How to decode/decipher Mozilla Firefox proprietary .jsonlz4 format? (sessionstore-backups/recovery.jsonlz4)

I'm trying to get a handle on Mozilla Firefox's proprietary file format .jsonlz4, used, for example, for sessionstore-backups/recovery.jsonlz4, but to no avail.

How do I get back my data, specifically, some long text I've typed in some textareas of a crashed session? It's my data!

4

6 Answers

As per , the following appears to work most reliably:

• in about:config, toggle the devtools.chrome.enabled setting from the default of false to a value of true

• open Scratchpad from within Firefox:

  • either with fn+Shift+F4 on a MacBook,
  • or Shift+F4,
  • or through the menu bar through Tools → Web Developer → Scratchpad

• in the menu bar within Scratchpad of Firefox, change Environment from Content to Browser (omitting this step would subsequently result in errors like Exception: ReferenceError: OS is not defined at the next step)

• use code like the following within the Scratchpad of Firefox:

  var file = "/Users/…/sessionstore-backups/recovery.baklz4";
  //OS.File.read(file, { compression: "lz4" }).then(bytes =>
  // OS.File.writeAtomic(file + ".uncompressed", bytes));
  OS.File.read(file, { compression: "lz4" }).then(bytes => { OS.File.writeAtomic(file + ".uncompressed.stringify", JSON.stringify(JSON.parse(new TextDecoder().decode(bytes)), null, 1))
  });

  The final parameter to JSON.stringify handles how many spaces would be used at each line; putting 0 causes the whole thing to be printed on a single line, putting 1 splits the lines properly (putting 2 would create too much useless whitespace and increases the size of the file for little benefit)

• click the Run button

• run fgrep :textarea /Users/…/sessionstore-backups/recovery.baklz4.uncompressed.stringify from within the Terminal app

5

Unfortunately, due to a non-standard header, standard tools won't work. There's an open proposal to change that. Apparently the Mozilla header was devised before a standard lz4 frame format existed; it does wrap a standard lz4 block.

That said, the same bug report includes a few alternative methods. I'll list them briefly:

• Use the dejsonlz4 tool, which includes binary builds for Windows and should be easy to build on *nix
  • lz4json is a similar tool, but relies on an external liblz4 and is somewhat easier to build on *nix but harder on Windows (outside WSL)
• Use this fairly simple Python script: (requires the lz4 package via pip or your package manager) -- the script appears to be python3 but is trivially adaptable to python2
• There is a webextension available that should be able to open these. NB: while source is available, I have not verified it, and the permissions it requests are a bit concerning (especially the response to concerns)
• In theory, you should be able to strip the first 8 bytes (e.g. with dd if=original.jsonlz4 of=stripped.lz4 bs=8 skip=1) and that should leave you with a valid lz4 block. Note that this is distinct from a lz4 frame. While most programming languages have libraries that can easily decode a block, finding a prebuilt tool to do so is more difficult, e.g. the liblz4-tool package only accepts the frame format.

5

I found the following methods to be working, testing on Ubuntu 20.04:

Method 1: Using the mozlz4 binary from GitHub:

Download the linux binary for mozlz4 from . Then run the following:

chmod u+x mozlz4-linux
./mozlz4-linux -x filename.jsonlz4

Method 2: Using the lz4json package from Ubuntu repos:

Ubuntu 20.04 repos have a package named lz4json. I haven't checked whether it is present on previous Ubuntu versions.

To install and use that, run

sudo apt install lz4 lz4json
lz4jsoncat ~/.mozilla/firefox/*default*/sessionstore-backups/recovery.jsonlz4

The above output will show a minified json. To make it readable, you can use the 'jq' json parser:

sudo apt install jq
# then pipe the output of the previous command through jq to make it readable:
lz4jsoncat ~/.mozilla/firefox/*default*/sessionstore-backups/recovery.jsonlz4 | jq

If you just want to see the list of URLs and the page titles, you can use this:

lz4jsoncat ~/.mozilla/firefox/*default*/sessionstore-backups/recovery.jsonlz4 \ | jq '.["windows"] | .[0] | .["tabs"] | .[] | .["entries"] | .[0] | .url,.title' \ | grep -v 'New Tab' | grep -v 'about:newtab' | sed 's/"http/\n"http/g'

2

I was able to extract the URLs from the {profile-dir}/sessionstore-backups/recovery.jsonlz4 file using the following free online tool designed expressly for this purpose:

The same site offers a similar tool for decrypting jsonlz4 files from the {profile-dir}/bookmarkbackups directory.

2

On UNIX® and UNIX-like systems, like Mac OS X with MacPorts, FreeBSD, OpenBSD or NetBSD with pkgsrc, the following fork of lz4json could also be used to compile cleanly out of the box, e.g., on Mac OS X w/ MacPorts:

sudo port install lz4
git clone
cd lz4json
make
./lz4jsoncat ~/Library/Application\ Support/Firefox/Profiles/CHANGE\ \
| python -m json.tool | fgrep :textarea | more

The Python script from @Bob's answer is broken with latest python3 lz4 lib, but it can be fixed if you replace

import lz4

with

import lz4.block as lz4

Your Answer

Sign up or log in

Sign up using Google Sign up using Facebook Sign up using Email and Password Submit

Post as a guest

Post Your Answer Discard

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy