Baudot ITA2 encoder

What Baudot–Murray ITA2 is and how letter and figure shifts work

This endpoint follows that classic state machine: encode walks each character, uppercases Latin letters, inserts LTRS or FIGS when the required case differs from the current case, and appends the five-bit pattern for the character.

Input contract, output fields, and how unmapped characters are reported

For encode, text is ordinary plain text; for decode, text is a stream of ITA2 codes. Encode accepts A–Z in either case, digits zero through nine, space, carriage return, line feed, and the figures punctuation set including hyphen, apostrophe, comma, exclamation, colon, parentheses, plus, question mark, ampersand, period, slash, semicolon, hash, and the pound sign. Decode accepts space-separated five-bit groups such as 00011, decimal integers from zero to thirty-one, hex tokens like 0x03, or a continuous binary string whose length is a multiple of five. POST a JSON body with required string field text. Optional mode defaults to encode and also accepts decode, plus short aliases such as enc, e, to_baudot, dec, d, and from_baudot. Field aliases for the payload include input, value, data, s, message, plaintext, and ciphertext so scripts can reuse existing shapes. Any other character—at-signs, percent, emoji, accented letters outside the ITA2 alphabet—is skipped, recorded once in first-seen order in the unmapped array, and counted in unmapped_count while mapped characters still produce codes. If nothing in the string is mappable, the task fails with invalid_input instead of returning an empty success. Malformed tokens, out-of-range decimals, and continuous binary of non-multiple length fail with invalid_input and are not billed. On success the body includes result (binary groups on encode, plain text on decode), mode, encoding set to baudot-ita2, codes as an array of integers zero through thirty-one including any shift codes inserted, code_count, mapped, unmapped, unmapped_count, shift_count, input_length, output_length, and a short formula string restating the ITA2 shift rule. Successful API calls use the published base price of $0.002 per item; validation failures do not consume balance.

Where a Baudot ITA2 encoder earns its keep in practice

History-of-computing and coding-theory courses use a Baudot ITA2 encoder so students can see why five bits plus two shift codes can cover letters and figures without a sixth bit, then reverse a hand-built stream and check their shift insertions. Amateur radio and vintage Telex restoration projects need a deterministic oracle that produces the same bit groups a mechanical teleprinter would have selected for a given message, including automatic FIGS before a digit run and LTRS when letters resume. Documentation and museum exhibit software generate labeled examples for captions without standing up a full RTTY modem stack. Capture-the-flag and classical-crypto challenges sometimes hide flags in ITA2 binary or decimal dumps; solvers need both directions and a clear invalid_input when a token is not a legal five-bit group.