Morse Code Translator

Depth: how the encode direction and separators actually behave

When you click Convert, the translator reads whatever sits in the input pane and decides the direction for you: mostly-letters text encodes to Morse, while a string made mostly of dots, dashes, spaces, and slashes decodes back to letters, so you rarely have to flip a switch by hand. In the encode direction, a single space separates the code for two letters inside the same word and a forward slash marks the gap between words, which matters when you compare output against a printed key that runs everything together - "SOS OK" becomes "... --- ... / --- -.-" with exactly one slash, not a blank run you have to count. Any character the ITU table does not cover, such as an emoji or a curly bracket, is written as a single "?" rather than silently dropped, so a "?" appearing in your Morse output is a signal that one source character had no defined code, not a bug. Empty input is treated as a no-op: clicking Convert on a blank pane leaves the output pane untouched instead of clearing prior work, which lets you re-run the same text after a small edit without losing the last result. Because the whole transform is a local string operation, a 5,000-character message and a 5-character one both finish in the same click with no upload step, and the decode path tolerates ragged spacing - two or three spaces between groups still parse the same as one.