MozLz4a compression/decompression utility

mozlz4a.py

#!/usr/bin/env python3
# vim: sw=4 ts=4 et tw=100 cc=+1
#
####################################################################################################
#                                           DESCRIPTION                                            #
####################################################################################################
#
# Decompressor/compressor for files in Mozilla's "mozLz4" format. Firefox uses this file format to
# compress e. g. bookmark backups (*.jsonlz4).
#
# This file format is in fact just plain LZ4 data with a custom header (magic number [8 bytes] and
# uncompressed file size [4 bytes, little endian]).
#
####################################################################################################
#                                           DEPENDENCIES                                           #
####################################################################################################
#
# - Tested with Python 3.10
# - LZ4 bindings for Python, version 4.x: https://pypi.python.org/pypi/lz4
#
####################################################################################################
#                                             LICENSE                                              #
####################################################################################################
#
# Copyright (c) 2015-2022, Tilman Blumenbach
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without modification, are permitted
# provided that the following conditions are met:
#
# 1. Redistributions of source code must retain the above copyright notice, this list of conditions
#    and the following disclaimer.
# 2. Redistributions in binary form must reproduce the above copyright notice, this list of
#    conditions and the following disclaimer in the documentation and/or other materials provided
#    with the distribution.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR
# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
# FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
# IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
# OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

import argparse
import sys

import lz4.block


class BinFileArg:
    def __init__(self, mode):
        self._mode = mode

    def __call__(self, arg):
        objs = {
            "r": sys.stdin.buffer,
            "w": sys.stdout.buffer,
        }

        if arg == "-":
            return objs[self._mode]

        try:
            return open(arg, self._mode + "b")
        except OSError as e:
            raise argparse.ArgumentTypeError(
                "failed to open file for %s: %s" % (
                    "reading" if self._mode == "r" else "writing",
                    e
                )
            )


def decompress(file_obj):
    if file_obj.read(8) != b"mozLz40\0":
        raise ValueError("Invalid magic number")

    return lz4.block.decompress(file_obj.read())


def compress(file_obj):
    compressed = lz4.block.compress(file_obj.read())
    return b"mozLz40\0" + compressed


def get_argparser():
    p = argparse.ArgumentParser(
        description="MozLz4a compression/decompression utility"
    )

    p.add_argument(
        "-d", "--decompress", "--uncompress",
        action="store_true",
        help="Decompress the input file instead of compressing it."
    )

    p.add_argument(
        "in_file",
        type=BinFileArg("r"),
        help="Path to input file. `-' means standard input."
    )
    p.add_argument(
        "out_file",
        type=BinFileArg("w"),
        nargs="?",
        default="-",
        help="Path to output file. `-' means standard output (and is the default)."
    )

    return p


def main():
    args = get_argparser().parse_args()

    try:
        with args.in_file as fh:
            if args.decompress:
                data = decompress(fh)
            else:
                data = compress(fh)
    except Exception as e:
        print(
            "Could not compress/decompress file `%s': %s" % (
                args.in_file.name,
                e
            ),
            file=sys.stderr
        )
        sys.exit(4)

    try:
        with args.out_file as fh:
            fh.write(data)
    except Exception as e:
        print(
            "Could not write to output file `%s': %s" % (
                args.out_file.name,
                e
            ),
            file=sys.stderr
        )
        sys.exit(5)


if __name__ == "__main__":
    sys.exit(main())

Script updated:

• Now works with the latest 4.x release of the Python lz4 package.
• Compression/decompression to/from stdout/stdin is now supported.

Can anyone explain how to use this script to add a search engine to the lz4 file, compress and use it in Firefox? Thanks!
I have all dependencies installed, including pip and lz4, yet the only thing I get after running ./mozlz4a.py search.json.mozlz4 is this

@simurq I'd say the script is not enough for parsing the search engines.

Here is an HTML page under a free license you can save that does it for you (but also save the linked JS files):

https://www.jeffersonscher.com/ffu/searchjson.html

FYI I have a tool that can process the search engines correctly: https://github.com/lilydjwg/mozlz4-tool

it's 6 am and i can not understand why simple { echo 0x184D2204 | xxd -r ; tail -c+9 previous.jsonlz4 ; } | lz4 -dc or anything like that would not work

God, if I could upvote this a hundred times I would. Why would anyone design a NEW file format (jsonlz4) for something that we already have so many STANDARD, well-defined formats for? Why does Mozilla require end-users to download, compile, test different potential hacks to get their data out of a custom format? Just use a STANDARD format!

The reason is because they started using it before the lz4 standard was finished. So that's why it's non-standard lz4. Lz4 seems like an OK choice, other than their use of it in a non-standard way. It's very fast to compress/decompress. And why not update it to use standard lz4 now?

I definitely agree it's a PITA to work with though. I wanted to do it in java but was unable to, so now I have to use python.

The reason is because they started using it before the lz4 standard was finished. So that's why it's non-standard lz4. Lz4 seems like an OK choice, other than their use of it in a non-standard way. It's very fast to compress/decompress. And why not update it to use standard lz4 now?

Yes, a mundane reason does exist, acting as the primary dependency for sustaining the real reason for a questionable practice. In a similarly redundant fashion, the questionable nature of the practice is, in itself, purely a thought exercise because the answer would always be the mundane reason, unless you're a disgruntled employee that still represents your employer, but this isn't a likely candidate for such a scenario to involve, because the majority of the public would stop listening about 5 words past "mozlz4 file format". In short, you are doing Mozilla a service by offering the mundane reason externally.

Now, if you consider that the only way Mozilla generates a positive revenue from Firefox is by making Google the default search engine, and the fact that this would be paid not based upon the simple fact that a Google employee downloads Firefox to confirm that it is the default, but rather by users with a Firefox user agent conducting Google searches, it becomes quite clear that profit is gained by making it difficult to change the default search engine, because this is enough to disuade the majority of users who ever have a moment where they feel like using a different search engine.

Also, they use this non-standard text compression format for this to effectively compress a 10kb file into a 5kb file sitting in a folder alongside 19 JSON files, 1 LZ4 file, 7 txt files, 10 sqlite files, 6 db files, and 2 js files where this file is the one and only mozlz4 file among them.

You can call this circumstantial evidence, but you might want to ask yourself:

1. Are you a courtroom?
2. What else are you allowing to slide this easily?

They also use lz4 for the session file containing tabs and history, which can be much larger. Getting access to the contents of that file is why I'm interested in lz4 at all. Although I find your conspiracy theory interesting, and it may be true, I'm not really interested about default search engines.

Not sure what you mean by, "are you a courtroom?" Yes I am. Or no I'm not. Who cares.

Also, you need to chill out.