Investment Objective
" b"The Fund seeks to track the performance of a benchmark index " b"that measures the investment return of large-capitalization " b"stocks.
Fees and Expenses
") def test_text_and_objective() -> None: print("text / objective extraction", flush=True) txt = edgar.to_text(_HTML) check("to_text strips tags", "Fund Summary" in txt and "" not in txt and "Investment Objective" in txt) doc = ("Random cover page text. " "Vanguard 500 Index Fund\nProspectus\nInvestment Objective\n" "The Fund seeks to track the performance of a benchmark index " "that measures the investment return of large-capitalization " "stocks.\nFees and Expenses\nmore text about Vanguard 500 " "Index Fund share classes.") obj = edgar.extract_objective(doc, "Vanguard 500 Index Fund") check("objective extracted", obj is not None and obj.startswith("The Fund seeks to track") and "large-capitalization stocks" in obj, repr(obj)) check("objective is bounded", obj is not None and len(obj) < 300, repr(obj and len(obj))) doc2 = ("Fund ABC\nThe ETF seeks to invest at least 80% of assets in " "US investment grade debt securities of emerging markets.") obj2 = edgar.extract_objective(doc2, "Fund ABC") check("variant: 'The ETF seeks to ...' on the cover", obj2 is not None and "debt securities" in obj2 and "US" in obj2, repr(obj2)) check("no objective -> None", edgar.extract_objective("no fund here at all", "Ghost Fund") is None) # a document that merely MENTIONS the fund deep in its body (family # filings, funds investing in it) must not yield that fund's objective: # the name is not on the cover (within the first COVER_GATE chars) body = ("Section %d discusses portfolio construction, risk management " "and liquidity. ") filler = ("XYZ Growth Fund\nProspectus\nInvestment Objective\nThe Fund " "seeks to provide total return while generating moderate " "current income.\n" + "".join(body % i for i in range(40)) + "Comparison: similar to Vanguard 500 Index Fund.\n") check("mention-only doc -> None", edgar.extract_objective(filler, "Vanguard 500 Index Fund") is None) # the same doc with the name on the cover IS about the fund oncover = ("Vanguard 500 Index Fund\nProspectus\nInvestment Objective\n" "The Fund seeks to track large-capitalization stocks.\n" + "".join(body % i for i in range(40))) check("name on cover -> extracted", (edgar.extract_objective(oncover, "Vanguard 500 Index Fund") or "").startswith("The Fund seeks to track")) # a leveraged ETF's cover cites the underlying fund's name; the name # is a reference, not a title -> not our fund's document doc4 = ("Acme 2X Leveraged QQQ Daily ETF\nProspectus\nThe Fund seeks\n" "daily investment results that are 200% of the performance of\n" "the Invesco QQQ Trust. The Fund seeks to achieve daily\n" "investment results equal to two times the performance of the\n" "index, before fees and expenses.") check("underlying reference on cover -> None", edgar.extract_objective(doc4, "Invesco QQQ Trust") is None) # ---------------------------------------------------------------- classify def test_classify() -> None: print("category classification", flush=True) c = edgar.classify_category check("US index fund -> equity", c("The Fund seeks to track large-capitalization stocks.", "Vanguard 500 Index Fund") == "equity") check("aggregate bond -> fixed_income", c("The Fund seeks to track the performance of an index of US " "investment grade debt securities.", "iShares Core US Aggregate Bond ETF") == "fixed_income") check("balanced -> mixed", c("The Fund invests in a combination of equity and debt " "securities.", "Example Balanced Fund") == "mixed") check("gold -> alternatives", c("The Fund seeks to track the price of gold bullion.", "SPDR Gold Shares") == "alternatives") check("money market -> money_market", c("The Fund seeks to maintain stability of principal. It " "invests in money market instruments.", "XX Money Market Fund") == "money_market") check("unknown -> other", c("The Fund does unusual things.", "Mystery") == "other") # ---------------------------------------------------------------- edgar live def test_edgar_live() -> None: print("EDGAR live (network)", flush=True) try: name = None meta = json.loads((pathlib.Path("~/prog/fin/stocks").expanduser() / "vtsax.json").read_text())["chart"]["result"][0]["meta"] name = meta.get("longName") if not name: raise urllib.error.URLError("no local vtsax.json") res = edgar.fetch_fund(name, ticker="vtsax", max_docs=8) except Exception as e: print(f" SKIP (no network / data): {type(e).__name__}: {e}") return check("vtsax objective fetched", res is not None and "seeks to" in res["objective"] and "stock market" in res["objective"].lower(), repr(res and res.get("objective"))) check("vtsax classified equity", res is not None and res["category"] == "equity", repr(res and res.get("category"))) check("provenance recorded", res is not None and res["url"].startswith("https://www.sec.gov") and res["form"], repr(res)) def test_curated() -> None: print("curated", flush=True) import fundlab.fundinfo as fi cur = fi.load_curated() check("curated file loads", len(cur) >= 10, str(len(cur))) check("curated entries well-formed", all(set(v) >= {"objective", "category"} and v["category"] in ("equity", "fixed_income", "mixed", "alternatives", "money_market", "other") for v in cur.values())) check("curated objectives look right", all("seeks" in v["objective"].lower() for v in cur.values())) def main() -> int: test_pool() test_text_and_objective() test_classify() test_curated() test_edgar_live() print(f"\n{PASS} passed, {FAIL} failed") return 1 if FAIL else 0 if __name__ == "__main__": sys.exit(main())