quit and reopen firefox to sync the sessionstore

navigate to file:///data/data/org.mozilla.firefox/files/mozilla/

   then click XXXXXXXX.default/sessionstore.js

   longtap >> select all >> copy

open termux and run this:

   fn=~/sd/Download/sessionstore-$(date +%Y-%m%d-%H%M%S).js;

   termux-clipboard-get > $fn;

then run this (on phone or desktop) to convert the js to a list of links:

   echo '<style>html{font-family:sans-serif;background:#222} a{line-height:1.5em;padding:.5em;display:block;text-decoration:none;color:#0cf} strong{padding-left:.5em;color:#f4b} a:hover{background:#000}</style><meta charset="utf-8" />' >$fn.html;

   cat $fn | jq '[.. | .entries? | .[]? | {url,title}]' | jq -r '.[] | "<a href=\"\(.url|@html)\">\(.url) <strong>\(.title|@html)</strong></a>"' >> $fn.html;

(this locates "entries" on every level and, if it contains "url" and "title", prints those as an html link)

also unfortunately this trick doesn't work for chrome
since access to the data folder is restricted

== END OF GUIDE ==





notes on playing with the js and figuring out how to parse it

open the json in firefox on a desktop, hit F12, console

the current structure is roughly:

{ "windows": [
    { "tabs": [
        { "entries": [
            { "url": "https://ocv.me/",
              "title": "https://ocv.me/"
          }]},
        {
          "entries": [
            { "url": "https://arstechnica.com/",
              "title": "https://arstechnica.com/"
            },
            { "url": "https://arstechnica.com/some/article",
              "title": "article title maybe",
	  }]}],
      "closedTabs": [
        (same as "tabs")
]}]}


notes
  all entries seems to have a "title" attribute, assuming this

  "originalURI" is punycode, "url" is human-readable unicode
  
  looks like "entries" is tab groups or something
  
  apparently "_closedWindows" can appear at the same level as "windows",
  probably same structure? haven't seen it

  "formdata" has a dontcare attribute called "url"


so we want to look for entries:[{"url"}] at any level and collect tabs from there


prettyprint json:
  var j = JSON.parse(document.childNodes[0].textContent); document.write('<pre>' + JSON.stringify(j,null,4)); document.close();

count all urls:
  var j = JSON.parse(document.childNodes[0].textContent); var r=[]; function dive(j) { if (typeof j == 'string') return; for (var k in j) { var o = j[k]; if (o == null) continue; if (o.url) r.push(o.url); dive(o); }} dive(j); console.log(r.length);

show url list:
  document.write('<meta charset="utf-8" /><pre>'+r.join('\n'))


"selected": 628
count: 689