import java.util.Vector;
import java.util.StringTokenizer;
import java.net.URL;
import java.net.MalformedURLException;
import java.io.InputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;

/**
 * <H2>Netrand Project</H2>
 * <H4>Software Engineering - CS536</H4>
 * <H4>University of Wisconsin - Milwaukee</H4>
 * <H4>Authors:</H4>
 * <UL>
 * <LI>Spring 1998 - Francis William Kasper
 * </UL>
 * <BR>File: HTMLLinkExtractor.java <BR>
 *  Note: 
 *  This file was originally part of a Web Crawler program written by
 *  Tim Macinta in 1997 that gathered links form the internet and formed
 *  a web search database.  The files containing the logic for crawling
 *  across the internet were taken from this program and slightly modified
 *  for the purpose of the NetRand project.
 *  <HR>
 *  This LinkExtractor can extract URLs from HTML files.
 */

public class HTMLLinkExtractor implements LinkExtractor {
  
  Vector urls = new Vector(6, 9);  // list of URLs
  int next_url = 0;                // next URL to return
  int url_count = 0;               // number of URLs
  URL base = null;                 // base URL

  /** Creates a new HTMLLinkExtractor that will enumerate all the
   *  URLs in the give "cache_file".
   */
  public HTMLLinkExtractor(File cache_file, URL base_url) throws IOException {
    this.base = base_url;
    InputStream in = new FileInputStream(cache_file);
    int state = 0;
    StringBuffer sb = new StringBuffer();
    int i = in.read();
    while (i >= 0) {
      switch (state) {
      case 0:
	if (i == '<') state = '<';
	break;
      case '<':
	if (i == '>') {
	  state = 0;
	  analyze(sb.toString());
	  sb.setLength(0);
	} else {
	  sb.append((char) i);
	}
      }
      i = in.read();
    }
    if (sb.length() > 0) analyze(sb.toString());
    in.close();
  }

  /** Analyzes "param", which should be the contents between a '<' and a '>',
   *  and adds any URLs that are found to the list of URLs.
   */
  public void analyze(String param) {
    StringTokenizer st = new StringTokenizer(param);
    if (st.countTokens() < 2) return;
    String first_word = st.nextToken().toLowerCase();
    if (first_word.equals("a")) {
      analyzeAnchor(st.nextToken(""));
    } else if (first_word.equals("frame")) {
      analyzeFrame(st.nextToken(""));
    } else if (first_word.equals("base")) {
      extractBase(st.nextToken(""));
    } 
  }

  /** Analyzes the <a> tag.  */
  void analyzeAnchor(String anchor) {
    String href = extract(anchor, "href");
    if (href == null) return;
    try {
      addURL(new URL(base, href));
    } catch (MalformedURLException e) {
      anchor = anchor.toLowerCase();
      // java doesn't understand mailto and will throw an exception
      if (!href.startsWith("mailto:")) {
	e.printStackTrace();
      }
    }
  }

  /** Analyzes the <frame> tag.  */
  void analyzeFrame(String frame) {
    String src = extract(frame, "src");
    if (src == null) return;
    try {
      addURL(new URL(base, src));
    } catch (MalformedURLException e) {
      e.printStackTrace();
    }
  }

  /** Extracts the base URL from the <base> tag. */
  void extractBase(String b) {
    String b2 = extract(b, "href");
    if (b2 != null) {
      try {
	base = new URL(base, b2);
      } catch (MalformedURLException e) {
	e.printStackTrace();
      }
    }
  }

  /** Adds "url" to the list of URLs. */
  public void addURL(URL url) {
    urls.addElement(url);
    url_count++;
  }

  public boolean hasMoreElements() {
    return url_count != next_url;
  }

  public Object nextElement() {
    Object ob = urls.elementAt(next_url);
    next_url++;
    return ob;
  }
  
  /** Resets this enumeration.  */
  public void reset() {
    next_url = 0;
  }

  /** Returns the value in "line" associated with "key", or null if "key"
   *  is not found.  For instance, if line were "a href="blah blah blah"
   *  and "key" were "href" this method would return "blah blah blah".
   *  <p>
   *  Keys are case insensitive.
   */
  String extract(String line, String key) {
    try {
      key = key.toLowerCase();
      String lower_case = line.toLowerCase();
      int i = lower_case.indexOf(key);
      if (i < 0) return null;
      i += key.length();
      if (line.charAt(i) != '=') return null;
      i++;
      int i2;
      if (line.charAt(i) == '"') {
	i++;
	i2 = line.indexOf('"', i);
	if (i2 < 0) {
	  return line.substring(i);
	} else {
	  return line.substring(i, i2);
	}
      } else {
	int targ = line.length();
	for (i2 = i; i < targ; i++) {
	  if (Character.isSpace(line.charAt(i))) break;
	}
	return line.substring(i, i2);
      }
    } catch (StringIndexOutOfBoundsException e) {}
    return null;
  }

}