fix: keep every entry heading on listing pages (#774)

A blog index, archive or category page holds several <article> siblings.
BODY_XPATH's "(.//article)[1]" keeps the first entry only, so the page
looks under-extracted and escalates to justext, which emits paragraphs
and drops every heading.

Give the recall pass a listing expression that takes the container of
the sibling articles, so the retry returns all entries with their heads
and wins the escalation. Recall only: on an article page the same shape
is a related-posts strip, whose container would hide the body.

trafilatura/main_extractor.py

@@ -35,6 +35,7 @@ from .xpaths import (
     COMMENTS_DISCARD_XPATH,
     COMMENTS_XPATH,
     DISCARD_IMAGE_ELEMENTS,
+    LISTING_BODY_XPATH,
     OVERALL_DISCARD_XPATH,
     PRECISION_DISCARD_XPATH,
     TEASER_DISCARD_XPATH,
@@ -654,8 +655,10 @@ def _extract(tree: HtmlElement, options: Extractor) -> tuple[_Element, str, set[
     if options.links is True:
         potential_tags.add("ref")
     result_body = Element("body")
+    # kept out of the confident passes: see LISTING_BODY_XPATH
+    expressions = BODY_XPATH if options.focus != "recall" else LISTING_BODY_XPATH + BODY_XPATH
     # iterate
-    for expr in BODY_XPATH:
+    for expr in expressions:
         # select tree if the expression has been found
         subtree = next((s for s in expr(tree) if s is not None), None)
         if subtree is None:

trafilatura/xpaths.py

@@ -105,6 +105,13 @@ BODY_XPATH = [
         """
     ),
 ]
+
+# Listing layouts (blog index, archive, category) hold several <article> siblings in one container,
+# each a teaser with its own heading. BODY_XPATH's "(.//article)[1]" keeps the first entry only, so
+# the page escalates to justext, which flattens every heading into a paragraph (#774). Recall only:
+# on an article page the same shape is a related-posts strip, whose container would hide the body.
+LISTING_BODY_XPATH = [XPath("(.//article[preceding-sibling::article])[1]/..")]
+
 # starts-with(@id, "article") or
 # or starts-with(@id, "story") or contains(@class, "story")
 # starts-with(@class, "content ") or contains(@class, " content")