Get URL of current active tab from Firefox via command line

Question

I am running Mozilla Firefox 54.0.

Given an already opened session of Firefox and multiple open tabs, is there a way for me to extract the currently active tab (the tab I am looking at) via command line?

I could not find anything in a List of command line arguments or on the Mozilla Developer page.

My question differs from this question, as it neither works the intended way for me nor do I want all the tabs; I want one specific tab, the tab I am looking at.

Does anybody have an idea?

Is there maybe a way to interface with a running instance of Firefox?

Thanks for Reading

Edit: The Solution:

import json

f= open('~/.mozilla/firefox/RANDOM.default/sessionstore-backups/recovery.js' )

jdata = json.loads(f.read())

f.close()

CurrentTab = jdata.get("windows")[0].get("tabs")[jdata["windows"][0]["sel‌​ected"]-1].get("entr‌​ies")[HistLen-1].get‌​("url")

 while

HistLen = len(jdata.get("windows")[0].get("tabs")[jdata["windows"][0][‌​"selected"]-1].get("‌​entries"))

The HistLen was necessary because else I was always getting some older page I had previously opened in that tab.

Thanks for Reading

Answer 1

A solution has been provided here which is a combination of sed and python2. Here is a little bit more clear version of it:

sed -n "$(
python2 <<< $'import json
f = open("/home/username/.mozilla/firefox/RANDOM.default/sessionstore-backups/recovery.js", "r")
jdata = json.loads(f.read())
f.close()
print str(jdata["windows"][0]["selected"])')p" <(python2 <<< $'import json
f = open("/home/username/.mozilla/firefox/RANDOM.default/sessionstore-backups/recovery.js", "r")
jdata = json.loads(f.read())
f.close()
for win in jdata.get("windows"):
 for tab in win.get("tabs"):
  i = tab.get("index") - 1
  print tab.get("entries")[i].get("url")'
)

The file it's using is:

/home/username/.mozilla/firefox/RANDOM.profile/sessionstore.js

in more recent versions you should change it with:

/home/username/.mozilla/firefox/RANDOM.default/sessionstore-backups/recovery.js

Note that this file get regenerated every 15 second so after window being instantly changed it doesn't give you the correct URL, you have to wait some second.

---

How does this work?

At the first part it look for the id of active tab, it's something between 1 to count of open tabs. let's say it's "3", the code corresponding to this purpose is:

str(jdata["windows"][0]["selected"])

Next it returns a list of URLs (All open tabs) and feds it to the stdin of sed:

for win in jdata.get("windows"):
 for tab in win.get("tabs"):
  i = tab.get("index") - 1
  print tab.get("entries")[i].get("url")

So we are doing something like:

sed -n 3p <<< "URL1
URL2
URL3"

which brings us to "URL3".

Answer 2

I have thought about this problem as well and finally came to a solution, inspired by this gist. I thoroughly inspected the objects and unfortunately, I did not find any attribute assigning a tab to be the active tab.

What I instead found was the tabs has a lastAccessed key and from there I computed the most recent tab.

import json
import lz4.block
import pathlib
from time import time

# Set up path and regex for files
path = pathlib.Path.home().joinpath('.mozilla/firefox')
# files = path.glob('*default*release*/sessionstore-backups/recovery.jsonlz4')
files = path.glob('5ov6afhq.default-release-1/sessionstore-backups/recovery'+
                  '.jsonlz4')

for f in files:
    # decompress if necessary
    b = f.read_bytes()
    if b[:8] == b'mozLz40\0':
        b = lz4.block.decompress(b[8:])

    # load as json
    j = json.loads(b)
    if 'windows' in j.keys():
        for w in j['windows']:

            # Variables for keeping track of most recent tab
            most_recent_tab_index = ''
            min_time = 1000

            # run through tabs
            for t in w['tabs']:
                # Firefox does not 0-index
                i = t['index'] - 1

                # Convert time to seconds
                access_time = int((int(time()*1000) - t['lastAccessed'])/600)

                if access_time < min_time:
                    most_recent_tab_index = t['entries'][i]['url']

            print("MOST RECENT TAB: ", most_recent_tab_index)

Answer 3

If you use an X11-based desktop, you can try your luck with xprop -id with the window ID found by xlsclients -l. It needs some clever use of grep (possibly with grep -B... -A...) to filter out the parts you need.