trunk/src/uri.cpp

/*
 * Copyright (C) 2010-2012 TSUBAKIMOTO Hiroya <z0rac@users.sourceforge.jp>
 *
 * This software comes with ABSOLUTELY NO WARRANTY; for details of
 * the license terms, see the LICENSE.txt file included with the program.
 */
#include "mailbox.h"
#include "win32.h"
#include <cassert>
#include <cstdlib>
#include <cstring>

#ifdef _DEBUG
#include <iostream>
#define DBG(s) s
#define LOG(s) (cout << s)
#else
#define DBG(s)
#define LOG(s)
#endif

/*
 * Functions of the class uri
 */
namespace {
  string::size_type find(const string& s,
                         string::value_type c, string::size_type i = 0)
  {
    while (i < s.size()) {
      if (s[i] == c) return i;
      i += (i + 1 < s.size() && IsDBCSLeadByte(BYTE(s[i]))) + 1;
    }
    return string::npos;
  }

  string decode(const string& s)
  {
    string result;
    string::size_type i = 0;
    while (i < s.size()) {
      string::size_type n = find(s, '%', i);
      if (n == string::npos || s.size() - n < 3) break;
      result.append(s, i, n - i);
      char hex[3] = { s[n + 1], s[n + 2] };
      char* e;
      char c = char(strtoul(hex, &e, 16));
      i = n + (*e ? (c = '%', 1) : 3);
      result.push_back(c);
    }
    if (i < s.size()) result.append(s.c_str() + i);
    return result;
  }

  string encode(const string& s, const char* chs)
  {
    string result;
    string::size_type i = 0;
    while (i < s.size()) {
      const char* const t = s.c_str();
      const char* p = t + i;
      while (BYTE(*p) > 32) {
        string::size_type n =
          IsDBCSLeadByte(BYTE(*p)) ? (p[1] != 0) * 2 : !strchr(chs, *p);
        p += n;
        if (n == 0) break;
      }
      if (!*p) break;
      string::size_type n = p - t;
      result.append(s, i, n - i);
      char hex[4] = { '%' };
      _ltoa(s[n] & 255, hex + 1, 16);
      result.append(hex);
      i = n + 1;
    }
    if (i < s.size()) result.append(s.c_str() + i);
    return result;
  }
}

uri::uri(const string& uri)
{
  string::size_type i = find(uri, '#');
  if (i != string::npos) _part[fragment] = decode(uri.substr(i + 1));
  
  string t(uri, 0, min(find(uri, '?'), i));
  i = find(t, ':');
  for (; i != string::npos; i = find(t, ':', i)) {
    if (++i == t.size() || t[i] != '/') continue;
    if (++i == t.size() || t[i] != '/') continue;
    _part[scheme] = decode(t.substr(0, i - 2));
    t.erase(0, i + 1);
    break;
  }
  i = find(t, '/');
  if (i != string::npos) {
    _part[path] = decode(t.substr(i + 1));
    t.erase(i);
  }
  i = find(t, '@');
  if (i != string::npos) {
    _part[user] = decode(t.substr(0, min(find(t, ';'), i)));
    t.erase(0, i + 1);
  }
  i = find(t, ':');
  for (; i != string::npos; i = find(t, ':', i)) {
    string::size_type n = i;
    while (++i < t.size() && strchr("0123456789", t[i])) continue;
    if (i < t.size()) continue;
    _part[port] = t.substr(n + 1);
    t.erase(n);
  }
  _part[host] = decode(t);
}

uri::operator string() const
{
  string uri = encode(_part[scheme], "#%") + "://";
  if (!_part[host].empty()) {
    if (!_part[user].empty()) uri += encode(_part[user], "@/#%") + '@';
    const string& h = _part[host];
    uri += encode(h, *h.begin() == '[' && *h.rbegin() == ']' ? "/#%" : ":/#%");
    if (!_part[port].empty()) uri += ':' + _part[port];
  }
  uri += '/' + encode(_part[path], "#%");
  if (!_part[fragment].empty()) uri += '#' + encode(_part[fragment], "%");
  return uri;
}