#include <x/csv.H> for (size_t row=0; row<3; row++) { std::vector<std::string> v; for (size_t col=0; col<5; col++) { std::ostringstream o; o << row*10 + col; v.push_back(o.str()); } x::tocsv(v.begin(), v.end(), std::ostreambuf_iterator<char>(std::cout)); std::cout << std::endl; }
The first two parameters to
x::tocsv()
are a beginning and and ending input iterator which iterate over strings.
x::tocsv() formats this input sequence using
a "comma-separated-value" format.
The strings may contain any characters, including newlines. Therefore,
parsing a "comma-separated-value" format cannot be
done one line at a time. A character input iterator is required with
x::fromcsv():
#include <x/csv.H> std::istream &i; // ... std::istreambuf_iterator<char> b(i), e; while (b != e) { std::cout << "Line: " << std::endl; std::vector<std::string> rows; b=x::fromcsv(b, e, rows); if (b == e || *b++ != '\n') { std::cerr << "Formatting error" << std::endl; } else { for (std::vector<std::string>::const_iterator b=rows.begin(), e=rows.end(); b != e; ++b) { std::cout << " Col: " << *b << std::endl; } } }
This example shows the proper way to parse "comma-separated-values":
Define a beginning and an ending input iterator to the formatted character input data.
Until the iterators compare equal, invoke
x::fromcsv(), which parses a single
"CSV" formatted list of string values from the character input.
After x::fromcsv() returns, the updated
input iterator is expected to point to the newline sequence that
separated the CSV lists. Otherwise it's a formatting error.
#include <x/csv.H> std::istream &i; // ... std::map<std::string, size_t> headers; std::istreambuf_iterator<char> b(i), e; b=x::csvheaderrow(b, e, headers); if (b == e || *b++ != '\n') { std::cerr << "Formatting error" << std::endl; } else { // ...
x::csvheaderrow() uses
x::fromcsv() to parse a CSV header line,
placing the column names into a map keyed by the column name, with the
value being the 0-based column number.