md: inline HTML

[cgilite] / markdown.awk

diff --git a/markdown.awk b/markdown.awk
index 98a3e8b2bda338bc58d531a4c074bcbbafaf498a..9c548ecf99fe43b5cfa4d051b11b22a0bd484d62 100755 (executable)
--- a/markdown.awk
+++ b/markdown.awk
@@ -24,6 +24,7 @@
 # - [x] Lists (ordered, unordered)
 # - [x] Code blocks (using indention)
 # - [x] Horizontal rules
+# - [x] Verbatim HTML block (gfm)
 #
 # Basic Markdown - Inline elements:
 # ---------------------------------
@@ -34,6 +35,8 @@
 # - [x] Images / reference style images
 # - [x] <automatic links>
 # - [x] backslash escapes
+# - [x] Verbatim HTML inline
+# - [ ] HTML escaping
 #
 # Extensions - Block elements:
 # ----------------------------
@@ -100,6 +103,11 @@ function inline( line, LOCAL, len, code, href, guard ) {
   } else if ( match(line, /^  \n/) ) {
     return "<br />\n" inline( substr(line, RLENGTH + 1) );
 
+  # Verbatim inline HTML
+  } else if ( match( line, /^(<!--([^-]|-[^-]|--[^>])*-->|<\?([^\?]|\?[^>])*\?>|<![A-Z][^>]*>|<!\[CDATA\[([^\]]|\][^\]]|\]\][^>])*\]\]>|<\/[A-Za-z][A-Za-z0-9-]*[[:space:]]*>|<[A-Za-z][A-Za-z0-9-]*([[:space:]]+[A-Za-z_:][A-Za-z0-9_\.:-]*([[:space:]]*=[[:space:]]*([[:space:]"'=<>`]+|"[^"]*"|'[^']*'))?)*[[:space:]]*\/?>)/) ) {
+    len = RLENGTH;
+    return substr( line, 1, len) inline(substr(line, len + 1));
+
   #  ``code spans``
   } else if ( match( line, /^`+/) ) {
     len = RLENGTH
@@ -197,7 +205,7 @@ function inline( line, LOCAL, len, code, href, guard ) {
     return "<strong>" inline( substr( line, 3, len - 4 ) ) "</strong>" inline( substr( line, len + 1 ) );
 
   #  __strong__
-  } else if ( match(line, "^__(([^_[:space:]]|" ieu ")|([^_[:space:]]|" ieu ")(" nu "|" ieu ")*([^_[:space:]]|" ieu "))__[[:space:]]") ) {
+  } else if ( match(line, "^__(([^_[:space:]]|" ieu ")|([^_[:space:]]|" ieu ")(" nu "|" ieu ")*([^_[:space:]]|" ieu "))__[[:space:][:punct:]]") ) {
     len = RLENGTH;
     return "<strong>" inline( substr( line, 3, len - 5 ) ) "</strong>" inline( substr( line, len) );
 
@@ -212,7 +220,7 @@ function inline( line, LOCAL, len, code, href, guard ) {
     return "<em>" inline( substr( line, 2, len - 2 ) ) "</em>" inline( substr( line, len + 1 ) );
 
   #  _em_
-  } else if ( match(line, "^_(([^_[:space:]]|" isu ")|([^_[:space:]]|" isu ")(" nu "|" isu ")*([^_[:space:]]|" isu "))_[[:space:]]") ) {
+  } else if ( match(line, "^_(([^_[:space:]]|" isu ")|([^_[:space:]]|" isu ")(" nu "|" isu ")*([^_[:space:]]|" isu "))_[[:space:][:punct:]]") ) {
     len = RLENGTH;
     return "<em>" inline( substr( line, 2, len - 3 ) ) "</em>" inline( substr( line, len ) );
 
@@ -233,6 +241,28 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent ) {
   if ( block == "" ) {
     return "";
 
+  # HTML #2 #3 #4 $5
+  } else if ( match( block, /(^|\n) ? ? ?(<!--([^-]|-[^-]|--[^>])*(-->|$)|<\?([^\?]|\?[^>])*(\?>|$)|<![A-Z][^>]*(>|$)|<!\[CDATA\[([^\]]|\][^\]]|\]\][^>])*(\]\]>|$))/) ) {
+    len = RLENGTH; st = RSTART;
+    return _block(substr(block, 1, st - 1)) substr(block, st, len) _block(substr(block, st + len));
+
+  # HTML #6
+  } else if ( match( tolower(block), /(^|\n) ? ? ?<\/?(address|article|aside|base|basefont|blockquote|body|caption|center|col|colgroup|dd|details|dialog|dir|div|dl|dt|fieldset|figcaption|figure|footer|form|frame|frameset|h[123456]|head|header|hr|html|iframe|legend|li|link|main|menu|menuitem|nav|noframes|ol|optgroup|option|p|param|section|source|summary|table|tbody|td|tfoot|th|thead|title|tr|track|ul)([[:space:]\n>]|\/>)([^\n]|\n[ \t]*[^\n])*(\n[[:space:]]*\n|$)/) ) {
+    len = RLENGTH; st = RSTART;
+    return _block(substr(block, 1, st - 1)) substr(block, st, len) _block(substr(block, st + len));
+
+  # HTML #1
+  } else if ( match( tolower(block), /(^|\n) ? ? ?<(script|pre|style)([[:space:]\n>]).*(<\/script>|<\/pre>|<\/style>|$)/) ) {
+    len = RLENGTH; st = RSTART;
+    match( tolower(substr(block, st, len)), /(<\/script>|<\/pre>|<\/style>)/);
+    len = RSTART + RLENGTH;
+    return _block(substr(block, 1, st - 1)) substr(block, st, len) _block(substr(block, st + len));
+
+  # HTML #7
+  } else if ( match( block, /^ ? ? ?(<\/[A-Za-z][A-Za-z0-9-]*[[:space:]]*>|<[A-Za-z][A-Za-z0-9-]*([[:space:]]+[A-Za-z_:][A-Za-z0-9_\.:-]*([[:space:]]*=[[:space:]]*([[:space:]"'=<>`]+|"[^"]*"|'[^']*'))?)*[[:space:]]*\/?>)([[:space:]]*\n)([^\n]|\n[ \t]*[^\n])*(\n[[:space:]]*\n|$)/) ) {
+    len = RLENGTH; st = RSTART;
+    return substr(block, st, len) _block(substr(block, st + len));
+
   # Horizontal rule
   } else if ( match( block, /(^|\n) ? ? ?((\* *){3,}|(- *){3,}|(_ *){3,})($|\n)/) ) {
     len = RLENGTH; st = RSTART;