Merge pull request #4 from Curious-chen/main

feat: 支持目录页解析详情地址
uncle-novel · Jul 10, 2023 · 6f54c14 · 6f54c14
2 parents 2a4aa0b + 65c16ba
commit 6f54c14
Show file tree

Hide file tree

Showing 4 changed files with 18 additions and 3 deletions.
diff --git a/README.md b/README.md
@@ -17,7 +17,7 @@
 <dependency>
   <groupId>com.unclezs</groupId>
   <artifactId>novel-analyzer</artifactId>
-  <version>1.0.27</version>
+  <version>1.0.28</version>
 </dependency>
 ```
 

diff --git a/pom.xml b/pom.xml
@@ -6,7 +6,7 @@
     <name>novel-analyzer</name>
     <groupId>com.unclezs</groupId>
     <artifactId>novel-analyzer</artifactId>
-    <version>1.0.27</version>
+    <version>1.0.28</version>
 
     <url>https://github.com/unclezs/novel-analyzer</url>
     <description>Uncle小说解析 SDK</description>

diff --git a/src/main/java/com/unclezs/novel/analyzer/core/model/TocRule.java b/src/main/java/com/unclezs/novel/analyzer/core/model/TocRule.java
@@ -53,6 +53,10 @@ public class TocRule implements Verifiable, Serializable {
    * 章节链接
    */
   private CommonRule url;
+  /**
+   * 章节详情页规则
+   */
+  private CommonRule detailPage;
   /**
    * 章节下一页规则（存在则会匹配下一页）
    */

diff --git a/src/main/java/com/unclezs/novel/analyzer/spider/TocSpider.java b/src/main/java/com/unclezs/novel/analyzer/spider/TocSpider.java
@@ -5,6 +5,7 @@
 import com.unclezs.novel.analyzer.core.comparator.ChapterComparator;
 import com.unclezs.novel.analyzer.core.helper.AnalyzerHelper;
 import com.unclezs.novel.analyzer.core.helper.DebugHelper;
+import com.unclezs.novel.analyzer.core.matcher.Matchers;
 import com.unclezs.novel.analyzer.core.matcher.matchers.RegexMatcher;
 import com.unclezs.novel.analyzer.core.model.AnalyzerRule;
 import com.unclezs.novel.analyzer.core.model.TocRule;
@@ -190,7 +191,17 @@ protected boolean loadPage(int page) throws IOException {
     try {
       // 解析小说详情，从目录页
       if (page == 1) {
-        this.novel = NovelMatcher.details(originalText, rule.getDetail());
+        String detailPageUrl = Matchers.match(originalText, tocRule.getDetailPage());
+        // 如果详情页不为空，则请求详情页
+        if (StringUtils.isNotBlank(detailPageUrl) && !detailPageUrl.equals(params.getUrl())) {
+          // 默认使用目录页的请求参数
+          RequestParams detailParams = RequestParams.create(detailPageUrl, rule.getToc().getParams());
+          String detailText = SpiderHelper.request(rule.getParams(), detailParams);
+          this.novel = NovelMatcher.details(detailText, rule.getDetail());
+        } else {
+          // 否则，从目录页解析详情
+          this.novel = NovelMatcher.details(originalText, rule.getDetail());
+        }
         this.novel.setUrl(params.getUrl());
       }
       List<Chapter> chapters = NovelMatcher.toc(originalText, tocRule);