Doc: modify input parameters for symmetry and Exx

[maki49]

• fix Should we set the default value of symmetry to 1? #2498 : set the default value of symmetry to 1 and modify corresponding tests
• symmetry becomes another parameter to achieve An user-friendly way to support "flexible default value" is needed for some parameters #2508 : current solution is to change the variable type of symmetry to string, using "default" and std::stoi to support flexible (not forced) default value.
• modify the description of symmetry param
• modify the description of some exx-params

[dyzheng]

Maybe Availability can be added in front of "### exx_hybrid_alpha"

[Satinelamp]

This is an excellent feature to let ABACUS decide how to set symmetry. But using a string type of symmetry parameter seems not necessary, you could provide a new integer option, e.g. symmetry=3, and use the following code to set symmetry to 0 or 1

    if (symmetry == 3)
    {   //deal with no-forced default value
        if (calculation == "nscf" || calculation == "get_S" || calculation == "istate" || calculation == "ienvelope")
            symmetry = 0;     //if md or exx, symmetry will be force-set to 0 or -1 later
        else
            symmetry = 1;
    }

the usage of std::stoi could be removed so that the code would be concise。
After all, inside the code of ABACUS, the parameter of symmetry is treated as an integer. Besides that, treating the symmetry option as an integer is also similar to VASP and other codes, which would make it more user-friendly.

[maki49]

This is an excellent feature to let ABACUS decide how to set symmetry. But using a string type of symmetry parameter seems not necessary, you could provide a new integer option, e.g. symmetry=3, and use the following code to set symmetry to 0 or 1

    if (symmetry == 3)
    {   //deal with no-forced default value
        if (calculation == "nscf" || calculation == "get_S" || calculation == "istate" || calculation == "ienvelope")
            symmetry = 0;     //if md or exx, symmetry will be force-set to 0 or -1 later
        else
            symmetry = 1;
    }

the usage of std::stoi could be removed so that the code would be concise。 After all, inside the code of ABACUS, the parameter of symmetry is treated as an integer. Besides that, treating the symmetry option as an integer is also similar to VASP and other codes, which would make it more user-friendly.

Thanks for your suggestion.

Currently, there are several parameters with the type of int or real(double) using a string-type initialization with a stoi or stod to support flexible-default like this (and most of them are exx related parameters) .
So setting it as another integer is not a common solution. But I agree that there might be a solution both common and user-friendly.

I think we should merge this PR anyhow (for 1-PR-1-feature principle).

Then I will reopen the issue #2508 and discuss with @PeizeLin for the common and user-friendly solution.