Std 634: Fill & Download for Free

I think the most dangerous line of code in C++ is this:Fred[8472] = 8743872; That looks harmless enough, but if Fred was defined to have, say, 8000 elements, then that line of code is going to attempt to write the integer 8743872 to some unknown location in RAM. And what comes from that could be just about anything.If you’re lucky, the operating system (or the program itself) will recognize that an attempted illegal memory access has occurred, and it will print an error message to your screen and your program will terminate.If you’re unlucky, the integer may be written on top of code that then gets executed, and just about anything could happen. The spurious code could damage critical operating-system files, or email pics of you having sex with your mistress to your wife, or enroll you in an Al Qaeda suicide bomber training camp, or start World War 3, or radio a hideous insult to a passing spaceship full of hostile aliens (who then obliterate Earth), or whatever. Oh my.So don’t do that. If you feel you must use arrays in C or C++, always do careful bounds checking. And whatever you do, do not write programs like this:// over-run-test-cpp.cpp #include <stdio.h> int main (void) {  int Fred[505] = {0};  Fred[634] = 87592;  printf("Fred[634] = %d\n", Fred[634]);  return 0; } On my system, even with heavy warnings flags, my compiler (gcc / g++) compiles that without error or warning! And it runs without error or warning, and gives the desired result! But the source code is committing illegal memory access and commanding over-writing of memory it doesn’t actually own!So, why didn’t that cause a compiler warning and/or run-time error? Because of optimization! The compiler, seeing that I’m not even coming close to filling-up the array, silently changes the index of the “87592” from “634” to “0”.BUT!!! What if someone then alters the program years later? The changes can have devastating consequences that the compiler won’t be able to optimize away:// evil-test-2.cpp #include <cstdio> int main (void) {  int Fred[500];  int i;  for ( i = 0 ; i < 100000 ; ++i )  {  Fred[i] = i;  }  printf("Fred[34927] = %d\n", Fred[34927]);  return 0; } Compilation string:g++ -I /rhe/include -D PLATFORM_IS_WIN64 -Wall -Wextra -Wfloat-equal -Wshadow -Wcast-qual -Wcast-align -Wconversion -Wcomments -Wundef -Wunused-macros -Wold-style-cast -Woverloaded-virtual -finput-charset=UTF-8 -std=gnu++14 -s -O2 evil-test-2.cpp -L/rhe/lib64 -L/lib -L/usr/lib -lm -o /rhe/bin64/test/evil-test-2.exe Compiler warnings issues:evil-test-2.cpp: In function ‘int main()’: evil-test-2.cpp:9:15: warning: iteration 500 invokes undefined behavior [-Waggressive-loop-optimizations]  Fred[i] = i;  ~~~~~~~~^~~ evil-test-2.cpp:7:20: note: within this loop  for ( i = 0 ; i < 100000 ; ++i )  ~~^~~~~~~~ evil-test-2.cpp:11:10: warning: array subscript is above array bounds [-Warray-bounds]  printf("Fred[34927] = %d\n", Fred[34927]);  ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Run-time results:$ evil-test-2 Segmentation fault (core dumped) This time the compiler was not able to optimize-away the problems, and everything blew up.So, do not count on always being able to get away with sloppy code! Not in any programming language! You might be able to get away with in in the short-term for various reasons (such as compiler optimization fixing your bugs for you, as in the example above), but in the long-term, it will come back and bite you, in very bad ways (program crashes, system crashes, hardware damage, data loss, financial losses, company bankruptcy, termination of employment, homelessness, or worse). So don’t do that!In the C++ programming language, specifically, you shouldn’t be using arrays or pointers at all, except when absolutely needed, and only if you know what you’re doing and are using careful and frequent bounds checking. Instead, use the “containers”, “iterators”, and “algorithms” provided by “The Standard Template Library” (“STL” for short). The STL exists to be used, and if you’re writing in C++ you should be using it heavily. Specifically, use it’s “vector” template for most of the things you’d use arrays for in C. And use the “.at()” method on vector objects, instead of [brackets], to invoke automatic bounds checking.