1 /*
2 * Copyright 2012-2025 CodeLibs Project and the Others.
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND,
13 * either express or implied. See the License for the specific language
14 * governing permissions and limitations under the License.
15 */
16 package org.codelibs.fess.crawler.interval;
17
18 import org.apache.logging.log4j.LogManager;
19 import org.apache.logging.log4j.Logger;
20 import org.codelibs.fess.crawler.interval.impl.DefaultIntervalController;
21 import org.codelibs.fess.helper.IntervalControlHelper;
22 import org.codelibs.fess.util.ComponentUtil;
23
24 /**
25 * FessIntervalController extends DefaultIntervalController to provide
26 * Fess-specific interval control functionality for web crawling operations.
27 * This controller manages delays and timing for various crawling states
28 * including processing delays, queue waiting times, and new URL discovery.
29 */
30 public class FessIntervalController extends DefaultIntervalController {
31
32 private static final Logger logger = LogManager.getLogger(FessIntervalController.class);
33
34 /**
35 * Default constructor.
36 */
37 public FessIntervalController() {
38 super();
39 }
40
41 /**
42 * Gets the delay time in milliseconds after processing a URL.
43 *
44 * @return the delay time in milliseconds after processing
45 */
46 public long getDelayMillisAfterProcessing() {
47 return delayMillisAfterProcessing;
48 }
49
50 /**
51 * Sets the delay time in milliseconds after processing a URL.
52 *
53 * @param delayMillisAfterProcessing the delay time in milliseconds after processing
54 */
55 public void setDelayMillisAfterProcessing(final long delayMillisAfterProcessing) {
56 this.delayMillisAfterProcessing = delayMillisAfterProcessing;
57 }
58
59 /**
60 * Gets the delay time in milliseconds when there are no URLs in the queue.
61 *
62 * @return the delay time in milliseconds when no URLs are available
63 */
64 public long getDelayMillisAtNoUrlInQueue() {
65 return delayMillisAtNoUrlInQueue;
66 }
67
68 /**
69 * Sets the delay time in milliseconds when there are no URLs in the queue.
70 *
71 * @param delayMillisAtNoUrlInQueue the delay time in milliseconds when no URLs are available
72 */
73 public void setDelayMillisAtNoUrlInQueue(final long delayMillisAtNoUrlInQueue) {
74 this.delayMillisAtNoUrlInQueue = delayMillisAtNoUrlInQueue;
75 }
76
77 /**
78 * Gets the delay time in milliseconds before processing a URL.
79 *
80 * @return the delay time in milliseconds before processing
81 */
82 public long getDelayMillisBeforeProcessing() {
83 return delayMillisBeforeProcessing;
84 }
85
86 /**
87 * Sets the delay time in milliseconds before processing a URL.
88 *
89 * @param delayMillisBeforeProcessing the delay time in milliseconds before processing
90 */
91 public void setDelayMillisBeforeProcessing(final long delayMillisBeforeProcessing) {
92 this.delayMillisBeforeProcessing = delayMillisBeforeProcessing;
93 }
94
95 /**
96 * Gets the delay time in milliseconds for waiting for new URLs.
97 *
98 * @return the delay time in milliseconds for waiting for new URLs
99 */
100 public long getDelayMillisForWaitingNewUrl() {
101 return delayMillisForWaitingNewUrl;
102 }
103
104 /**
105 * Sets the delay time in milliseconds for waiting for new URLs.
106 *
107 * @param delayMillisForWaitingNewUrl the delay time in milliseconds for waiting for new URLs
108 */
109 public void setDelayMillisForWaitingNewUrl(final long delayMillisForWaitingNewUrl) {
110 this.delayMillisForWaitingNewUrl = delayMillisForWaitingNewUrl;
111 }
112
113 /**
114 * Delays the crawler while waiting for new URLs to be available.
115 * This method calibrates CPU load, checks crawler status, applies
116 * interval control rules, and then calls the parent implementation.
117 * All operations are wrapped in exception handling to ensure robustness
118 * in test and production environments.
119 */
120 @Override
121 protected void delayForWaitingNewUrl() {
122 try {
123 ComponentUtil.getSystemHelper().calibrateCpuLoad();
124 } catch (final Exception e) {
125 if (logger.isDebugEnabled()) {
126 logger.debug("Failed to calibrate CPU load", e);
127 }
128 }
129
130 try {
131 final IntervalControlHelper intervalControlHelper = ComponentUtil.getIntervalControlHelper();
132 intervalControlHelper.checkCrawlerStatus();
133 intervalControlHelper.delayByRules();
134 } catch (final Exception e) {
135 if (logger.isDebugEnabled()) {
136 logger.debug("Failed to apply interval control rules", e);
137 }
138 }
139
140 try {
141 super.delayForWaitingNewUrl();
142 } catch (final Exception e) {
143 if (logger.isDebugEnabled()) {
144 logger.debug("Failed to execute parent delay logic", e);
145 }
146 }
147 }
148 }