# dollcode

[](https://github.com/isolyth/dollcode-ts)

January 25, 2026 project utilities

A base-3 text encoder using Unicode block characters. Each character's codepoint is encoded
using digits 3 mapped to ▖▘▌.

Encode Decode

Text to encode

Output

Copy to clipboard

▶ Custom characters

## About

This is a recreation of the text encoding functionality from the original dollcode site
(dollcode.v01dlabs.sh) that no longer exists. It was reverse-engineered using:

- The [dollcode.rs](https://codeberg.org/ember-ana/dollcode.rs) Rust crate (which only supports number encoding)
- Information from [noe.sh/dollcode](https://noe.sh/dollcode) about the base-3 encoding system
- Analysis of example encoded strings to determine the text encoding algorithm

Dollcode is a trinary (base-3) encoding system using Unicode block characters. The original
site supported encoding both pure numbers and text, but used different encoding schemes for
each. This implementation focuses on the text-compatible version.

## How It Works

- Each character is converted to its Unicode codepoint
- The codepoint is encoded using base-3 with digits {1,2,3}
- Digits are mapped to characters: 1→▖, 2→▘, 3→▌
- Character groups are separated by zero-width joiners (U+200D)

Example: Hello → ▘▖▘▌‍▌▖▌▘‍▌▘▘▌‍▌▘▘▌‍▌▘▌▌‍

## Technical Details

- Default Characters: ▖ (U+2596), ▘ (U+2598), ▌ (U+258C)
- Default Separator: Zero-width joiner (U+200D)
- Encoding: Base-3 system with digits {1,2,3} instead of {0,1,2}
- Character Order: Most significant bit first (MSB-first)

## Differences from Number-Only Encoding

The Rust crate at [dollcode.rs](https://codeberg.org/ember-ana/dollcode.rs) implements
number-only encoding, which produces different output than text encoding. This implementation
focuses on the text-compatible version that was available on the original dollcode.v01dlabs.sh site.

[LLM-readable markdown version of this page](/projects/dollcode/llms.txt)