What kind of compression algorithm is this? [closed]

Question

So I have a couple of compressed files, and the uncompressed versions as well. I do not have the software that created these files originally. I'm trying to figure out what the underlying algorithm is -- can you figure it out? Initially, I thought it might be some LZW variant, but I'm not sure. The data seems to make more sense when broken up into 6-bit words -- I see lots of repeating patterns then.

The two files are very similar, and the uncompressed versions differ by only a few bytes -- that could help establish where those differing bytes are in the compressed files. I've highlighted the differences.

Compressed file #1:

02 02 01 17 0E 11 92 14 C0 55 52 44 FF BC AE 47 DB E1 05 42 F8 70 DE 57 23 FF
54 1A 55 3D BF 54 10 E3 38 0C B2 FB C4 92 1C 20 DE 57 23 FF 54 1A 55 3D BE 5E
4C 96 B2 0E 32 80 CB 2F BC 48 70 83 79 5C 8F FD 50 69 54 F6 F9 96 48 A9 07 19
C2 30 F0 E1 BC AE 47 FE A8 34 AA 7B 7E 32 BF E5 1F EE A8 48 CA 11 87 87 0D E5
72 3F F5 41 A5 53 DB E5 24 5D F8 CA FF 4C B1 13 8C 71 18 7B C3 86 F2 B9 1F FA
A0 D2 A9 ED FD 55 97 BA 22 32 C0 CB 2F BC

Compressed file #2:

02 02 01 17 0E 11 92 14 C0 55 52 44 FF BC AE 47 DB E1 05 42 F8 70 DE 57 23 FF
54 1A 55 3D BF 54 10 E3 38 0D 36 D4 04 92 1C 20 DE 57 23 FF 54 1A 55 3D BE 5E
4C 96 B2 0E 32 80 D3 6D 40 48 70 83 79 5C 8F FD 50 69 54 F6 F9 96 48 A9 07 19
C2 30 F0 E1 BC AE 47 FE A8 34 AA 7B 7E 32 BF E5 1F EE A8 48 CA 11 87 87 0D E5
72 3F F5 41 A5 53 DB E5 24 5D F8 CA FF 4C B1 13 8C 71 18 7B C3 86 F2 B9 1F FA
A0 D2 A9 ED FD 55 97 BA 22 32 C0 D3 6D 40

Uncompressed file #1:

20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 2A 2A 2A 2A 2A 20 47 52 41 4E 44 20 54 4F 54 41 4C 53 20
2A 2A 2A 2A 2A 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44
53 20 52 45 41 44 20 20 20 20 20 20 20 20 20 20 20 20 20 20 32 32 38 37 0D 0A
0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 42 59 50
41 53 53 45 44 20 20 20 20 20 20 20 20 20 20 32 32 38 37 0D 0A 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 43 48 41 4E 47 45 44 20 20 20
20 20 20 20 20 20 20 20 20 20 20 30 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C
20 52 45 43 4F 52 44 53 20 4E 4F 54 20 4F 4E 20 58 52 45 46 20 20 20 20 20 20
20 20 20 20 30 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44
53 20 42 41 4E 4B 20 4E 4F 54 20 46 4F 55 4E 44 20 20 20 20 20 20 20 30 0D 0A
0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 57 52 49
54 54 45 4E 20 20 20 20 20 20 20 20 20 20 20 32 32 38 37

Uncompressed file #2:

20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 2A 2A 2A 2A 2A 20 47 52 41 4E 44 20 54 4F 54 41 4C 53 20
2A 2A 2A 2A 2A 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44
53 20 52 45 41 44 20 20 20 20 20 20 20 20 20 20 20 20 20 20 33 34 33 39 0D 0A
0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 42 59 50
41 53 53 45 44 20 20 20 20 20 20 20 20 20 20 33 34 33 39 0D 0A 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 43 48 41 4E 47 45 44 20 20 20
20 20 20 20 20 20 20 20 20 20 20 30 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C
20 52 45 43 4F 52 44 53 20 4E 4F 54 20 4F 4E 20 58 52 45 46 20 20 20 20 20 20
20 20 20 20 30 0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44
53 20 42 41 4E 4B 20 4E 4F 54 20 46 4F 55 4E 44 20 20 20 20 20 20 20 30 0D 0A
0D 0A 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20 20 54 4F 54 41 4C 20 52 45 43 4F 52 44 53 20 57 52 49
54 54 45 4E 20 20 20 20 20 20 20 20 20 20 20 33 34 33 39

As you can see, the output files are just plain ASCII text files. Any ideas?

Answer 1

This seems to be some proprietary encoding format, designed to shave some bits of specific types of messages.

It operates on 8 bit (ascii) input and outputs a bit stream using a mixture of a 5 and 6 bit token set, including some control characters.

The following tokens can be identified:

// 5 bit tokens:
00000 switch to 6 bit mode
00011 take the following 6 bits as N, and output N spaces
00100 A
00101 B
.....
11101 Z
11110 <crlf>
11111 space

// 6 bit tokens:
000001 switch to 5 bit mode
000011 take the following 6 bits as N, and output N spaces
001001 <crlf>
011000 1
011001 2
......
100000 9

// pure speculation:
010111 0
010010 *
000110 repeat the next 6 bit char N times
001100 space
00001 skip 3 bits, take the next 8 bits as ascii, and output N times

Without more examples it is hard to determine what happens at the beginning of the stream. It might be some magic value, or could contain some control values.