webMagic-Mobai
/
magic-Dependency
mirror of https://gitea.qinmian.online/Mobai/magic-Dependency.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

use group 0 of targetUrl instead of group 1 #588

master
yihua.huang 2017-06-04 07:11:13 +08:00
parent 9b77306098
commit 211ac74389
2 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
webmagic-extension/src/main/java/us/codecraft/webmagic/model/ModelPageProcessor.java
View File

@ -74,7 +74,7 @@ class ModelPageProcessor implements PageProcessor {
for (Pattern targetUrlPattern : urlPatterns) {
Matcher matcher = targetUrlPattern.matcher(link);
if (matcher.find()) {
page.addTargetRequest(new Request(matcher.group(1)));
page.addTargetRequest(new Request(matcher.group(0)));
}
}
}

6
webmagic-extension/src/main/java/us/codecraft/webmagic/model/PageModelExtractor.java
View File

@ -163,12 +163,12 @@ class PageModelExtractor {
private void initClassExtractors() {
Annotation annotation = clazz.getAnnotation(TargetUrl.class);
if (annotation == null) {
targetUrlPatterns.add(Pattern.compile("(.*)"));
targetUrlPatterns.add(Pattern.compile(".*"));
} else {
TargetUrl targetUrl = (TargetUrl) annotation;
String[] value = targetUrl.value();
for (String s : value) {
targetUrlPatterns.add(Pattern.compile("(" + s.replace(".", "\\.").replace("*", "[^\"'#]*") + ")"));
targetUrlPatterns.add(Pattern.compile(s.replace(".", "\\.").replace("*", "[^\"'#]*")));
}
if (!targetUrl.sourceRegion().equals("")) {
targetUrlRegionSelector = new XpathSelector(targetUrl.sourceRegion());
@ -179,7 +179,7 @@ class PageModelExtractor {
HelpUrl helpUrl = (HelpUrl) annotation;
String[] value = helpUrl.value();
for (String s : value) {
helpUrlPatterns.add(Pattern.compile("(" + s.replace(".", "\\.").replace("*", "[^\"'#]*") + ")"));
helpUrlPatterns.add(Pattern.compile(s.replace(".", "\\.").replace("*", "[^\"'#]*")));
}
if (!helpUrl.sourceRegion().equals("")) {
helpUrlRegionSelector = new XpathSelector(helpUrl.sourceRegion());