fix(extract): a label with no body under it is emitted, not overwritten

GREEN. A pending label prefix is now flushed as its own line when the next
thing out is a heading, when a second prefix would replace it, and at the end
of the document. Measured on R761: non-whitespace preservation goes from
1 283 395 against 1 283 393 to **1 283 395 against 1 283 395, ratio
1.000000** -- exact, which is what the invariant claims. ATX heading lines
2 761 of 2 761, unchanged.

Known-negatives held: `~/okf-test/dokumenter` still byte-identical against the
frozen `e1f4faa` export (`diff -r` empty), and that folder has 0 `.xml` files
of its 5, so it is a real control and not merely an unchanged number.

pytest -q: 1567 passed, 1 skipped.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

src/llm_ingestion_okf/extract.py

@@ -481,11 +481,25 @@ class _XmlTextExtractor:
             self._lines.append(f"{self._prefix}{line}")
             self._prefix = ""
         if prefix:
+            # A prefix still pending here belongs to a section that turned out
+            # to have no body line at all, and REPLACING it would drop it from
+            # the document. Measured on R761: exactly one `x)`, two characters,
+            # which is the whole distance between 0.999998 and exact.
+            if self._prefix:
+                self._lines.append(self._prefix.rstrip())
             self._prefix = prefix
 
     def _emit(self, line: str) -> None:
-        """Put a whole line out, ahead of whatever is being accumulated."""
+        """Put a whole line out, ahead of whatever is being accumulated.
+
+        A prefix still pending is FLUSHED first rather than carried past a
+        heading: the section it belongs to is above this one, and holding it
+        would either attach it to the wrong body or lose it outright.
+        """
         self._break()
+        if self._prefix:
+            self._lines.append(self._prefix.rstrip())
+            self._prefix = ""
         self._lines.append(line)
 
     def _text_of(self, element: Element) -> str:
@@ -554,6 +568,8 @@ class _XmlTextExtractor:
     def text(self, root: Element) -> str:
         self._walk(root, 0)
         self._break()
+        if self._prefix:
+            self._lines.append(self._prefix.rstrip())
         return "\n".join(self._lines)