mirror of
https://github.com/MarginaliaSearch/MarginaliaSearch.git
synced 2025-02-24 05:18:58 +00:00
(feature-extractor) Add mail.ru counter to non-adtech trackers
This commit is contained in:
parent
df85468c01
commit
d8073f0dde
@ -23,8 +23,12 @@ public class FeatureExtractor {
|
|||||||
private static final List<String> innocentTrackers = List.of(
|
private static final List<String> innocentTrackers = List.of(
|
||||||
"twitter.com",
|
"twitter.com",
|
||||||
"bing.com",
|
"bing.com",
|
||||||
"msn.com");
|
"msn.com",
|
||||||
private static final List<String> adtechTrackers = List.of("adform.net",
|
"mail.ru/counter"
|
||||||
|
);
|
||||||
|
private static final List<String> adtechTrackers = List.of(
|
||||||
|
"publir.com",
|
||||||
|
"adform.net",
|
||||||
"connect.facebook",
|
"connect.facebook",
|
||||||
"facebook.com/tr",
|
"facebook.com/tr",
|
||||||
"absbygoogle.com",
|
"absbygoogle.com",
|
||||||
@ -222,6 +226,12 @@ public class FeatureExtractor {
|
|||||||
|
|
||||||
String scriptText = scriptTag.html();
|
String scriptText = scriptTag.html();
|
||||||
|
|
||||||
|
if (scriptText.contains("_ga=") || scriptText.contains("ga('create'")) {
|
||||||
|
features.add(HtmlFeature.TRACKING);
|
||||||
|
}
|
||||||
|
if (scriptText.contains("_tmr")) {
|
||||||
|
features.add(HtmlFeature.TRACKING);
|
||||||
|
}
|
||||||
if (scriptText.contains("'pd.js'")) {
|
if (scriptText.contains("'pd.js'")) {
|
||||||
features.add(HtmlFeature.PARDOT);
|
features.add(HtmlFeature.PARDOT);
|
||||||
}
|
}
|
||||||
|
Loading…
Reference in New Issue
Block a user