feat(prospection): public holdings extraction gated behind human confirmation

.gitignore

@@ -6,6 +6,7 @@ __pycache__/
 .venv/
 *.mp4
 *.pdf
+!tests/fixtures/*.pdf
 .langgraph_api/
 .worktrees/
 .superpowers/sdd/

prospection/sources/geco_funds.py

@@ -0,0 +1,14 @@
+"""prospection/sources/geco_funds.py — exposition actions déclarée.
+
+Repli documenté : l'export GECO par SGP n'étant pas confirmé comme
+automatisable, la poche actions se lit dans `libelle_type_activite` du CSV
+AMF, déjà en base.
+"""
+from __future__ import annotations
+
+_EQUITY_LABELS = ("action", "diversifi", "internationa", "monde", "amérique", "amerique")
+
+
+def has_equity_funds(account: dict) -> bool:
+    label = (account.get("type_activite") or "").lower()
+    return any(marker in label for marker in _EQUITY_LABELS)

prospection/sources/holdings.py

@@ -0,0 +1,88 @@
+"""prospection/sources/holdings.py — positions publiées et niveau de preuve.
+
+Rien ne passe en `verified` sans confirmation humaine explicite : le
+validateur de rédaction ne voit jamais une ligne non confirmée.
+"""
+from __future__ import annotations
+
+import re
+from datetime import datetime
+from pathlib import Path
+
+from prospection.sources.ticker_aliases import ALIASES
+from prospection.store import prospects_db as db
+
+REPORT_PATHS = ("/reporting", "/nos-fonds", "/documentation", "/fonds", "/publications")
+
+_LINE_RE = re.compile(r"^(?P<label>[A-Za-zÀ-ÿ0-9 .&'\-]{3,60}?)\s+(?P<weight>\d{1,2}[.,]\d)\s*%$")
+
+
+def discover_report_urls(account: dict, client=None) -> list[str]:
+    site = (account.get("site") or "").rstrip("/")
+    if not site:
+        return []
+    if client is None:
+        import httpx
+
+        client = httpx.Client(timeout=20.0)
+
+    found = []
+    for path in REPORT_PATHS:
+        url = f"{site}{path}"
+        try:
+            response = client.get(url, headers={"User-Agent": "amplegest-prospection/1.0"},
+                                  follow_redirects=True)
+            response.raise_for_status()
+        except Exception:
+            continue
+        found.append(url)
+    return found
+
+
+def extract_holdings(pdf_path: Path, source_url: str) -> list[dict]:
+    import pdfplumber
+
+    rows: list[dict] = []
+    with pdfplumber.open(str(pdf_path)) as pdf:
+        lines = [l.strip() for page in pdf.pages
+                 for l in (page.extract_text() or "").splitlines()]
+
+    for line in lines:
+        match = _LINE_RE.match(line)
+        if not match:
+            continue
+        label = match.group("label").strip()
+        rows.append({
+            "fund_name": None,
+            "raw_label": label,
+            "ticker": resolve_ticker(label),
+            "weight_pct": float(match.group("weight").replace(",", ".")),
+            "as_of": None,
+            "source_url": source_url,
+            "confidence": "probable",
+            "retrieved_at": datetime.utcnow().isoformat(timespec="seconds"),
+            "confirmed_at": None,
+        })
+    return rows
+
+
+def resolve_ticker(label: str) -> str | None:
+    lowered = label.lower()
+    for alias, ticker in ALIASES.items():
+        if alias in lowered:
+            return ticker
+    return None
+
+
+def evidence_level(account_id: int) -> str:
+    confirmed = [h for h in db.confirmed_holdings(account_id) if h.get("ticker")]
+    if confirmed:
+        return "holdings_verified"
+    account = next((a for a in db.list_accounts() if a["id"] == account_id), None)
+    with db.connect() as conn:
+        any_holding = conn.execute(
+            "SELECT 1 FROM holdings WHERE account_id=? LIMIT 1", (account_id,)
+        ).fetchone()
+    if any_holding or (account and account.get("tier") in (1, 2)):
+        return "universe_inferred"
+    return "no_evidence"

prospection/sources/ticker_aliases.py

@@ -0,0 +1,24 @@
+"""prospection/sources/ticker_aliases.py — table d'alias curée, libellé -> ticker EDGAR.
+
+Volontairement petite et manuelle. Une correspondance devinée produirait une
+personnalisation fausse dans un email : mieux vaut ne rien dire.
+"""
+from __future__ import annotations
+
+ALIASES: dict[str, str] = {
+    "apple": "AAPL",
+    "advanced micro devices": "AMD",
+    "alphabet": "GOOGL",
+    "google": "GOOGL",
+    "microsoft": "MSFT",
+    "nvidia": "NVDA",
+    "tesla": "TSLA",
+    "amazon": "AMZN",
+    "meta platforms": "META",
+    "broadcom": "AVGO",
+    "eli lilly": "LLY",
+    "visa": "V",
+    "netflix": "NFLX",
+    "salesforce": "CRM",
+    "adobe": "ADBE",
+}

