parser: support diverse md inputs (bold headings, Works Cited, blockquote notes, junk filtering)

- Bibliography heading now accepts bold markers (## **BIBLIOGRAPHY**),
  alternative names (Works Cited, References), and numbered sub-headings
  within the section.
- Original heading text preserved in output instead of hardcoded
  "## Bibliography".
- New find_blockquote_notes() parser for PDF-to-markdown footnote format
  (> N text), wired into CLI as third fallback after pandoc and numbered.
- PDF junk filtered from bibliography entries: bare page numbers,
  blockquote footnotes, download banners, CC license URLs, and short
  running headers.

tests/test_cli.py

@@ -51,14 +51,14 @@ def test_reformat_draft_preserves_sections_after_bibliography():
     draft = """\
 ## Bibliography
 
-one entry.
+Yu, Charles. *Interior Chinatown* (New York: Pantheon, 2020).
 
 ## Appendix
 
 appendix text.
 """
     output = reformat_draft(draft, formatter=_fake_formatter)
-    assert "FORMATTED(one entry.)" in output
+    assert "FORMATTED(Yu, Charles. *Interior Chinatown* (New York: Pantheon, 2020).)" in output
     assert "## Appendix" in output
     assert "appendix text." in output
 
@@ -69,11 +69,27 @@ def test_reformat_draft_raises_when_no_bibliography():
 
 
 def test_reformat_draft_bibliography_heading_preserved():
-    draft = "## Bibliography\n\nentry.\n"
+    draft = '## Bibliography\n\nYu, Charles. *Interior Chinatown* (Pantheon, 2020).\n'
     output = reformat_draft(draft, formatter=_fake_formatter)
     assert "## Bibliography" in output
 
 
+def test_reformat_draft_preserves_works_cited_heading():
+    """When the input heading is ``## Works Cited``, the output must
+    keep that heading — not replace it with ``## Bibliography``."""
+    draft = '## Works Cited\n\nDavidson, Donald. "On the Very Idea." (1974).\n'
+    output = reformat_draft(draft, formatter=_fake_formatter)
+    assert "## Works Cited" in output
+    assert "## Bibliography" not in output
+
+
+def test_reformat_draft_preserves_bold_wrapped_heading():
+    """Bold markers in the heading should be preserved in the output."""
+    draft = '## **Works Cited**\n\nDavidson, Donald. "On the Very Idea." (1974).\n'
+    output = reformat_draft(draft, formatter=_fake_formatter)
+    assert "## **Works Cited**" in output
+
+
 def test_python_dash_m_invocation_actually_runs_main():
     """Regression test: `python -m cmos.cli` must actually invoke main().
 
@@ -216,6 +232,23 @@ def test_reformat_notes_auto_detects_numbered_format():
     assert "[^2]" not in output
 
 
+def test_reformat_notes_auto_detects_blockquote_format():
+    """PDF-to-markdown drafts use ``> N text`` blockquote footnotes.
+    reformat_notes must auto-detect after pandoc and numbered fail."""
+    draft = """\
+Some prose.
+
+> 1 Davidson, "On the Very Idea of a Conceptual Scheme."
+> 2 Frankenberry 1999, 526.
+"""
+    output = reformat_notes(draft, formatter=_fake_note_formatter)
+    assert '> 1 NOTE_FORMATTED(Davidson, "On the Very Idea of a Conceptual Scheme.")' in output
+    assert "> 2 NOTE_FORMATTED(Frankenberry 1999, 526.)" in output
+    # No pandoc or numbered markers should appear.
+    assert "[^1]" not in output
+    assert "[1] " not in output
+
+
 def test_reformat_notes_pandoc_input_still_round_trips_as_pandoc():
     """Regression: after the auto-detect change, pandoc-format input
     must still produce pandoc-format output. The original_prefix path
@@ -265,28 +298,28 @@ def test_reformat_draft_preserves_order_under_concurrency():
     # later ones if order were naively tied to completion.
     import time
 
+    # Map each entry to a sleep duration so earlier entries finish last.
+    order = {"2020": 4, "2021": 3, "2022": 2, "2023": 1, "2024": 0}
+
     def slow_fake(messy: str) -> str:
-        # Entries with lower index sleep longer so they finish last.
-        n = int(messy.split()[-1])
-        time.sleep(0.05 * (5 - n))
+        year = messy.strip()[-5:-1]  # extract "2020" etc.
+        time.sleep(0.05 * order.get(year, 0))
         return f"FORMATTED({messy})"
 
     draft = """\
 ## Bibliography
 
-entry 0
-entry 1
-entry 2
-entry 3
-entry 4
+Author, A. *Title Zero* (Publisher, 2020).
+Author, B. *Title One* (Publisher, 2021).
+Author, C. *Title Two* (Publisher, 2022).
+Author, D. *Title Three* (Publisher, 2023).
+Author, E. *Title Four* (Publisher, 2024).
 """
     output = reformat_draft(draft, formatter=slow_fake, concurrency=4)
     # Entries must appear in input order.
     lines = [l for l in output.splitlines() if l.startswith("FORMATTED(")]
-    assert lines == [
-        "FORMATTED(entry 0)",
-        "FORMATTED(entry 1)",
-        "FORMATTED(entry 2)",
-        "FORMATTED(entry 3)",
-        "FORMATTED(entry 4)",
-    ]
+    assert lines[0].startswith("FORMATTED(Author, A.")
+    assert lines[1].startswith("FORMATTED(Author, B.")
+    assert lines[2].startswith("FORMATTED(Author, C.")
+    assert lines[3].startswith("FORMATTED(Author, D.")
+    assert lines[4].startswith("FORMATTED(Author, E.")