Spaces:
Sleeping
Sleeping
File size: 8,301 Bytes
e813ce3 4438658 e813ce3 4438658 e813ce3 4438658 e813ce3 4438658 e813ce3 4438658 e813ce3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 | """SEC EDGAR integration β fetches 10-K/10-Q filings and extracts
MD&A (Item 7) and Risk Factors (Item 1A) for the analyst prompt.
Uses the EDGAR EFTS full-text search API (same endpoint used in
``src/discovery/insider_feed.py``) and BeautifulSoup for HTML parsing.
"""
import re
import time
from datetime import datetime, timedelta
import requests
from bs4 import BeautifulSoup
from langchain_core.tools import tool
from langchain_text_splitters import RecursiveCharacterTextSplitter
from src.core.logger import get_logger
logger = get_logger(__name__)
_SEC_HEADERS = {
"User-Agent": "PrimoGreedy/1.0 (contact@primogreedy.com)",
"Accept": "application/json",
}
_EFTS_URL = "https://efts.sec.gov/LATEST/search-index"
_MAX_SECTION_CHARS = 2000
_SPLITTER = RecursiveCharacterTextSplitter(
chunk_size=_MAX_SECTION_CHARS, chunk_overlap=200
)
# Regex patterns for section headers in 10-K/10-Q filings
_MDA_PATTERN = re.compile(
r"Item\s*7[\.\s\β\-]+.*?Management.s\s+Discussion|"
r"Item\s*7[\.\s\β\-]|"
r"Management.s\s+Discussion\s+and\s+Analysis",
re.IGNORECASE,
)
_RISK_PATTERN = re.compile(
r"Item\s*1A[\.\s\β\-]+.*?Risk\s+Factors|"
r"Item\s*1A[\.\s\β\-]|"
r"Risk\s+Factors",
re.IGNORECASE,
)
_NEXT_ITEM_PATTERN = re.compile(r"Item\s*\d+[A-Z]?[\.\s\β\-]", re.IGNORECASE)
# ---------------------------------------------------------------------------
# EFTS search β find the most recent 10-K or 10-Q for a ticker
# ---------------------------------------------------------------------------
def _search_filings(ticker: str) -> dict | None:
"""Query EDGAR EFTS for the most recent annual/quarterly filing.
Returns the first hit as a dict with ``file_url``, ``form_type``,
``file_date``, ``company_name``, or *None* if nothing found.
"""
two_years_ago = (datetime.now() - timedelta(days=730)).strftime("%Y-%m-%d")
today = datetime.now().strftime("%Y-%m-%d")
params = {
"q": "",
"forms": "10-K,10-Q",
"dateRange": "custom",
"startdt": two_years_ago,
"enddt": today,
"tickers": ticker,
}
try:
resp = requests.get(_EFTS_URL, params=params, headers=_SEC_HEADERS, timeout=10)
if resp.status_code != 200:
logger.info("EDGAR EFTS returned %d for %s", resp.status_code, ticker)
return None
hits = resp.json().get("hits", {}).get("hits", [])
if not hits:
return None
src = hits[0].get("_source", {})
file_num = src.get("file_num", "")
accession = (
src.get("accession_no", "")
.replace("-", "")
)
primary_doc = src.get("file_description", "")
if not primary_doc:
primary_doc = src.get("display_names", [""])[0] if src.get("display_names") else ""
entity_id = src.get("entity_id", "")
return {
"form_type": src.get("form_type", "10-K"),
"file_date": src.get("file_date", ""),
"company_name": src.get("display_names", [""])[0] if src.get("display_names") else ticker,
"entity_id": entity_id,
"accession": accession,
"file_num": file_num,
}
except requests.RequestException as exc:
logger.warning("EDGAR EFTS request failed for %s: %s", ticker, exc)
return None
def _fetch_filing_index(entity_id: str, accession: str) -> str | None:
"""Fetch the filing index page and return the URL of the primary HTML document."""
