# Status: # [.] Inline HTML # [ ] Autoescape # [.] Block elements # [x] Paragraph # [x] Line break # [x] Headers (atx and setex) # [x] Blockquotes # [ ] Lists # [ ] Horizontal rules # [x] Code blocks # [ ] Span elements # [ ] Links # [ ] Inline # [ ] Reference # [ ] Emphasis # [ ] Code # [ ] Images # [ ] Inline # [ ] Reference # [ ] Other # [ ] Autolinks # [ ] Escapes # # [ ] Non-standard features # [ ] Strikethrough ~~...~~ # [ ] Superscript ^...^ # [ ] Subscript ~...~ # [ ] Footnote [^...] (with user customizations) # [ ] Task list - [ ] (readonly) # [ ] Definition lists # [x] Single line # [ ] Multi-line # [ ] Fenced code blocks ```lang ``` # [ ] Tables # [ ] Links # [ ] Link destination ( inside anchors) # [ ] Nested links # No character classes, since they aren't supported on every system # # Limitations: # - definition lists always contain one element, i.e. multiple definitions # aren't merged into the same list # - emphasis which isn't surrounded by blanks is still treated as emphasis # e.g. "some_text_here" will become "sometexthere" # - emphasis inside inline code block is still converted to HTML # - backticks inside `` emphasis blocks aren't escaped function empty(value) { return value == "" } function _next() { __buffered_print() not_paragraph = 1 next } { not_paragraph_p1 = not_paragraph not_paragraph = 0 empty0 = empty($0) emptyP1 = empty(P1) } # Output is buffered on multiple lines, held in these variables: # P2 = previous non-empty line # P1 = previous line # $0 = current line # N1 = next new line (N1 is always set by the user) # Inline HTML /^[ \t]*<\/?[0-9A-Za-z]+/ { _next() } # Blockquotes blockquote_level > 0 && empty0 { for (P1 = P1 ORS; blockquote_level > 0; --blockquote_level) { P1 = P1 "" } not_paragraph_p1 = 1 } /^>[ \t]/ { match($0, /^[> \t]+[ \t]+/) level = substr($0, RSTART, RLENGTH) $0 = substr($0, RSTART + RLENGTH) gsub(/[ \t]/, "", level) level = length(level) for (;blockquote_level < level; ++blockquote_level) { P1 = P1 "
" } if (blockquote_level > level) { P2 = P2 ORS } for (;blockquote_level > level; --blockquote_level) { P2 = P2 "" } not_paragraph = 1 } /^>$/ { $0 = ""; not_paragraph = 1 } # Code blocks /^ / { if (!inside_codeblock) { P1 = P1 "
"
inside_codeblock = 1
}
_next()
}
inside_codeblock && ! /^[ \t]*$/ {
if (emptyP1) {
P2 = P2 ORS ""
}
else {
P1 = P1 ORS ""
}
inside_codeblock = 0
emptyP1 = 1
}
# Line break
/ $/ {
sub(/[ \t]* $/, "")
}
# atx-style headers
/^[ \t]*#+/ {
match($0, /#+/)
if (RLENGTH <= 6) {
## Remove prefix/suffix
gsub(/^[ \t]*#+[ \t]*/, "")
gsub(/[ \t]*#+[ \t]*$/, "")
$0 = "" $0 } !emptyP1 && empty0 && !not_paragraph_p1 { P1 = P1 "
" } # After all $0 parsing { __buffered_print() } function __buffered_print() { empty0 = empty($0) emptyP1 = empty(P1) if (empty0 && emptyP1) { ++P1_repeats } else if (!empty0 && emptyP1) { P1 = $0 } else { print P2 while (P1_repeats > 0) { print "" --P1_repeats } P2 = P1 P1 = $0 if (empty0) { P1_repeats = 1 } } if (!empty(N1)) { if (emptyP1) { P1 = N1 } else { print P2 while (P1_repeats > 0) { print "" --P1_repeats } P2 = P1 P1 = N1 N1 = "" } } } END { print P2 while (P1_repeats > 0) { print "" --P1_repeats } if (!empty(P1)) { print P1 } print if (!empty(N1)) { print N1 } }