The Big Programming Thread - Page 323

On July 15 2013 15:10 Iggyhopper wrote:
NOT w3schools.

JavaScript-
Interactive: http://www.codecademy.com/
Book/Resource: http://eloquentjavascript.net/

On July 15 2013 14:54 Release wrote:
What are good books/websites from which to learn html/css/javascript?

On July 17 2013 06:38 Arnstein wrote:
Just wondering, what's wrong with w3schools?

On July 18 2013 10:21 holdthephone wrote:
Alright, C++:

Have a text input file, want to read it into char array, but size of file is variable (could be 8 chars, 64, 512, etc...). How can I accommodate for the right amount of characters, besides simply creating an ultra huge char array?

On July 17 2013 06:30 HardlyNever wrote:
Anyone have experience with this JQuery plugin ?

http://jqueryvalidation.org/

I'm having trouble getting it to work on my form. Or anyone recommend another jquery plugin to validate my forms (they don't have to be too robust).

On July 18 2013 10:40 Pawsom wrote:


Vectors if its C++

On July 18 2013 18:09 Tobberoth wrote:

This would be my advice as well. You can use new (or malloc though if it's C++ I don't see why you would), but in modern C++, using vectors would probably be considered best practice.

#include <fstream>
#include <vector>
#include <boost/spirit/include/qi.hpp>

    namespace spirit = boost::spirit;
    namespace qi = boost::spirit::qi;
    std::vector<char> data;
    qi::parse(spirit::istream_iterator(std::ifstream("input.txt") >> std::noskipws), spirit::istream_iterator(), *qi::char_, data)

On July 19 2013 07:40 nunez wrote:


use vector!
don't use new unless you can't not heap it.

if you are also looking to parse the files contents check out boost spirit qi! excellent parser generator library for cpp.

example parsing file into vector of chars... look how cute.

#include <fstream>
#include <vector>
#include <boost/spirit/include/qi.hpp>

    namespace spirit = boost::spirit;
    namespace qi = boost::spirit::qi;
    std::vector<char> data;
    qi::parse(spirit::istream_iterator(std::ifstream("input.txt") >> std::noskipws), spirit::istream_iterator(), *qi::char_, data)

On July 19 2013 08:08 RoyGBiv_13 wrote:


A more generic question would be: "how do I keep a variable amount of memory reserved for X purpose". The more generic answer is: "on the heap, on the stack, or within ROM (a file on the file system, typically) [and secret option D, manually defined memory]"

You typically don't want to use the stack for unbounded amounts of memory, as it can cause a stack overflow, which could be a security hole. This is typically why you don't have unbounded recursion either.

In this case, you are trying to buffer a variable amount of memory in RAM for processing data typically stored in ROM. This would mean you will dynamically request memory from the heap. Boost's Array, std's Vector, and C's malloc library all use heap.

Typical OS behaviour states that your heap can continue to grow as long as you have pages left for the kernel to assign, so it is bounded by the amount of RAM on your system. Because the malloc library uses operating system hooks, it does not have (major) security issues like the stack does. Vector and Array are pretty good at heap management, trying to keep your data in a contiguous virtual block if possible, but calling malloc directly also has it's benefits. The keyword "new" is a wrapper around malloc that also calls the constructor for that class.

The third option would be to directly process the file in ROM without buffering the entire thing into RAM. This is a strategy used by many of the GNU utilities, as streaming data does not need a variable size buffer. Probably the fastest way, but definitely overkill in most applications.

[secret option D is just assigning a pointer to some place in memory, clearing it and using it, but why do this whenever such smart people have already build a malloc library?]