printf(" %s,", tbl[sprintf("%02x",i*8+j)]);
+		}
+		printf "\n"
+	}
+}
\ No newline at end of file
diff --git a/source3/script/gen-8bit-gap.sh.in b/source3/script/gen-8bit-gap.sh.in
index e66c654c22d..bcf64a4464f 100755
--- a/source3/script/gen-8bit-gap.sh.in
+++ b/source3/script/gen-8bit-gap.sh.in
@@ -31,9 +31,7 @@ echo " */"
 echo '#include "includes.h"'
 echo
 echo "static const uint16 to_ucs2[256] = {"
-sed -ne '/^[^[:space:]]*[[:space:]]*.x00/d' \
-    -e 's/^<U\(....\)>[[:space:]]*.x\(..\).*/  [0x\2] = 0x\1,/p' \
-    "$CHARMAP" | sort -u
+cat "$CHARMAP" | @AWK@ -f @srcdir@/script/gen-8bit-gap.awk
 echo "};"
 echo
 echo "static const struct charset_gap_table from_idx[] = {"