md2html: do not loosen tight lists with multi-line items

The list-interrupt preprocessor inserted a blank line before any bullet
that followed a non-item, non-4-space-indented line. When a list item
wrapped onto a 2-space-indented continuation line, the *next* item looked
like a paragraph->list transition, so a blank got injected between items
— turning a tight list into a loose one (every <li> wrapped in <p>), as
seen in docs/internals.md "Writing programs in C".

Track list state and only interrupt an actual paragraph: a bullet that
continues an already-open list (or its indented continuation lines) no
longer triggers a blank-line insert. The original paragraph->list
interrupt and the "1985." ordered-list guard are unchanged.

1 files changed, 25 insertions, 2 deletions

diff --git a/filters/html-converters/md2html b/filters/html-converters/md2html
index 79d4a17..d2906ec 100755
--- a/filters/html-converters/md2html
+++ b/filters/html-converters/md2html
@@ -31,6 +31,13 @@ _MD_ANYITEM = re.compile(r'^ {0,3}(?:[-*+]|\d{1,9}[.)])[ \t]')
 class _ListInterruptPreprocessor(Preprocessor):
 	def run(self, lines):
 		out, in_fence, fence = [], False, None
+		# Track whether we're already inside a list. A bullet that merely
+		# continues an existing list must NOT get a blank line inserted before
+		# it: that would turn a tight list into a loose one (every item wrapped
+		# in <p>), which is exactly what happens when a preceding item wraps onto
+		# an indented continuation line and we mistake the next bullet for a
+		# paragraph->list transition. We only want to interrupt a real paragraph.
+		in_list = False
 		for line in lines:
 			m = _MD_FENCE.match(line)
 			if m:
@@ -41,12 +48,28 @@ class _ListInterruptPreprocessor(Preprocessor):
 					in_fence, fence = False, None
 				out.append(line)
 				continue
-			if not in_fence and out:
+			if in_fence:
+				out.append(line)
+				continue
+			if not line.strip():
+				# a blank line doesn't end a list (loose lists / lazy
+				# continuation), so leave in_list untouched
+				out.append(line)
+				continue
+			is_item = bool(_MD_ANYITEM.match(line))
+			indented = line[:1] in (' ', '\t')
+			# a non-indented, non-item line after a list closes that list
+			if in_list and not is_item and not indented:
+				in_list = False
+			if out:
 				prev = out[-1]
-				if (prev.strip() and not _MD_ANYITEM.match(prev)
+				if (prev.strip() and not in_list
+						and not _MD_ANYITEM.match(prev)
 						and not prev.startswith(('    ', '\t'))
 						and (_MD_BULLET.match(line) or _MD_ORDERED1.match(line))):
 					out.append('')
+			if is_item:
+				in_list = True
 			out.append(line)
 		return out