ZLib VS Snappy

So the real issue here is that the lack of tree validation before the tree construction, I believe. I'm surprised that this check was not yet implemented (I actually checked libwebp to make sure that I was missing one). Given this blind spot, an automated test based on the domain knowledge is likely useless to catch this bug.

[1] https://github.com/madler/zlib/blob/master/examples/enough.c

In the source code of the Node.js opensource project, lib folder contains JavaScript code, mostly wrappers over C++ and function definitions. On the contrary, src folder contains C++ implementations of the functions, which pulls dependencies from the V8 project, the libuv project, the zlib project, the llhttp project, and many more - which are all placed at the deps folder.

A minimal viable Deflate decompressor is not exactly complex[1], although slower than mainline zlib.

[1] https://github.com/madler/zlib/tree/master/contrib/puff

ZLibrary not zlib for anyone worried about archiving algorithms. https://zlib.net/ and all the Linux repositories are safe (AFAIK).

>The following is my educated guess at what to do next and I am a week-end warrior programmer at best. If someone with more experience comments, I would listen to them over me< If you have no programming experience, this may be a lost cause. It looks like the save file has a header area followed by a list of "zlib" compressed chunks. zlib.net may have a program that can open the save file and let you see the raw information inside it. Something after the header is garbled and it is preventing Save-Editors from decompressing that chunk OR from parsing what is inside of it. If you are lucky, the garbled bit will jump out at you and you can repair it.

These appears to be the relevant changes:

2022-07-30: https://github.com/madler/zlib/commit/eff308af425b67093bab25...

2022-08-08: https://github.com/madler/zlib/commit/1eb7682f845ac9e9bf9ae3...

The second commit definitely fixed a null pointer dereference, I am not sure if the CVE is referencing something else that was fixed by the first commit.

It is actually possible to estimate the compression level without actually compressing everything again, because different levels use different strategies which can be identified in some cases. zlib in particular has three strategies [1] and preflate [2] leverages this to store the deflate stream reconstruction data without much bits.

[1] https://github.com/madler/zlib/blob/21767c6/deflate.c#L134-L...

[2] https://github.com/deus-libri/preflate/blob/master/preflate_...

It's still a heap of old school C spaghetti https://github.com/madler/zlib/commit/eff308af425b67093bab25...

Another example of Nim being really fast is the supersnappy library. This library benchmarks faster than Google’s C or C++ Snappy implementation.

It doesn't destroy performance for the simple reason that nowadays memory access has higher latency than pure compute. If you need to use compute to produce some data to be stored in memory, your overall throughput could very well be faster than without compression.

There have been a large amount of innovation on fast compression in recent years. Traditional compression tools like gzip or xz are geared towards higher compression ratio, but memory compression tends to favor speed. Check out those algorithms:

* lz4: https://lz4.github.io/lz4/

* Google's snappy: https://github.com/google/snappy

* Facebook's zstd in fast mode: http://facebook.github.io/zstd/#benchmarks