csv_reader.cxx

#include "csv_reader.h" 
#include <cgv/utils/file.h>
 
namespace cgv {
    namespace utils {
 
const char* csv_reader_base::get_ws() const 
{
    return ((flags & CSV_SPACE) != 0) ? " " : ""; 
}
const char* csv_reader_base::get_sep() const 
{
    if ((flags & CSV_COMMA) != 0) {
        if ((flags & CSV_TAB) != 0)
            return ",\t";
        else
            return ",";
    }
    else
        if ((flags & CSV_TAB) != 0)
            return "\t";
    return "";
}
bool csv_reader_base::has_heading() const 
{
    return (flags & CSV_HEADING) != 0; 
}
bool csv_reader_base::fail() const 
{
    return failed;
}
bool csv_reader_base::find_column_index(const std::string& col_name, size_t& col_index) const
{
    auto iter = std::find(col_names.begin(), col_names.end(), col_name);
    if (iter == col_names.end())
        return false;
    col_index = unsigned(iter - col_names.begin());
    return true;
}
csv_reader_base::csv_reader_base(const std::string& file_name, CSV_Flags _flags)
{
    flags = _flags;
    if (!cgv::utils::file::read(file_name, content, true)) {
        last_error = "cannot read <" + file_name + ">";
        failed = true;
        return;
    }
    cgv::utils::split_to_lines(content, lines, true);
    while (li < lines.size()) {
        if (!lines[li].empty())
            break;
        ++li;
    }
    if (li >= lines.size()) {
        last_error = "all lines in csv file are empty";
        failed = true;
        return;
    }
    if (has_heading()) {
        std::vector<cgv::utils::token> tokens;
        cgv::utils::split_to_tokens(lines[li], tokens, "", false, "", "", get_sep());
        for (auto tok : tokens)
            col_names.push_back(cgv::utils::to_string(tok));
        ++li;
        nr_cols = col_names.size();
    }
}
 
bool csv_reader_base::parse_next_line(std::vector<cgv::utils::token>& tokens) const
{
    while (li < lines.size()) {
        if (lines[li].empty()) {
            ++li;
            continue;
        }
        // split with seperator extration to support empty entries 
        std::vector<cgv::utils::token> sep_tokens;
        cgv::utils::split_to_tokens(lines[li], sep_tokens, get_sep(), false, "'\"", "'\"", get_ws());
        // filter out separators
        tokens.clear();
        bool last_was_empty = true;
        for (auto tok : sep_tokens) {
            if (tok == "," || tok == "\t") {
                if (last_was_empty)
                    tokens.push_back(cgv::utils::token(tok.begin, tok.begin));
                else
                    last_was_empty = true;
            }
            else {
                tokens.push_back(tok);
                last_was_empty = false;
            }
        }
        if (nr_cols == 0)
            nr_cols = tokens.size();
        else
            if (tokens.size() < col_names.size()) {
                ++li;
                continue;
            }
        ++li;
        return true;
    }
    return false;
}
 
    }
}