import com.google.common.cache.LoadingCache;

import jakarta.annotation.PostConstruct;
import jakarta.annotation.PreDestroy;

/**
 * Helper class for managing crawler statistics and performance metrics.
 * This class provides functionality to track, record, and report statistics
 * about crawler operations including timing data, performance metrics, and
 * operational events. It uses an internal cache to maintain statistics

 */
package org.codelibs.fess.crawler.rule.impl;

import java.util.ArrayList;
import java.util.List;

import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.rule.Rule;
import org.codelibs.fess.crawler.rule.RuleManager;

/**
 * Implementation of the {@link RuleManager} interface.
 * Manages a list of rules for the crawler.

import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.entity.ResultData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />
	<include path="crawler/transformer.xml" />

	<component name="ruleManager" class="org.codelibs.fess.crawler.rule.impl.RuleManagerImpl" instance="prototype">
		<postConstruct name="addRule">
			<arg>sitemapsRule</arg>
		</postConstruct>
		<postConstruct name="addRule">

import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.builder.RequestDataBuilder;
import org.codelibs.fess.crawler.client.AbstractCrawlerClient;
import org.codelibs.fess.crawler.client.AccessTimeoutTarget;
import org.codelibs.fess.crawler.entity.RequestData;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.exception.ChildUrlsException;

{"index":{"_index":"fess_config.scheduled_job","_id":"default_crawler"}}
{"name":"Default Crawler","target":"all","cronExpression":"0 0 * * *","scriptType":"groovy","scriptData":"return container.getComponent(\"crawlJob\").logLevel(\"info\").gcLogging().execute(executor);","jobLogging":true,"crawler":true,"available":true,"sortOrder":1,"createdBy":"system","createdTime":0,"updatedBy":"system","updatedTime":0}
{"index":{"_index":"fess_config.scheduled_job","_id":"suggest_indexer"}}

User-agent: BruteBot
Disallow: /
Allow: /foo/bar/
Crawl-delay: 1314000

# welcome!
User-agent: Googlebot
Crawl-delay: 1

User-agent: *
Disallow: /private/
Disallow: /help        # disallows /help.html, /help/index.html, etc.
Allow: /help/faq.html
Crawl-delay: 3

User-agent: Crawler
Disallow: /aaa

User-agent: Crawler/1.0
Disallow: /bbb

User-agent: Crawler/2.0
Disallow: /ccc

User-agent: Hoge Crawler

 */
package org.codelibs.fess.crawler.service;

import java.util.List;
import java.util.Map;

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.fess.crawler.entity.OpenSearchUrlQueue;
import org.codelibs.fess.crawler.service.impl.OpenSearchUrlQueueService;
import org.codelibs.fess.crawler.util.OpenSearchCrawlerConfig;

<components namespace="fessCrawler">
    <include path="crawler/container.xml"/>
    <include path="crawler/client.xml"/>
    <include path="crawler/rule.xml"/>
    <include path="crawler/filter.xml"/>
    <include path="crawler/interval.xml"/>
    <include path="crawler/extractor.xml"/>
    <include path="crawler/mimetype.xml"/>
    <include path="crawler/encoding.xml"/>
    <include path="crawler/urlconverter.xml"/>
    <include path="crawler/log.xml"/>

import org.codelibs.fess.app.service.FailureUrlService;
import org.codelibs.fess.crawler.builder.RequestDataBuilder;
import org.codelibs.fess.crawler.client.CrawlerClient;
import org.codelibs.fess.crawler.entity.RequestData;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.entity.UrlQueue;
import org.codelibs.fess.crawler.log.LogType;
import org.codelibs.fess.exception.ContainerNotAvailableException;