Run-length encoding

Run-length encoding is the bridge between the two halves of this course: it is lossless compression and a traceable program in one artifact. Walk a row of pixels, count each run of one colour, and store it as a (length, colour) pair. Watch out fill up — and watch the most common bug quietly drop the last run.

Compress a row of pixels by replacing each run of the same colour with a (length, colour) pair.

Run-length encoding — the forgotten last run execution-derived · CPython

8    else:

variables

state after line 1 runs

variable	value
`pixels`	`["W","W","W","B","B","W"]`

step 1 / 24

Click a line, drag the slider, or use the ← → keys.

What you are looking at

Six pixels — W W W B B W — should compress to three pairs:[3, W] [2, B] [1, W]. The encoder only saves a run when the colour changes, so the very last run needs one extra save after the loop.

Buggy — saves a run only on a colour change, and the loop ends on a colour that never changes again. The final [1, W] is never saved; out ends with two pairs instead of three. Decode it and the image comes back wrong.
Procedural — the same loop plus one out.append after it, to flush the last run. Now out holds all three pairs and round-trips back to the original row.

Every state you see came from running the program under CPython's tracer at build time — see how GlassBox stays honest.