txt_filter.cpp (XML toupper)

/* txt_filter: Operates on plaintext/xml files,
 * replacing any <tagnames> to uppercase <TAGNAMES>.
 * Equivalent to running
#!/bin/sh
cat input.xml | \
sed 's@\<\([a-z]\+\)\>@\U\1@g' | \
sed 's@\<\([a-z]\+\)\( \+.*\)\>@\U\1\L\2@g' > output.xml
 */

#include <iostream>
#include <fstream>
#include <string>

using namespace std;

int main()
{
    const string filename = "input.txt";
    const string filename_out = "output.txt";
    ifstream istrm(filename, ios::in);
    if (!istrm.is_open()) {
        cerr << "failed to open " << filename << endl;
        return 1;
    }
    ofstream ostrm(filename_out, ios::out);

    string linebuf, tagname;
    string tag, tag_upper;
    int count_replaced = 0;
//  locale loc = locale();
    while (!istrm.eof()) {
        getline(istrm, linebuf, '\n');
//      cerr << linebuf << endl;

        // Find an opening XML tag: '<', alphabetics [a-zA-Z], '>'
        const size_t pos1 = linebuf.find('<');
        const size_t pos2 = linebuf.find('>');
        if ((pos1 == -1UL) || (pos2 == -1UL) ||
            ( !isalpha(linebuf[pos1+1]) && !(linebuf[pos1+1] == '/') )
        ) {
            ; // No tag, skip filtering
        } else {
            // Found a line containing a tag, ex.
            // '   <ingredient amount="3" unit="spoon">Flour</ingredient>'

            // First pass: find the end of tag name
            const size_t pos_tag_start = pos1 + 1;
            size_t pos_tag_end = pos2;
            tag = linebuf.substr(pos_tag_start, pos_tag_end - pos_tag_start);
            clog << "Parsing tagline: " << tag << endl;
            for (size_t i = 0; i<tag.length(); i++) {
                if ((tag[i] == ' ') || (tag[i] == '\t') || (tag[i] == '\n')) {
                    pos_tag_end = i + pos_tag_start;
                    break;
                }
            }
//          tag_upper = toupper(tag, loc);

            // Second pass: filter only the name
            tagname = linebuf.substr(pos_tag_start, pos_tag_end - pos_tag_start);
            clog << "Found tag: " << tagname << endl;
            tag_upper = tagname; // allocate by copy, then patch
            for (size_t i = 0; i<tag_upper.length(); i++) {
                tag_upper[i] = toup1per(tagname[i]);
            }
            linebuf.replace(pos_tag_start, pos_tag_end - pos_tag_start, tag_upper);
            count_replaced++;
        }
//      cout << linebuf << endl;
        ostrm << linebuf << endl;
    }
    cout << "Replaced " << count_replaced << " tags to uppercase." << endl;
    return 0;
}