Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-05-02 14:46:18 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

avoid "Uncaught TypeError: Cannot read property 'querySelectorAll' of undefined" from outlinks script

Browse source
This commit is contained in:
Noah Levitt 2016-08-25 13:10:30 -07:00
parent ed7e01210d
commit 1c5c9417d2
2 changed files with 8 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

12
brozzler/browser.py
View file

@ -304,11 +304,13 @@ class Browser:
var __brzl_framesDone = new Set(); var __brzl_framesDone = new Set();
var __brzl_compileOutlinks = function(frame) { var __brzl_compileOutlinks = function(frame) {
__brzl_framesDone.add(frame); __brzl_framesDone.add(frame);
var outlinks = Array.prototype.slice.call( if (frame && frame.document) {
frame.document.querySelectorAll('a[href]')); var outlinks = Array.prototype.slice.call(
for (var i = 0; i < frame.frames.length; i++) { frame.document.querySelectorAll('a[href]'));
if (frame.frames[i] && !__brzl_framesDone.has(frame.frames[i])) { for (var i = 0; i < frame.frames.length; i++) {
outlinks = outlinks.concat(__brzl_compileOutlinks(frame.frames[i])); if (frame.frames[i] && !__brzl_framesDone.has(frame.frames[i])) {
outlinks = outlinks.concat(__brzl_compileOutlinks(frame.frames[i]));
}
} }
} }
return outlinks; return outlinks;

2
setup.py
View file

@ -32,7 +32,7 @@ def find_package_data(package):
setuptools.setup( setuptools.setup(
name='brozzler', name='brozzler',
version='1.1b6.dev77', version='1.1b6.dev78',
description='Distributed web crawling with browsers', description='Distributed web crawling with browsers',
url='https://github.com/internetarchive/brozzler', url='https://github.com/internetarchive/brozzler',
author='Noah Levitt', author='Noah Levitt',