tests/fixtures/reporting_mensuel.pdf

@@ -0,0 +1,76 @@
+%PDF-1.3
+%���
+1 0 obj
+<<
+/Count 1
+/Kids [3 0 R]
+/MediaBox [0 0 595.28 841.89]
+/Type /Pages
+>>
+endobj
+2 0 obj
+<<
+/OpenAction [3 0 R /FitH null]
+/PageLayout /OneColumn
+/Pages 1 0 R
+/Type /Catalog
+>>
+endobj
+3 0 obj
+<<
+/Contents 4 0 R
+/Parent 1 0 R
+/Resources 6 0 R
+/Type /Page
+>>
+endobj
+4 0 obj
+<<
+/Filter /FlateDecode
+/Length 217
+>>
+stream
+x�m�Ak�@��{~�w)�`�Y]us�$,�ĥ�װa�V#���
+%=
+q��7v)S�
+endstream
+endobj
+5 0 obj
+<<
+/BaseFont /Helvetica
+/Encoding /WinAnsiEncoding
+/Subtype /Type1
+/Type /Font
+>>
+endobj
+6 0 obj
+<<
+/Font <</F1 5 0 R>>
+/ProcSet [/PDF /Text /ImageB /ImageC /ImageI]
+>>
+endobj
+7 0 obj
+<<
+/CreationDate (D:20260829102135Z)
+>>
+endobj
+xref
+0 8
+0000000000 65535 f 
+0000000015 00000 n 
+0000000102 00000 n 
+0000000205 00000 n 
+0000000285 00000 n 
+0000000574 00000 n 
+0000000671 00000 n 
+0000000758 00000 n 
+trailer
+<<
+/Size 8
+/Root 2 0 R
+/Info 7 0 R
+/ID [<C91A12C767E0F45E98F95EC734815254><C91A12C767E0F45E98F95EC734815254>]
+>>
+startxref
+813
+%%EOF

tests/test_prospection_holdings.py

@@ -0,0 +1,73 @@
+"""tests/test_prospection_holdings.py — positions publiques et niveau de preuve."""
+from __future__ import annotations
+
+from pathlib import Path
+
+import pytest
+
+from prospection.sources import geco_funds, holdings
+from prospection.store import prospects_db as db
+
+FIXTURE = Path(__file__).parent / "fixtures" / "reporting_mensuel.pdf"
+
+
+@pytest.fixture(autouse=True)
+def tmp_db(tmp_path, monkeypatch):
+    monkeypatch.setattr(db, "DB_PATH", tmp_path / "prospects.db")
+    monkeypatch.setattr(holdings, "DB_PATH", tmp_path / "prospects.db", raising=False)
+    db.init_db()
+
+
+def _account_id() -> int:
+    return db.upsert_account({"no_amf": "GP-19990001", "nom": "Boutique Gestion SAS",
+                              "site": "https://boutique-gestion.fr", "statut": "vivant",
+                              "independante": 1})
+
+
+def test_has_equity_funds_reads_the_amf_activity_label():
+    assert geco_funds.has_equity_funds({"type_activite": "Actions"}) is True
+    assert geco_funds.has_equity_funds({"type_activite": "Monétaire"}) is False
+
+
+def test_extract_holdings_reads_lines_and_weights():
+    rows = holdings.extract_holdings(FIXTURE, "https://x.fr/r.pdf")
+    labels = {r["raw_label"]: r["weight_pct"] for r in rows}
+    assert labels["Microsoft Corp"] == 5.4
+    assert labels["NVIDIA Corp"] == 4.8
+
+
+def test_extract_holdings_never_marks_a_row_verified():
+    rows = holdings.extract_holdings(FIXTURE, "https://x.fr/r.pdf")
+    assert rows
+    assert all(r["confidence"] == "probable" for r in rows)
+    assert all(r["confirmed_at"] is None for r in rows)
+
+
+def test_resolve_ticker_maps_known_labels_and_gives_up_otherwise():
+    assert holdings.resolve_ticker("Microsoft Corp") == "MSFT"
+    assert holdings.resolve_ticker("Air Liquide SA") is None
+
+
+def test_evidence_level_requires_a_confirmed_edgar_holding():
+    account_id = _account_id()
+    holding_id = db.upsert_holding({"account_id": account_id, "fund_name": "Fonds A",
+                                    "raw_label": "Microsoft Corp", "ticker": "MSFT",
+                                    "weight_pct": 5.4, "as_of": "2026-07-31",
+                                    "source_url": "https://x.fr/r.pdf",
+                                    "confidence": "probable",
+                                    "retrieved_at": "2026-08-01T00:00:00"})
+    assert holdings.evidence_level(account_id) == "universe_inferred"
+    db.confirm_holding(holding_id, "2026-08-29T09:00:00")
+    assert holdings.evidence_level(account_id) == "holdings_verified"
+
+
+def test_evidence_level_is_no_evidence_without_any_holding():
+    assert holdings.evidence_level(_account_id()) == "no_evidence"
+
+
+def test_discover_report_urls_returns_empty_when_every_path_is_missing():
+    class _Client:
+        def get(self, url, headers=None, follow_redirects=True):
+            raise RuntimeError("404")
+
+    assert holdings.discover_report_urls({"site": "https://x.fr"}, client=_Client()) == []