// $Id: $

#include "Options.h"
#include "Iterators.h"

URL_Iterator::~URL_Iterator (void)
{
}

int
URL_Iterator::destroy (void)
{
  // Commit suicide.
  delete this;
  return 0;
}

HTML_Body_Iterator::HTML_Body_Iterator (URL &url)
  : url_ (url)
{
}

int
HTML_Body_Iterator::next (ACE_CString &url)
{
  if (this->url_.stream ().eof () == 0)
    {
      // Initialize this to the NUL-string.
      url = ACE_CString ("");

      char *s;

      size_t href_begin_len = ACE_OS::strlen ("A HREF=\"");
      size_t http_len = ACE_OS::strlen ("http://");
      size_t mailto_len = ACE_OS::strlen ("mailto:");
      size_t href_end_len = ACE_OS::strlen ("/A>");

      for (char c;
           (c = this->url_.stream ().get_char ()) != EOF;
           )
        {
          // We need to skip over tokens like "mailto".
          int skip = 0; 
          const char *s;

          if (c == '<' 
              && (s = this->url_.stream ().peek_str (0, href_begin_len)) != 0
              && ACE_OS::strncasecmp (s,
                                      "A HREF=\"",
                                      href_begin_len) == 0)
            {
              this->url_.stream ().seek (href_begin_len);

              if ((s = this->url_.stream ().peek_str (0, mailto_len)) == 0)
                return 0;
              else if (ACE_OS::strncasecmp (s,
                                            "mailto:",
                                            mailto_len) == 0)
                {
                  this->url_.stream ().seek (mailto_len);
                  skip = 1;
                }
              else
                {
                  if ((s = this->url_.stream ().peek_str (0, http_len)) == 0)
                    return 0;
                  else if (ACE_OS::strncasecmp (s,
                                                "http://",
                                                http_len) == 0)
                    // Skip over the "http://".
                    this->url_.stream ().seek (http_len);
                  else
                    {
                      // If we're not prefixed with "http://" assume
                      // that we're a "relative" URL, so let's prepend
                      // our current server's hostname and the dirname
                      // portion of our URI onto this.
                      url += ACE_CString (this->url_.url_addr ().get_host_name ());

                      if (this->url_.stream ().peek_char (0) != '/')
                        {
                          const char *dir =
                            ACE::dirname (this->url_.url_addr ().get_path_name ());
                          if (ACE_OS::strncmp (s,
                                               "../",
                                               3) == 0)
                            {
                              // Remove a level of the dir.
                              dir = ACE::dirname (dir);
                              this->url_.stream ().seek (3);
                            }
              
                          if (dir[0] != '/')
                            url += ACE_CString ("/");
                          if (ACE_OS::strcmp (dir, "/") == 0)
                            url += ACE_CString ("/");                            
                          else
                            {
                              url += ACE_CString (dir);
                              if (url[url.length ()] != '/')
                                url += ACE_CString ("/");
                            }
                        }
                    }
                }

              // Find the end of the URL.
              while ((c = this->url_.stream ().get_char ()) != EOF
                     && c != '\"')
                continue;

              if (c == EOF)
                ACE_ERROR_RETURN ((LM_ERROR,
                                   "unexpected EOF\n"),
                                   0);

              if (skip == 0 && this->url_.stream ().recv_len () > 0)
                url += ACE_CString (this->url_.stream ().recv (),
                                    this->url_.stream ().recv_len () - 1);
              
              // Skip to the end of this "</A>" terminator.
              while ((c = this->url_.stream ().get_char ()) != EOF)
                {
                  if (c == '<'
                      && (s = this->url_.stream ().peek_str (0, href_end_len)) != 0
                      && ACE_OS::strncmp (s,
                                          "/A>",
                                          href_end_len) == 0)
                    {
                      this->url_.stream ().seek (href_end_len);
                      break;
                    }
                }

              if (skip)
                continue;
              else
                return 1;
            }
        }

      return 0;
    }
  else
    return 0;
}

HTTP_Header_Iterator::HTTP_Header_Iterator (URL &url)
  : url_ (url),
    end_of_header_ (0)
{
}

int
HTTP_Header_Iterator::next (ACE_CString &line)
{
  if (this->end_of_header_)
    return 0;
  else
    {
      for (char c;
           (c = this->url_.stream ().get_char ()) != EOF;
           )
        {
          // Check to see if we're at the end of the header line.
          if (c == '\r' && this->url_.stream ().peek_char (0) == '\n')
            {
              line.set (this->url_.stream ().recv (),
                        this->url_.stream ().recv_len () - 1,
                        1);

              // Check to see if we're at the end of the header.
              if (this->url_.stream ().peek_char (1) == '\r'
                  && this->url_.stream ().peek_char (2) == '\n')
                {
                  this->end_of_header_ = 1;
                  // We're at the end of the header section.
                  this->url_.stream ().seek (3);
                }
              else
                // We're at the end of the line.
                this->url_.stream ().seek (1);

              return 1;
            }
          // Handle broken Web servers that use '\n' instead of
          // '\r\n'.
          else if (c == '\n')
            {
              line.set (this->url_.stream ().recv (),
                        (this->url_.stream ().recv_len ()),
                        1);

              // Check to see if we're at the end of the header.
              if (this->url_.stream ().peek_char (0) == '\n')
                {
                  // We're at the end of the header section.
                  this->url_.stream ().seek (1);
                  this->end_of_header_ = 1;
                }

              return 1;
            }

        }

      ACE_ASSERT (!"yikes, found EOF while in header!");
    }
  return 0;
}

URL_Download_Iterator::URL_Download_Iterator (URL &url)
  : url_ (url)
{
}

int
URL_Download_Iterator::next (ACE_CString &buffer)
{
  size_t len = BUFSIZ;

  const char *buf = this->url_.stream ().recv (len);

  if (buf == 0)
    return 0;
  else
    {
      buffer.set (buf, len, 1);
      return 1;
    }
}