focusedspider.sql

一个java+Mysql写的主题爬虫,可以通过给定初始的url爬取到相关的网页

# MySQL-Front 5.0  (Build 1.126)

/*!40101 SET @OLD_SQL_MODE=@@SQL_MODE */;
/*!40101 SET SQL_MODE='STRICT_TRANS_TABLES,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION' */;
/*!40111 SET @OLD_SQL_NOTES=@@SQL_NOTES */;
/*!40103 SET SQL_NOTES='ON' */;
/*!40014 SET @OLD_UNIQUE_CHECKS=@@UNIQUE_CHECKS */;
/*!40014 SET UNIQUE_CHECKS=0 */;
/*!40014 SET @OLD_FOREIGN_KEY_CHECKS=@@FOREIGN_KEY_CHECKS */;
/*!40014 SET FOREIGN_KEY_CHECKS=0 */;


# Host: localhost    Database: focusedspider
# ------------------------------------------------------
# Server version 5.0.67-community-nt

DROP DATABASE IF EXISTS `focusedspider`;
CREATE DATABASE `focusedspider` /*!40100 DEFAULT CHARACTER SET latin1 */;
USE `focusedspider`;

#
# Table structure for table complete_queue
#

CREATE TABLE `complete_queue` (
  `id` int(11) NOT NULL auto_increment,
  `url` varchar(1024) NOT NULL default '',
  `relativity` float default '0',
  `priority` float default '0',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table complete_queue
#
LOCK TABLES `complete_queue` WRITE;
/*!40000 ALTER TABLE `complete_queue` DISABLE KEYS */;

/*!40000 ALTER TABLE `complete_queue` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table crawledurl
#

CREATE TABLE `crawledurl` (
  `id` int(11) NOT NULL auto_increment,
  `encodeURL` varchar(1500) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table crawledurl
#
LOCK TABLES `crawledurl` WRITE;
/*!40000 ALTER TABLE `crawledurl` DISABLE KEYS */;

/*!40000 ALTER TABLE `crawledurl` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table discard_queue
#

CREATE TABLE `discard_queue` (
  `id` int(11) NOT NULL auto_increment,
  `relativity` float NOT NULL default '0',
  `url` varchar(1024) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table discard_queue
#
LOCK TABLES `discard_queue` WRITE;
/*!40000 ALTER TABLE `discard_queue` DISABLE KEYS */;

/*!40000 ALTER TABLE `discard_queue` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table error_queue
#

CREATE TABLE `error_queue` (
  `id` int(11) NOT NULL auto_increment,
  `url` varchar(1024) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table error_queue
#
LOCK TABLES `error_queue` WRITE;
/*!40000 ALTER TABLE `error_queue` DISABLE KEYS */;

/*!40000 ALTER TABLE `error_queue` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table keywords
#

CREATE TABLE `keywords` (
  `word` varchar(20) NOT NULL default '',
  `weight` float NOT NULL default '0'
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table keywords
#
LOCK TABLES `keywords` WRITE;
/*!40000 ALTER TABLE `keywords` DISABLE KEYS */;

/*!40000 ALTER TABLE `keywords` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table processing_queue
#

CREATE TABLE `processing_queue` (
  `id` int(11) NOT NULL auto_increment,
  `url` varchar(1024) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table processing_queue
#
LOCK TABLES `processing_queue` WRITE;
/*!40000 ALTER TABLE `processing_queue` DISABLE KEYS */;

/*!40000 ALTER TABLE `processing_queue` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table seeds
#

CREATE TABLE `seeds` (
  `id` int(11) NOT NULL auto_increment,
  `url` varchar(1024) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table seeds
#
LOCK TABLES `seeds` WRITE;
/*!40000 ALTER TABLE `seeds` DISABLE KEYS */;

/*!40000 ALTER TABLE `seeds` ENABLE KEYS */;
UNLOCK TABLES;

#
# Table structure for table waiting_queue
#

CREATE TABLE `waiting_queue` (
  `id` int(11) NOT NULL auto_increment,
  `url` varchar(1024) NOT NULL default '',
  PRIMARY KEY  (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

#
# Dumping data for table waiting_queue
#
LOCK TABLES `waiting_queue` WRITE;
/*!40000 ALTER TABLE `waiting_queue` DISABLE KEYS */;

/*!40000 ALTER TABLE `waiting_queue` ENABLE KEYS */;
UNLOCK TABLES;

/*!40014 SET FOREIGN_KEY_CHECKS=@OLD_FOREIGN_KEY_CHECKS */;
/*!40014 SET UNIQUE_CHECKS=@OLD_UNIQUE_CHECKS */;
/*!40111 SET SQL_NOTES=@OLD_SQL_NOTES */;
/*!40101 SET SQL_MODE=@OLD_SQL_MODE */;