Unicode Steganography: The Complete Technical Deep Dive

An exhaustive technical exploration into how modern text layout engines (CoreText, HarfBuzz, DirectWrite) handle non-printing Unicode codepoints and how covert channels are established in UTF-8 strings.

1. The Unicode Architecture & Character Pipelines

To understand digital text steganography, one must understand how digital text travels from an application buffer to physical screen pixels:

1. Character Representation (Codepoints): Characters are abstract integers defined in the Unicode Standard (e.g., 'A' = U+0041).
2. Byte Encoding (UTF-8): In UTF-8, codepoints are encoded into 1 to 4 bytes. Plane 14 tag codepoints (U+E0000 to U+E007F) encode into exactly 4 bytes each: 0xF3 0xA0 0x80 ....
3. Shaping & Glyph Mapping (HarfBuzz / CoreText): The layout engine analyzes the codepoint stream, applies OpenType substitutions and ligatures, and queries the font file for glyph outlines.
4. Rasterization: If a codepoint is defined as non-printing or lacks a visual outline, the rasterizer allocates an advance width of zero pixels.

2. Channel Capacity & Overhead

Because each Plane 14 tag codepoint requires 4 bytes in UTF-8, encoding a 100-character ASCII secret produces 400 bytes of tag characters. When attached to a single emoji (typically 4 bytes), the resulting string is 404 bytes in memory, yet visually identical to the single 4-byte emoji.

Modern mobile messaging buffers handle strings of tens of thousands of bytes with ease, making Unicode steganography one of the most efficient, lightweight covert channels in computer science.