Previously the loop would break after the first heading was found,
so only h1 was captured. Now all three heading levels are extracted
in order from the first 7 lines.

Rules:
- If there is only h1, only use h1
- If there is also a h2 AFTER h1, also add h2
- If there is also a h3 AFTER h2, also add h3
- If there is no h1 but h2, add h2
- If there is no h1 and no h2 but h3, add h3
- If there is no h1, h2 or h3, filename only contains the time component

@@ -26,2 +34,4 @@
+            if level == 0:
+                continue
             heading_text = stripped[level:].strip()
             if level == 1: