Allow array<char*> (preserves orig functionality)

[helmingstay]

Overview

The idea is to provide additional functionality, while leaving orig functionality in-place. Allowing for the array<char*> simplified my life in writing several methods for a class inheriting from CSVReader.

Commit message

• Add convenience type aliases
• In set_column_names and read_header, new function takes array<char*>
• Add template function to construct array from parameter pack, pass to above

[breznak]

can you explain on a high-level what the additional functionality is?

[helmingstay]

Sorry for slow response. I've been thinking critically about this. I'm relatively new to TMP, and your library has been a nice tool to learn on. And I see now that I can keep the parameter pack intact and hand it off within the class - in particular, to generate a vector to be filled by read_row.

In the end, I used a helper template to feed individual elements of an auto-generated vector into read_row to be filled. In essence, this is the opposite of the pull request - starting with a vector<char *>, and ending with a parameter pack.

Thanks for the nice library! -Christian

On Mon, May 23, 2016 at 5:45 AM, breznak notifications@github.com wrote:

can you explain on a high-level what the additional functionality is?

— You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub https://github.com/ben-strasser/fast-cpp-csv-parser/pull/23#issuecomment-220958179

A man, a plan, a cat, a ham, a yak, a yam, a hat, a canal – Panama!

[ben-strasser]

Hi,

it's nice to see people using the library and thanks for the patch. :)

However, I will not merge the patch because of several reasons:

• Unless I'm mistaken your code only works with const char* parameters as column names. The old code works also with std::string and everything convertible to std::string. This is a breaking interface change.
• Originally the code compiled with GCC 4.6. If I recall correctly, then this is still the case. template aliases however require GCC 4.8. I'm not saying that I never want to switch to GCC 4.8 however, the new feature should be worth it.
• I do not see the advantage of the new code.

Best Regards Ben Strasser