add isExtractLinks to OOSpider #575
parent
8999ea9320
commit
7ffc6998ef
|
@ -23,6 +23,8 @@ class ModelPageProcessor implements PageProcessor {
|
|||
|
||||
private Site site;
|
||||
|
||||
private boolean extractLinks = true;
|
||||
|
||||
public static ModelPageProcessor create(Site site, Class... clazzs) {
|
||||
ModelPageProcessor modelPageProcessor = new ModelPageProcessor(site);
|
||||
for (Class clazz : clazzs) {
|
||||
|
@ -45,8 +47,10 @@ class ModelPageProcessor implements PageProcessor {
|
|||
@Override
|
||||
public void process(Page page) {
|
||||
for (PageModelExtractor pageModelExtractor : pageModelExtractorList) {
|
||||
if (extractLinks) {
|
||||
extractLinks(page, pageModelExtractor.getHelpUrlRegionSelector(), pageModelExtractor.getHelpUrlPatterns());
|
||||
extractLinks(page, pageModelExtractor.getTargetUrlRegionSelector(), pageModelExtractor.getTargetUrlPatterns());
|
||||
}
|
||||
Object process = pageModelExtractor.process(page);
|
||||
if (process == null || (process instanceof List && ((List) process).size() == 0)) {
|
||||
continue;
|
||||
|
@ -83,4 +87,12 @@ class ModelPageProcessor implements PageProcessor {
|
|||
public Site getSite() {
|
||||
return site;
|
||||
}
|
||||
|
||||
public boolean isExtractLinks() {
|
||||
return extractLinks;
|
||||
}
|
||||
|
||||
public void setExtractLinks(boolean extractLinks) {
|
||||
this.extractLinks = extractLinks;
|
||||
}
|
||||
}
|
||||
|
|
|
@ -97,4 +97,9 @@ public class OOSpider<T> extends Spider {
|
|||
return this;
|
||||
}
|
||||
|
||||
public OOSpider setIsExtractLinks(boolean isExtractLinks){
|
||||
modelPageProcessor.setExtractLinks(isExtractLinks);
|
||||
return this;
|
||||
}
|
||||
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue