Trynig to add a new, persistent, Firefox profile with Selenium. AFAIK, when executing FirefoxProfile(), a new profile is generated using a temporary file. Ideally, this profile should be able to remain available to subsequent processes - even after the creator is closed.
Problem:
Create a new Firefox profile from within Python code. This should return a FirefoxProfile object that is usable with the Firefox webdriver Selenium uses.
The profile created should persist after the process ends - i.e. it should be a full-fledged profile, not just a temporary profile.
Some pointers:
The profiles.ini file seems to be key. I have read some code that uses the Java class ProfilesIni to modify profile information. If this class is available for Python code, it should probably take care of most of this.
If the only way to do this is to manually modify the profiles.ini file, that's acceptable. A better, more standardized solution (one that uses a library or Selenium code) would be preferable, however.
Thanks very much!
If this helps anyone, what needs to be done is run:
firefox[.exe] -CreateProfile <profile_name>
The .exe in brackets is intended to provide for it being run under Windows.
Yes, this does not use the Selenium library in Python, but it does provide the desired result.
Related
I would like to create a new session for each tab I open and then control the sessions individually using Selenium python. Is this possible?
#reynoldsnp: Firefox has an official addon that does this, but I'm not sure if you can get selenium to interact with the addon. addons.mozilla.org/en-GB/firefox/addon/multi-account-containers If you figure out a way to do it, I would love to know how.
(I can't comment yet due to my reputation score, therefore quoted comment).
I don't know how to actually interact with the extension but if you have a known set of sites you would like to open:
Try this:
Make a firefox profile for your use with selenium. Multiple profiles
Windows 8/8.1/10:
Press Win + R on your keyboard.
Type firefox --new-instance --ProfileManager
Open Firefox in that profile by selecting the new profile in the setup wizard. Install the extension in that profile.
Set up the containers you would like in the extension, to, by default, open up with a specific site.
Ensure that the checkbox is ticked.
Start selenium with that profile like this:
from selenium import webdriver
profile = webdriver.FirefoxProfile('path/to/your/profile') # on windows found here: %APPDATA%/Mozilla/Firefox/Profiles
driver = webdriver.Firefox(firefox_profile=profile)
Navigate between tabs, effectively containers, using selenuium.
First, no, you cannot. While tabs runs as a process, they are attached to the session ID which initially open the browser. This is how the protocol works https://www.w3.org/TR/webdriver/#new-session
They have, however, a unique ID which you can use to identify them by and switch between them.
driver.window_handles
will give you the list of open tabs. Each tab is fully isolated. You can now choose between
driver.switch_to_window("any open tab taken from windows handles list")
driver.do_something
driver.switch_to_window("any other tab from windows handles list")
driver.do_something_else_on_other_tab
# or (this option can let you run in parallel)
driver a = ChromeDriver()
driver b = ChromeDriver()
a.do_something
b.do_something
As suggested (and I personally do myself) open new session for each tab you want, that way you can parallel them and run much faster, all in all.
I am not sure the performance difference is that significant between multiple browsers or multiple tabs... they should use almost the same resources.
I didn't find an answer to this. To automate some tests I have used Firefox and Selenium. I thought to give a try to Selenium 3.0 and its Marionette interface. Everything works with an old Firefox binary and the old webdriver way, so my code as such works.
For my tests I have created a Firefox profile, which I then pass to Selenium. Previously I did it like this:
p = webdriver.FirefoxProfile("profilename")
driver = webdriver.Firefox(firefox_profile=p, firefox_binary="/path/to/ff")
This works. My browser starts and whatever modifications it makes to the profile (cookies), seem to be there when I restart the browser with the same profile.
With Marionette/geckodriver the instructions tell me to use capabilities["profile"] for this, and they state this should be "Base64-encoded zip of a profile directory to use as the profile for the Firefox instance".
Ok. First, how do I create a base64-encoded zip in Python? Or do I just use shell for this?
Second, how does this work in practice? If I zip and base64 encode my profile directory, how do I then get the modified version back after my tests complete? If I created this file in shell and kept passing the same file for consecutive attempts, every modification, for example a login cookie, would be lost, and I would need to start from the scratch again, which in this case is not desirable.
I can keep using the old system at least for now, this is just to satisfy my curiosity.
Hannu
I am new to web programming.
I am trying to add parameters to one of my Chrome extension.
I know I can enter, for example, "window.localStorage.setItem()" in the extension console. However, I cannot find a way to navigate my webdriver to that extension background page. I have seen, in the past, people would use chrome//extensions:extension_id as url to get to that page, but now this method seems not to work.
Is there any way that I can go to that page directly without telling my webdriver to click programmer mode and then click the extension?
Thanks in advance. This has been bothered me for hours.
Probably one thing to correct here is that chrome://extensions/ opens up the Extensions on Chrome browser. So this shall work fine :
driver.get("chrome://extensions/")
Just a note now the chrome://extensions/extension_id does not take you to the extension page anymore.
In case you are interested in the Details or Options of an extension, you shall try and access it via the links only since the uri is not consistent for different extensions as well. e.g
Extension 1
chrome-extension://gighmmpiobklfepjocnamgkkbiglidom/options/index.html
Extension 2 ::
chrome-extension://fngmhnnpilhplaeedifhccceomclgfbg/options_pages/support.html
The general format has extension_id but not certainly defines the entire uri to get you to the page.
I would suggest if you want to work around with the Extensions, there is this Management API which seems useful to dive in.
I'm not sure how to find this information, I have found a few tutorials so far about using Python with selenium but none have so much as touched on this.. I am able to run some basic test scripts through python that automate selenium but it just shows the browser window for a few seconds and then closes it.. I need to get the browser output into a string / variable (ideally) or at least save it to a file so that python can do other things on it (parse it, etc).. I would appreciate if anyone can point me towards resources on how to do this. Thanks
using Selenium Webdriver and Python, you would simply access the .page_source property to get the source of the current page.
for example, using Firefox() driver:
from selenium import webdriver
driver = webdriver.Firefox()
driver.get('http://www.example.com/')
print(driver.page_source)
driver.quit()
There's a Selenium.getHtmlSource() method in Java, most likely it is also available in Python. It returns the source of the current page as string, so you can do whatever you want with it
Ok, so here is how I ended up doing this, for anyone who needs this in the future..
You have to use firefox for this to work.
1) create a new firefox profile (not necessary but ideal so as to separate this from normal firefox usage), there is plenty of info on how to do this on google, it depends on your OS how you do this
2) get the firefox plugin: https://addons.mozilla.org/en-US/firefox/addon/2704/ (this automatically saves all pages for a given domain name), you need to configure this to save whichever domains you intend on auto-saving.
3) then just start the selenium server to use the profile you created (below is an example for linux)
cd /root/Downloads/selenium-remote-control-1.0.3/selenium-server-1.0.3
java -jar selenium-server.jar -firefoxProfileTemplate /path_to_your_firefox_profile/
Thats it, it will now save all the pages for a given domain name whenever selenium visits them, selenium does create a bunch of garbage pages too so you could just delete these via a simple regex parsing and its up to you, from there how to manipulate the saved pages
I'm looking for a solution similar to presented in this question, except it has to deal with python.
the answer that was accepted is language agnostic. You will need to setup a firefox profile then when starting up selenium rc tell it to use that profile.
You just need to run your tests to use *firefox against the rc that has your profile and when the browser loads it will be the one you are after