Added parse_ini_string, parse_ini_file functions

[pmswga]

At the moment, the functions do not process an array of values.

Example:

[Section3.1]
var1[]=val_1_1
var1[]=val_1_2
var1[]=val_1_3

This section will skip.

[quasilyte]

This is a first review round. I'm sure @DrDet will find something else here, but let's start with something simple.

[DrDet]

In general, I'm not sure it's correct way of parsing ini string. For example, PHP use autogenerated parser for that. Can you explain your parsing method somehow?

[pmswga]

general, I'm not sure it's correct way of parsing ini string. For example, PHP use autogenerated parser for that. Can you explain your parsing method somehow?

I didn't use the automatically generated parsers, because:

1. I didn't think about it
2. I'm not sure that I could correctly implement it in runtime

Therefore, the solution that came to my mind is to check the resulting strings by regular expressions.

[Tsygankov-Slava]

I have found several places where something can be improved. It would be great if you took these comments into attention.

[Tsygankov-Slava]

In this case, it's better to use enum.

enum qoutes_type {
    SINGLE_QOUTES,
    DOUBLE_QUOTES,
    ALL_QUOTES
} qoutes_type_t;

[Tsygankov-Slava]

The variable named flag contains little information about what exactly is there, I think it's better to call this variable quotes_type.

Then the function declaration will look like this:

string clear_quotes(const string &str, qoutes_type_t quotes_type);

[Tsygankov-Slava]

In this case, it's better to take advantage of the opportunities that C++11 standard gives us and use auto.

auto clear_str = f$preg_replace(string("/ +/"), string(" "), str);

[Tsygankov-Slava]

Firstly, also it's better to use auto here.
Secondly, it looks very strange that your for goes to ini_string_copy.size(), suddenly there will be a situation that you turn to ini_string_copy[ini_string_copy.size()]) (for example, if a string of the form [text somehow gets to your function). It's not very good.
It's better to fix on

for (string::size_type i = 0; i < ini_string_copy.size(); ++i) {
...
}

[mkornaukhov]

Type of i is deduced as int, it is not an intention.

[Tsygankov-Slava]

Type of i is deduced as int, it is not an intention.

Thanks, you're right.
Fixed