Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-04-20 23:56:34 -04:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Merge branch 'master' into simple-behaviors

Browse Source
This commit is contained in:
Noah Levitt 2015-02-02 14:48:26 -08:00
commit 48b8754f40
2 changed files with 132 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

135
umbra/behaviors.d/default.js
View File

@ -2,19 +2,116 @@
//
// Scrolls to the bottom of the page. That's it at the moment.
//
var umbraAboveBelowOrOnScreen = function(e) {
var eTop = e.getBoundingClientRect().top;
if (eTop < window.scrollY) {
return -1; // above
} else if (eTop > window.scrollY + window.innerHeight) {
return 1; // below
} else {
return 0; // on screen
}
}
var UMBRA_IFRAME_SOUNDCLOUD_EMBEDDED_SELECTOR = "iframe";
var UMBRA_THINGS_TO_CLICK_SOUNDCLOUD_EMBEDDED_SELECTOR = "button.sc-button-play, button.playButton";
var MAX_IFRAME_RECURSE_DEPTH = 1; //0-based
var umbraState = {'idleSince':null};
var umbraAlreadyClicked = {};
var umbraFinished = false;
var umbraIntervalFunc = function() {
var needToScroll = (window.scrollY + window.innerHeight < document.documentElement.scrollHeight);
// console.log('intervalFunc umbraState.idleSince=' + umbraState.idleSince + ' needToScroll=' + needToScroll + ' window.scrollY=' + window.scrollY + ' window.innerHeight=' + window.innerHeight + ' document.documentElement.scrollHeight=' + document.documentElement.scrollHeight);
if (needToScroll) {
window.scrollBy(0, 200);
umbraState.idleSince = null;
} else if (umbraState.idleSince == null) {
var umbraSoundCloudEmbeddedElements = [];
getUmbraSoundCloudEmbeddedElements(umbraSoundCloudEmbeddedElements);
var clickedSomething = false;
var somethingLeftBelow = false;
var somethingLeftAbove = false;
var missedAbove = 0;
for (var i = 0; i < umbraSoundCloudEmbeddedElements.length; i++) {
var targetId = umbraSoundCloudEmbeddedElements[i].id;
var target = umbraSoundCloudEmbeddedElements[i].target;
if (!(targetId in umbraAlreadyClicked)) {
var where = umbraAboveBelowOrOnScreen(target);
if (where == 0) { // on screen
// var pos = target.getBoundingClientRect().top;
// window.scrollTo(0, target.getBoundingClientRect().top - 100);
console.log("clicking at " + target.getBoundingClientRect().top + " on " + target.outerHTML);
if (target.click != undefined) {
target.click();
}
umbraAlreadyClicked[targetId] = true;
clickedSomething = true;
umbraState.idleSince = null;
break;
} else if (where > 0) {
somethingLeftBelow = true;
} else if (where < 0) {
somethingLeftAbove = true;
}
}
}
if (!clickedSomething) {
if (somethingLeftAbove) {
console.log("scrolling UP because everything on this screen has been clicked but we missed something above");
window.scrollBy(0, -500);
umbraState.idleSince = null;
} else if (somethingLeftBelow) {
console.log("scrolling because everything on this screen has been clicked but there's more below document.body.clientHeight=" + document.body.clientHeight);
window.scrollBy(0, 200);
umbraState.idleSince = null;
} else if (window.scrollY + window.innerHeight < document.documentElement.scrollHeight) {
console.log("scrolling because we're not to the bottom yet document.body.clientHeight=" + document.body.clientHeight);
window.scrollBy(0, 200);
umbraState.idleSince = null;
} else if (umbraState.idleSince == null) {
umbraState.idleSince = Date.now();
}
}
if (umbraState.idleSince == null) {
umbraState.idleSince = Date.now();
}
}
//try to detect sound cloud "Play" buttons and return them as targets for clicking
var getUmbraSoundCloudEmbeddedElements = function(soundCloudEmbeddedElements, currentIframeDepth, currentDocument,
iframeElement) {
//set default values for parameters
currentIframeDepth = currentIframeDepth || 0;
currentDocument = currentDocument || document;
if (currentIframeDepth > MAX_IFRAME_RECURSE_DEPTH) {
return;
}
//collect all buttons on current document first
var button = [];
button = currentDocument.querySelectorAll(UMBRA_THINGS_TO_CLICK_SOUNDCLOUD_EMBEDDED_SELECTOR);
var cssPathIframe = iframeElement ? getElementCssPath(iframeElement) : "";
for (var i = 0; i < button.length; i++) {
soundCloudEmbeddedElements.push({"id" : cssPathIframe + getElementCssPath(button.item(i)), "target" : button.item(i)});
}
//now get all buttons in embedded iframes
var iframe = [];
iframe = currentDocument.querySelectorAll(UMBRA_IFRAME_SOUNDCLOUD_EMBEDDED_SELECTOR);
for (var i = 0; i < iframe.length; i++) {
getUmbraSoundCloudEmbeddedElements(soundCloudEmbeddedElements, currentIframeDepth + 1, iframe[i].contentWindow.document.body, iframe[i]);
}
}
// If we haven't had anything to do (scrolled, clicked, etc) in this amount of
@ -32,4 +129,30 @@ var umbraBehaviorFinished = function() {
return false;
}
//copied from http://stackoverflow.com/questions/4588119/get-elements-css-selector-without-element-id
var getElementCssPath = function(element) {
var names = [];
while (element.parentNode){
if (element.id){
names.unshift('#' + element.id);
break;
} else {
if (element == element.ownerDocument.documentElement) {
names.unshift(element.tagName);
}
else {
for (var c = 1, e = element; e.previousElementSibling; e = e.previousElementSibling, c++);
names.unshift(element.tagName + ":nth-child(" + c + ")");
}
element = element.parentNode;
}
}
return names.join(" > ");
}
var umbraIntervalId = setInterval(umbraIntervalFunc, 100);

5
umbra/browser.py
View File

@ -231,7 +231,7 @@ class Chrome:
# returns websocket url to chrome window with about:blank loaded
def start(self):
timeout_sec = 60
timeout_sec = 600
new_env = os.environ.copy()
new_env["HOME"] = self.user_home_dir
chrome_args = [self.executable,
@ -242,6 +242,7 @@ class Chrome:
"--window-size=1100,900", "--no-default-browser-check",
"--disable-first-run-ui", "--no-first-run",
"--homepage=about:blank", "--disable-direct-npapi-requests",
"--disable-web-security",
"about:blank"]
self.logger.info("running {}".format(chrome_args))
self.chrome_process = subprocess.Popen(chrome_args, env=new_env, start_new_session=True)
@ -270,7 +271,7 @@ class Chrome:
time.sleep(0.5)
def stop(self):
timeout_sec = 60
timeout_sec = 300
self.logger.info("terminating chrome pid {}".format(self.chrome_process.pid))
self.chrome_process.terminate()