visual-cs

LZ77 Compression — Sliding Window

← All visualizations

LZ77 Compression Explorer

Watch the sliding window scan the input character by character. The search buffer holds already-encoded data; the lookahead buffer holds what's next. Each step emits a (offset, length, next-char) token.

12
6
Speed
Step —
Full input string
Window view
Search buffer (already encoded)
Lookahead buffer
Run compression to begin…
Output tokens
#
Match in buffer
Offset
Length
Next
Decoded segment
Input size
—
characters
Tokens emitted
—
triples
Chars via back-ref
—
copy from buffer
Compression ratio
—
tokens / chars

Legend

Already processed
Search buffer
Current position
Lookahead buffer
Best match found

Algorithm

1
Slide the window: search buffer holds the last S encoded chars.
2
Look ahead up to L chars and find the longest match in the search buffer.
3
Emit (offset, length, next) — offset back from current, length of match, and the first unmatched char.
4
Advance the window by length + 1 characters.
5
Repeat until the input is exhausted.

Compression

Tokens emitted 0
vs Input 0
Fewer tokens = better compression. Tokens include back-references that expand on decode.

Presets