161sh/CatgirlIntelligenceAgency
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Big brain web developers were using onload and onerror handlers to load JS without script tags...

Browse Source
This commit is contained in:
Viktor Lofgren 2023-06-30 17:10:25 +02:00
parent 7d86586594
commit d2fdaafc7a
2 changed files with 28 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
code/processes/converting-process/src/main/java/nu/marginalia/converting/processor/logic/DocumentValuator.java
View File

@ -36,8 +36,15 @@ public class DocumentValuator {
var scriptVisitor = new ScriptVisitor();
parsed.getElementsByTag("script").traverse(scriptVisitor);
int value = scriptVisitor.score();
return scriptVisitor.score();
for (var links : parsed.head().getElementsByTag("link")) {
if (links.hasAttr("onerror") || links.hasAttr("onload")) {
value += 1;
}
}
return value;
}
private static class ScriptVisitor implements NodeVisitor {
@ -56,7 +63,6 @@ public class DocumentValuator {
}
else if (node instanceof TextNode tn) {
visitScriptText(tn);
}
}
@ -73,8 +79,7 @@ public class DocumentValuator {
String srcAttr = el.attr("src");
if (srcAttr.contains("wp-content") || srcAttr.contains("wp-includes") || srcAttr.contains("jquery")) {
penalty += 0.49;
}
else if (!Strings.isBlank(srcAttr)) {
} else if (!Strings.isBlank(srcAttr)) {
penalty += 1;
}
}

20
code/processes/converting-process/src/main/java/nu/marginalia/converting/processor/logic/FeatureExtractor.java
View File

@ -76,6 +76,19 @@ public class FeatureExtractor {
}
}
// 500 IQ web developers use <link> error or load handlers
// sneakily load JS without explicit script tags
for (var link : doc.head().getElementsByTag("link")) {
if (link.hasAttr("onerror")) {
features.add(HtmlFeature.JS);
break;
}
if (link.hasAttr("onload")) {
features.add(HtmlFeature.JS);
break;
}
}
if (features.contains(HtmlFeature.JS) && adblockSimulator.hasAds(doc.clone())) {
features.add(HtmlFeature.ADVERTISEMENT);
}
@ -117,8 +130,13 @@ public class FeatureExtractor {
}
private boolean hasTrackingScript(Element scriptTag) {
return hasTrackingScript(scriptTag.attr("src"));
}
private boolean hasTrackingScript(String scriptText) {
for (var tracker : trackers) {
if (scriptTag.attr("src").contains(tracker)) {
if (scriptText.contains(tracker)) {
return true;
}
}