Mirrors/FairEmail
1
0
Fork 0
mirror of https://github.com/M66B/FairEmail.git synced 2026-01-02 19:10:11 +01:00
Code Issues Packages Projects Releases Wiki Activity

Remove HTML elements with a namespace

Browse Source
This commit is contained in:
M66B
2019-05-13 11:03:15 +02:00
parent a6c59b973b
commit d2737219e0
1 changed files with 33 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
app/src/main/java/eu/faircode/email/HtmlHelper.java
View File

@@ -76,6 +76,7 @@ public class HtmlHelper {
static final int PREVIEW_SIZE = 250;
private static final int TRACKING_PIXEL_SURFACE = 25;
private static final List<String> heads = Collections.unmodifiableList(Arrays.asList(
"h1", "h2", "h3", "h4", "h5", "h6", "p", "ol", "ul", "table", "br", "hr"));
private static final List<String> tails = Collections.unmodifiableList(Arrays.asList(
@@ -115,6 +116,38 @@ public class HtmlHelper {
boolean paranoid = prefs.getBoolean("paranoid", true);
Document parsed = Jsoup.parse(html);
// <html xmlns:v="urn:schemas-microsoft-com:vml"
// xmlns:o="urn:schemas-microsoft-com:office:office"
// xmlns:w="urn:schemas-microsoft-com:office:word"
// xmlns:m="http://schemas.microsoft.com/office/2004/12/omml"
// xmlns="http://www.w3.org/TR/REC-html40">
// <o:p>&nbsp;</o:p></span>
// Default XHTML namespace: http://www.w3.org/1999/xhtml
String ns = null;
for (Element h : parsed.select("html"))
for (Attribute a : h.attributes()) {
if (a.getKey().startsWith("xmlns:") &&
a.getValue().startsWith("http://www.w3.org/")) {
ns = a.getKey().split(":")[1];
break;
}
}
for (Element e : parsed.select("*"))
if (e.tagName().contains(":")) {
String tag = e.tagName();
if (ns != null && e.tagName().startsWith(ns)) {
e.tagName(tag.split(":")[1]);
Log.i("Updated tag=" + tag + " to=" + e.tagName());
} else {
e.remove();
Log.i("Removed tag=" + tag);
}
}
Whitelist whitelist = Whitelist.relaxed()
.addTags("hr", "abbr")
.removeTags("col", "colgroup", "thead", "tbody")