BIRSAx2/mdcat mdcat-2.13.0

on GitHub

Added

• Custom heading markers and GFM alert icons/labels via text in a
  [theme.styles] entry in ~/.config/mdcat/config.toml: set h2-h6 to
  override the marker written before heading text, or alert_note-
  alert_caution to override the icon and label written for that alert
  (GH #23).
• pulldown-cmark-mdcat: Theme gained h2_marker-h6_marker and
  alert_note_label-alert_caution_label fields (all public Strings)
  controlling the text previously hardcoded in the renderer.
• --toc prints a numbered table of contents, generated from the document's
  headings and nested by level, before its content. Entries link to
  FILE#slug (a GitHub-style anchor) via OSC 8 so that terminals and later
  viewers that resolve such anchors can jump to that heading; on standard
  input entries are plain text, since there's no file to link to.
• --image-protocol/$MDCAT_IMAGE_PROTOCOL/defaults.image_protocol in
  ~/.config/mdcat/config.toml forces a specific inline image protocol
  (none, iterm2, kitty, or sixel), overriding terminal auto-detection.
  Useful inside tmux/screen, where the outer terminal's capabilities usually
  aren't visible to auto-detection. Precedence: --image-protocol wins over
  $MDCAT_IMAGE_PROTOCOL, which wins over the config default.
• --tabs/defaults.tabs in ~/.config/mdcat/config.toml expands literal
  tabs in the input to spaces, using the given tab stop width, before
  parsing. Off by default, so tabs pass through unchanged; without this, a
  tab inside text content throws off line-wrapping and alignment width
  calculations, since terminals render it as jumping to the next tab stop
  rather than occupying a single column.

Fixed

• Syntax-highlighted fenced code blocks nested inside list items (or other
  indented blocks) lost their indentation on every line after the first,
  because only untagged (non-highlighted) code blocks re-applied indentation
  after each line.
• Inline images (badges, icons) rendered too small on Sixel and iTerm2, only
  filling part of their allotted terminal row, since they were sent at
  native pixel resolution rather than scaled to a whole number of cell
  rows. Sixel images also drifted into a descending staircase when several
  appeared in a row (e.g. consecutive README badges), and iTerm2 never
  applied any explicit sizing at all (GH #22).