From 1f7324ea9af0d4df0f3e0567195bec1d743171da Mon Sep 17 00:00:00 2001
From: Robert Pelnar <robert.pelnar@seznam.cz>
Date: Wed, 23 May 2018 20:57:25 +0200
Subject: [PATCH] New ML support - postbuild.sh - initial version

---
 lang/postbuild.sh | 99 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 99 insertions(+)
 create mode 100644 lang/postbuild.sh

diff --git a/lang/postbuild.sh b/lang/postbuild.sh
new file mode 100644
index 000000000..43fe8c8a0
--- /dev/null
+++ b/lang/postbuild.sh
@@ -0,0 +1,99 @@
+#!/bin/sh
+# postbuild.sh
+#
+# Output folder and elf file:
+OUTDIR="../../output"
+OUTELF="$OUTDIR/Firmware.ino.elf"
+#
+# AVR gcc tools used:
+OBJCOPY=C:/arduino-1.6.8/hardware/tools/avr/bin/avr-objcopy.exe
+
+function finish
+{
+ echo
+ echo "postbuild.sh finished... press enter key"
+ read
+ exit
+}
+
+echo "postbuild.sh started..."
+
+#check input files
+echo " checking files:"
+if [ ! -e $OUTDIR ]; then echo "  folder '$OUTDIR' not found!"; finish; fi
+echo "  folder OK"
+if [ ! -e $OUTELF ]; then echo "  elf file '$OUTELF' not found!"; finish; fi
+echo "  elf OK"
+if ! ls $OUTDIR/sketch/*.o >/dev/null 2>&1; then echo "  no object files in '$OUTDIR/sketch/'!"; finish; fi
+echo "  objects OK"
+
+#run progmem.sh to examine content of progmem1
+./progmem.sh 1
+
+#create sorted list of strings from progmem1.var and lang_en.txt
+#lines from progmem1.var will contain addres (8 chars) and english text
+#lines from lang_en.txt will contain linenumber and english text
+#after sort this will generate pairs of lines (line from progmem1 first)
+#result of sort is compiled with simple script and stored to file textaddr.txt
+echo "compiling progmem1.var and lang_en.txt"
+addr=''
+text=''
+(cat progmem1.var | sed -E "s/^([^ ]*) ([^ ]*) (.*)/\1 \"\3\"/";\
+ cat lang_en.txt | sed "/^$/d;/^#/d" | sed = | sed '{N;s/\n/ /}') |\
+ sort -k2 |\
+ sed "s/\\\/\\\\\\\/g" | while read num txt; do
+ if [ ${#num} -eq 8 ]; then
+  if [ -z "$addr" ]; then
+   addr=$num
+  else
+   if [ "$text" == "$txt" ]; then
+    addr="$addr $num"
+   else
+    echo "ADDR NF $addr $text"
+    addr=$num
+   fi
+  fi
+  text=$txt   
+ else
+  if [ -z "$addr" ]; then
+   echo "TEXT NF $num $txt"
+  else
+   if [ "$text" == "$txt" ]; then
+    if [ ${#addr} -eq 8 ]; then
+     echo "ADDR OK $addr $num"
+    else
+     echo "$addr" | sed "s/ /\n/g" | while read ad; do
+      echo "ADDR OK $ad $num"
+     done
+    fi
+    addr=''
+    text=''
+   else
+    echo "TEXT NF $num $txt"
+   fi
+  fi
+ fi
+done > textaddr.txt
+
+#check for messages declared in progmem1, but not found in lang_en.txt
+echo "checking textaddr.txt..."
+if cat textaddr.txt | grep "^ADDR NF"; then echo "Some strings not found in lang_en.txt!"; finish; fi
+
+#update progmem1 id entries in binary file
+echo "extracting binary..."
+$OBJCOPY -I ihex -O binary $OUTDIR/Firmware.ino.hex ./firmware.bin
+
+#update progmem1 id entries in binary file
+echo "updating binary..."
+#dd if=/dev/zero of=test.bin bs=512 count=64 2>/dev/null
+time cat textaddr.txt | grep "^ADDR OK" | cut -f3- -d' ' | sed "s/^0000/0x/" |\
+ awk '{ hi = int($2 / 256); lo = int($2 - 256 * hi); printf("%d \\\\x%02x\\\\x%02x\n", strtonum($1), hi, lo); }' |\
+ while read addr data; do
+  echo -n -e $data | dd of=./firmware.bin bs=1 count=2 seek=$addr conv=notrunc oflag=nonblock 2>/dev/null
+ done
+
+#update progmem1 id entries in binary file
+echo "converting to hex..."
+$OBJCOPY -I binary -O ihex ./firmware.bin ./firmware.hex
+
+finish