if not entity_id or not accession:
return None
index_url = (
f"https://www.sec.gov/Archives/edgar/data/{entity_id}/{accession}/"
)
time.sleep(0.5)
try:
resp = requests.get(
index_url,
headers={**_SEC_HEADERS, "Accept": "text/html"},
timeout=10,
)
if resp.status_code != 200:
return None
soup = BeautifulSoup(resp.text, "html.parser")
for a_tag in soup.find_all("a", href=True):
href = a_tag["href"]
if href.endswith(".htm") or href.endswith(".html"):
if not href.startswith("http"):
href = f"https://www.sec.gov{href}" if href.startswith("/") else f"{index_url}{href}"
return href
return None
except requests.RequestException as exc:
logger.warning("EDGAR filing index fetch failed: %s", exc)
return None
# ---------------------------------------------------------------------------
# HTML parser β extract MD&A and Risk Factors sections
# ---------------------------------------------------------------------------
def _extract_section(full_text: str, start_pattern: re.Pattern, label: str) -> str:
"""Find a section by *start_pattern* and return text up to the next Item header."""
match = start_pattern.search(full_text)
if not match:
return ""
start = match.end()
remainder = full_text[start:]
end_match = _NEXT_ITEM_PATTERN.search(remainder, pos=200)
if end_match:
section_text = remainder[: end_match.start()]
else:
section_text = remainder[:_MAX_SECTION_CHARS * 2]
section_text = section_text.strip()
if not section_text:
return ""
if len(section_text) > _MAX_SECTION_CHARS:
chunks = _SPLITTER.split_text(section_text)
section_text = chunks[0] if chunks else section_text[:_MAX_SECTION_CHARS]
return section_text
def parse_mda_risk_factors(html: str) -> str:
"""Extract MD&A and Risk Factors from a 10-K/10-Q HTML document.
Returns a formatted string ready for ``{sec_context}`` in the prompt,
or an empty string if extraction fails.
"""
try:
soup = BeautifulSoup(html, "html.parser")
for tag in soup(["script", "style", "meta", "link"]):
tag.decompose()
full_text = soup.get_text(separator="\n", strip=True)
mda = _extract_section(full_text, _MDA_PATTERN, "MD&A")
risk = _extract_section(full_text, _RISK_PATTERN, "Risk Factors")
if not mda and not risk:
return ""
parts = ["SEC FILING GROUND TRUTH:"]
if mda:
parts.append(f"\nMD&A SUMMARY (Item 7):\n{mda}")
if risk:
parts.append(f"\nRISK FACTORS (Item 1A):\n{risk}")
return "\n".join(parts)
except Exception as exc:
logger.warning("SEC filing parse error: %s", exc)
return ""
# ---------------------------------------------------------------------------
# Public @tool β used by analyst_node
# ---------------------------------------------------------------------------
@tool
def get_sec_filings(ticker: str) -> str:
"""Fetch the most recent 10-K or 10-Q filing from SEC EDGAR and extract
MD&A and Risk Factors sections for investment analysis.
Args:
ticker: US stock ticker symbol (e.g. AAPL, MSFT)
"""
if "." in ticker:
return ""
filing = _search_filings(ticker)
if not filing:
logger.info("No SEC filings found for %s", ticker)
return ""
logger.info(
"Found %s for %s (filed %s)",
filing["form_type"], ticker, filing["file_date"],
)
doc_url = _fetch_filing_index(filing["entity_id"], filing["accession"])
if not doc_url:
return (
f"SEC FILING GROUND TRUTH:\n"
f"Found {filing['form_type']} filed {filing['file_date']} "
f"but could not retrieve document."
)
time.sleep(0.5)
try:
resp = requests.get(
doc_url,
headers={**_SEC_HEADERS, "Accept": "text/html"},
timeout=15,
)
if resp.status_code != 200:
return ""
if len(resp.text) > 2_000_000:
logger.info("SEC filing too large (%d bytes), truncating", len(resp.text))
html = resp.text[:2_000_000]
else:
html = resp.text
except requests.RequestException as exc:
logger.warning("SEC filing fetch failed for %s: %s", ticker, exc)
return ""
result = parse_mda_risk_factors(html)
if result:
header = f"[Source: {filing['form_type']} filed {filing['file_date']}]"
return f"{result}\n{header}"
return ""
|