feat(control-pipeline): GuidanceIngester engine for supervisory guidance (Parser 3)

Add services/guidance_ingester.py — extracts guidance documents (pdfplumber for
PDF, an HTML stripper otherwise; pdfplumber is imported lazily so the module and
its tests load without it) and tags them as a SEPARATE interpretative source:
source_class=supervisory_guidance / authority_weight=70 / bindingness=
interpretative / use_for_primary=false, with references_out to the binding norms
they interpret (Art. N DSGVO / § N BDSG). Guidance therefore ranks below binding
law for obligation questions yet stays retrievable as interpretation context.

supervisory_guidance is reused deliberately: the live re-ranker already weights
it 70 and 8k+ chunks use it (no classifier change, no schema drift). EDPB is the
first target; technical standards (weight 80) are a later separate class.

Tested: 6 unit tests on the text + metadata path (PDF extraction is exercised in
the container), ruff + mypy clean.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

control-pipeline/tests/test_guidance_ingester.py

@@ -0,0 +1,72 @@
+"""Unit tests for the GuidanceIngester engine (Parser 3).
+
+Pure tests on the text + metadata path (PDF extraction is a lazy pdfplumber
+wrapper, exercised in the container). Covers: normalisation, HTML stripping,
+references_out to the interpreted norm, the interpretative (non-primary)
+metadata and the self-test gate.
+"""
+
+from services.guidance_ingester import (
+    GuidanceSpec,
+    build_upload_unit,
+    extract_html,
+    guidance_meta,
+    guidance_refs_out,
+    normalize_text,
+    self_test,
+)
+
+SPEC = GuidanceSpec(
+    source_id="edpb_dpo",
+    short="EDPB DPO",
+    title="EDPB Leitlinien zum Datenschutzbeauftragten",
+    publisher="EDPB",
+    url="https://edpb.europa.eu/guidelines/dpo",
+    interpreted_reg="DSGVO",
+    version_date="2017-04-05",
+)
+
+
+def test_normalize_text_collapses_whitespace_and_blank_runs():
+    assert normalize_text("a   b\t c\n\n\n\nd") == "a b c\n\nd"
+
+
+def test_extract_html_strips_tags():
+    assert "Hallo Welt" in extract_html("<p>Hallo <b>Welt</b></p>")
+
+
+def test_guidance_refs_out_links_to_interpreted_reg():
+    text = "Gemaess Artikel 37, Art. 38 und Article 9 der Verordnung sowie § 38 BDSG."
+    refs = guidance_refs_out("DSGVO", text)
+    assert "Art. 37 DSGVO" in refs
+    assert "Art. 38 DSGVO" in refs
+    assert "Art. 9 DSGVO" in refs
+    assert "§ 38 BDSG" in refs
+
+
+def test_guidance_meta_is_interpretative_not_primary():
+    meta = guidance_meta(SPEC, "Diese Leitlinie erlaeutert Artikel 37 DSGVO im Detail.")
+    assert meta["source_class"] == "supervisory_guidance"
+    assert meta["authority_weight"] == 70
+    assert meta["use_for_primary"] is False
+    assert meta["bindingness"] == "interpretative"
+    assert meta["chunk_scope"] == "guidance"
+    assert meta["regulation_short"] == "EDPB DPO"
+    assert meta["interprets"] == "DSGVO"
+    assert meta["issuer"] == "EDPB"
+    assert "Art. 37 DSGVO" in meta["references_out"]
+
+
+def test_self_test_passes_long_and_flags_short():
+    ok, _ = self_test("x" * 300)
+    assert ok
+    bad, problems = self_test("too short")
+    assert not bad and "too short" in problems[0]
+
+
+def test_build_upload_unit_tags_collection_and_version():
+    unit = build_upload_unit(SPEC, "A" * 300 + " Artikel 35 DSGVO", "run9")
+    assert unit.document_version == "run9-edpb_dpo"
+    assert unit.collection == "bp_compliance_datenschutz"
+    assert unit.filename == "edpb_dpo.txt"
+    assert unit.meta["use_for_primary"] is False