backup/pywb
1
0
Fork 0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-15 00:03:28 +01:00
Code Issues Projects Releases Activity

wombat: cache partial tag document.write() calls, eg. document.write("<scr"), document.write("ipt>") and parse and write full tag.

Browse Source 
(detect if no tags added, not 100% fool-proof but covers common cases at this point)
This commit is contained in:
Ilya Kreymer 2015-09-02 05:36:34 -07:00
parent 6df114b948
commit ddcd03f1fc
1 changed files with 20 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
pywb/static/wombat.js
View File

@ -1086,13 +1086,19 @@ var wombat_internal = function($wbwindow) {
}
}
var write_buff = "";
//============================================
function rewrite_html(string, check_end_tag) {
if (!string) {
return string;
}
var inner_doc = new DOMParser().parseFromString(string, "text/html");
if (write_buff) {
string = write_buff + string;
write_buff = "";
}
var inner_doc = new DOMParser().parseFromString(string, "text/html");
if (!inner_doc) {
return string;
@ -1105,7 +1111,7 @@ var wombat_internal = function($wbwindow) {
}
var new_html = "";
// if original had <html> tag, add full document HTML
if (string && string.indexOf("<html") >= 0) {
new_html = inner_doc.documentElement.outerHTML;
@ -1114,10 +1120,15 @@ var wombat_internal = function($wbwindow) {
new_html = inner_doc.head.innerHTML;
new_html += inner_doc.body.innerHTML;
if (check_end_tag && inner_doc.all.length > 3) {
var end_tag = "</" + inner_doc.all[3].tagName.toLowerCase() + ">";
if (ends_with(new_html, end_tag) && !ends_with(string, end_tag)) {
new_html = new_html.substring(0, new_html.length - end_tag.length);
if (check_end_tag) {
if (inner_doc.all.length > 3) {
var end_tag = "</" + inner_doc.all[3].tagName.toLowerCase() + ">";
if (ends_with(new_html, end_tag) && !ends_with(string, end_tag)) {
new_html = new_html.substring(0, new_html.length - end_tag.length);
}
} else if (string[0] != "<" || string[string.length - 1] != ">") {
write_buff += string;
return;
}
}
}
@ -1790,6 +1801,9 @@ var wombat_internal = function($wbwindow) {
var new_write = function(string) {
new_buff = rewrite_html(string, true);
if (!new_buff) {
return;
}
var res = orig_doc_write.call(this, new_buff);
check_wombat(this.defaultView);
return res;