#!/bin/bash
# Generates an extractor that saves C-array encoded data to a file to assist with extracting loadable, binary firmware from within drivers.
#
# Copyright (C) 2022 DiffieHellman
#
# This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License version 3 for more details.

#return invalid usage if there isn't 2 args
[ "$#" -eq 1 ] || { echo "Usage: ./generate_extractor.sh <input_file>"; exit 2; }

#make the script automatically exit if any subcommand fails
#set -euo pipefail
set -euxo pipefail

input="$1"
output="output.c"
squeezed_input="tmp.c"

#create head of C extractor with {u,s}{8,16,32} mappings
echo "/*
Copyright (C) 2022 DiffieHellman
C-array to file extractor. No copyright claims are made on C-array data that is to be extracted.

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License version 3 for more details.
*/

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <stdint.h>

typedef uint8_t u1Byte;
typedef uint32_t u4Byte;
typedef uint8_t u8;
typedef uint16_t u16;
typedef uint32_t u32;

typedef int8_t s8;
typedef int16_t s16;
typedef int32_t s32;
" > "$output"

#convert all tabs, newlines and line feeds in the file into spaces and squeeze any repeated spaces to simplify regex
tr '\t' ' ' < "$input" | tr '\n' ' ' | tr '\r' ' ' | tr --squeeze-repeats ' ' > "$squeezed_input"

#extract C arrays while also preventing arbitrary code execution. grep in -z mode outputs \0 at the end of each match, which is changed to 2 newlines. sadly no UTF-8 support
grep -ozE "((|const )(u|s)(8|16|32|64|1Byte)|(|unsigned )((|u)int(|_fast)(|8_t|16_t|32_t)|char|short|long|long long)) "\
"[[:alnum:]_]+ ?\[[A-Za-z ]*\]"\
"[A-Za-z_ ]*= ?"\
"{[[:space:][:digit:]A-Fa-f,x.]+} ?;" "$squeezed_input" | sed 's/\x0/\n\n/g' >> "$output"

#get names of all arrays previously extracted
names=$(grep '\[' "$output" | sed 's/ *\[.*//' | sed 's/.* //')

#todo: handle automatically outputting arrays like: u8 long_winded_name[LongWindedName] = {

#array to file function
echo 'void openWriteClose(char filename[], char array[], int size)
        {
	FILE *fp = fopen(filename, "w");
        if (!fp){perror("Could not open file"); exit(1);}
        ssize_t result = write(fileno(fp), array, size);
        if (!result){perror("Failed to write file"); exit(1);}
        fclose(fp);
        }

int main(void)
	{' >> "$output"

#convert space separted string to array
conv_array=($names)
#loop over array
for name in ${conv_array[@]}
do
	echo "	openWriteClose(\"${name}\", ${name}, sizeof(${name}));" >> "$output"
done

echo "
	return 0;
	}" >> "$output"

rm "$squeezed_input"

#compile and run the extractor
gcc "$output" -o extractor && ./extractor && rm extractor && rm "$output"
grep '\[' "$input"