Letter to Number Translator

Enter letters in the main source box to produce their numbers, or swap the boxes to decode numeric input. Uppercase and lowercase source letters share the same positions, and the decoded result uses uppercase. The word CAB becomes 3-1-2 under the default convention. Check a short word that you already know before decoding a complete message. If a result seems shifted by one letter throughout, compare the zero-based rule in the workbench rather than changing individual letters until the sentence happens to look familiar.

The Letter to Number Translator does not infer a numbering system from the apparent meaning of a message.

Each letter number needs a visible boundary. Hyphens or spaces separate numeric values, while a slash represents a space in the decoded text. The main encoder places hyphens between tokens and uses slash tokens for spaces, making its own output suitable for reverse input. A continuous string such as 123 is rejected rather than guessed as 1-2-3 or 12-3. Both could be valid alphabet positions in a different message. Preserve separators from a worksheet or ask the sender which grouping was intended before trying to solve a missing-boundary puzzle.

In the Letter to Number Translator, numeric separators are not word spaces. Use / when a decoded gap belongs between words.

The normal rule assigns A=1, B=2 and Z=26. A zero-based rule assigns A=0 through Z=25. Reversing the alphabet assigns A the highest selected value and Z the lowest. These are distinct conventions, so the workbench displays the entire selected key next to the result. Use the same key in both directions if you want to check an encoding. Changing a switch after pasting numbers changes their interpretation immediately; it does not mean that the original sender used that key or that the new sentence is the intended solution.

The Letter to Number Translator main boxes remain A1Z26 even after you experiment with a different key in the workbench.

This implementation keeps a small explicit punctuation set: period, comma, exclamation mark and question mark. They are encoded as bracketed tokens such as [!] so they cannot be mistaken for an alphabet position. The reverse reader accepts those exact tokens. Literal digits in a forward message are rejected because a number already has a meaning in the code; silently mixing a source digit with an encoded letter would make the result ambiguous. Other punctuation is also rejected instead of being dropped from a message without a visible explanation.

The Letter to Number Translator uses an inspectable format. If your source uses a different punctuation convention, rewrite its separators deliberately before decoding.

Read every token as a number under the chosen key. An out-of-range token is a reason to inspect the convention, grouping or source transcription, rather than discard that token. Under A1Z26, 27 does not represent a letter. A zero can be meaningful with A=0 but is invalid under the default key. If a puzzle contains coordinates, dates or telephone-keypad groups, this page may be the wrong first step. Compare its instructions with the visible alphabet strip before deciding that every number sequence is an A1Z26 message.

The Letter to Number Translator supplies the mechanical substitution; it cannot identify a hidden cipher merely because a few decoded letters form a word.

Label the numbering rule, alphabet direction and word separator when you share a result. Keep the original plain text if the message will be edited later. The code does not preserve letter case, so reverse decoding alone cannot tell you which letters were originally capitalized. It is also easy for another reader to decode once the convention is known. Use it for puzzles, classroom exercises, word games and checking a number-letter mapping, rather than for private passwords or secure communications. A cleanly labelled code is more useful than a guessed but plausible sentence.

The Letter to Number Translator main source allows 500 text units. Its larger 2,500-unit reverse box accepts the expanded output of a full-length source.