Shell/Prompt/Concept/PS1 vs PS2

• PS1 vs PS2 (Diátaxis overview)

  • Conceptual overview

    • In interactive shells, a prompt is the UI that signals “the shell is ready for input”.
    • Most shells support multiple “prompt states” depending on whether the shell is:
      • ready for a new command
      • waiting for you to finish the current command
    • PS1 and PS2 are the classic names for these states in Bourne-style shells, and the concepts map cleanly to many other shells even if the exact variable names differ.

  • Key idea

    • PS1 = “new command” prompt
    • PS2 = “continuation / incomplete command” prompt

  • Reference model (mental model)

    • Think of the shell as having a tiny parser running while you type.
    • If the parser decides “I have a complete command”, you see PS1.
    • If the parser decides “this isn’t complete yet; I need more input”, you see PS2.

  • PS1: Primary prompt

    • Purpose
      • Signals the shell is ready for a fresh command.
    • Typically includes context
      • username / hostname
      • current working directory
      • git branch / dirty state
      • exit status of the last command
      • active environment (venv / conda / nix / etc.)
    • Practical effect
      • You spend most of your interactive time looking at PS1.

  • PS2: Secondary (continuation) prompt

    • Purpose
      • Signals the shell is waiting for you to finish the current command.
    • When it appears (common cases)
      • unmatched quotes (single or double)
      • a trailing line continuation \
      • incomplete structures (e.g., if, case, { ... }, subshells, etc.)
      • multiline commands (some heredocs / command substitutions depending on shell)
    • Typical style
      • intentionally minimal (often just > ) so it visually reads as “you’re still in the middle of something”.

  • Relationship to Diátaxis

    • Tutorial (learning path)

      • Learn to notice when you’re in PS2 mode so you don’t think the terminal “hung”.
      • Practice: intentionally trigger PS2 with a multiline command, then close it properly.

    • How-to guide (task-focused)

      • Configure PS1 to include useful context (cwd, git status, etc.).
      • Keep PS2 short and obvious so continuation is visually clear.

    • Reference (lookup)

      • PS1 = primary prompt string
      • PS2 = continuation prompt string
      • Both are expanded/rendered by the shell according to its prompt-expansion rules.

    • Explanation (deeper reasoning)

      • PS1 and PS2 exist because interactive input is sometimes multi-step.
      • PS2 is a UX affordance: it tells you “the shell is still collecting input for the same command”.

  • Quick examples

    • Example: PS1 (new command)
      • user@host ~/Documents/GitHub main %
    • Example: PS2 (unfinished quote)
      • echo "hello
      • > world"
    • Example: PS2 (line continuation)
      • echo hello \
      • > world

  • Practical heuristics

    • If you see PS2:
      • check for an unclosed quote
      • check for unclosed parens/braces
      • check whether you ended the previous line with \
    • If you want to bail out:
      • Ctrl-C usually cancels the current partial command and returns to PS1