Dancing Men cipher

Encode mode walks the input, keeps alphabetic characters, skips whitespace, emits space-separated figure tokens, and lists any non-letter symbols in unmapped_letters so you can see what did not enter the cipher. Decode mode splits tokens on whitespace and common separators, rebuilds the message from known poses, and reports unknown tokens in unmapped_figures instead of inventing letters. Empty input, a string with no A–Z letters on encode, or a decode stream with no recognized figures fails with invalid_input. Use it free in the browser for literary-cipher lessons, Holmes puzzle worksheets, and reversible figure demos, or call the API at $0.002 per successful request when pipelines need reproducible Dancing Men transforms.

What the Dancing Men cipher is and what this tool returns

The Dancing Men cipher is the stick-figure monoalphabetic code that appears in Conan Doyle’s Holmes adventure of the same name: each letter of a secret message is drawn as a tiny man in a distinct pose, and Holmes reconstructs the alphabet from frequency and cribs just as he would for any simple substitution. Each token follows a compact grammar: face left or right, left arm up horizontal down or bent, right arm with the same four options, and legs together, apart, stride-left, or stride-right. That grammar yields far more than twenty-six combinations; the endpoint uses one stable assignment for A through Z so H always becomes the same figure string and that string always returns H. Historically the men were ink drawings on paper; this endpoint only exposes the letter-to-description map so teachers, puzzle authors, and test authors can work with the codes without shipping image assets. The JSON response always includes result for the transformed string, encoding fixed to dancing_men, mode set to encode or decode, input_length and output_length for the raw strings you sent and received, letter_count and figure_count for how many A–Z letters or pose tokens were produced or consumed, unmapped_letters and unmapped_figures arrays for symbols or tokens the table could not map, and a formula string that restates the pose grammar for documentation and automated checks. There is no streaming path and no silent fallback to Baconian, Morse, or a Vigenère keyword: either the payload is valid for the chosen mode, or the request fails with invalid_input. If you need polyalphabetic ciphers, ROT variants, or modern authenticated encryption, use those dedicated endpoints; this landing answers only the Dancing Men stick-figure contract that literary-cipher chapters already describe.

How text and mode are read and how unmapped items are reported

still yields figures for H and I while reporting the exclamation mark. In decode mode the stream is split on spaces, tabs, newlines, commas, semicolons, and pipes; each non-empty token is looked up in the inverse table. The preferred payload fields are text for the message or figure stream and optional mode for encode or decode. Text accepts aliases such as input, value, s, message, plaintext, and ciphertext so forms and SDK clients can reuse existing column names. Mode defaults to encode when omitted and also accepts short aliases such as enc, encrypt, e for encode and dec, decrypt, d for decode; any other string fails validation. In encode mode only alphabetic letters are kept and mapped, case is ignored on the way in, whitespace is skipped without being treated as an error, and the result is space-separated pose tokens; characters that are neither letters nor whitespace are collected in order into unmapped_letters so a string like HI! If no letter remains after filtering, the engine rejects the request with invalid_input so a digit-only string does not look like a successful empty cipher. Known tokens append an uppercase letter; unknown tokens append nothing to result and are listed in unmapped_figures so you can see which figures failed recognition. If every token is unknown, the request fails with invalid_input rather than returning an empty success. Missing text, empty text, non-string types, and unknown modes likewise raise invalid_input. Pair the free browser run on this page with the priced API when you need the identical mathematical result inside an automated pipeline. Invalid inputs return invalid_input and are not charged, which keeps validation cheap while successful transforms remain a fixed item price.

Where Dancing Men fits in lessons, puzzles, and test fixtures

Product and education tools use the transform when a learner pastes a figure stream and the UI must show both the recovered letters and any unmapped_figures under the documented table. On this page the work runs entirely in your browser; through the API each successful lookup costs $0.002, failed validation is not charged, and nothing is stored after the response returns. That split — free interactive exploration and a priced, identical API — matches the rest of the KIT tier-A catalog and lets you move from a classroom demo to an automated pipeline without changing the transformed result.