Add scrape of ETF top/major holdings

yfinance/base.py

@@ -36,6 +36,7 @@
 from .scrapers.analysis import Analysis
 from .scrapers.fundamentals import Fundamentals
 from .scrapers.holders import Holders
+from .scrapers.holdings import Holdings
 from .scrapers.quote import Quote, FastInfo
 from .scrapers.history import PriceHistory
 
@@ -68,6 +69,7 @@ def __init__(self, ticker, session=None, proxy=None):
         self._price_history = None  # lazy-load
         self._analysis = Analysis(self._data, self.ticker)
         self._holders = Holders(self._data, self.ticker)
+        self._holdings = Holdings(self._data, self.ticker)
         self._quote = Quote(self._data, self.ticker)
         self._fundamentals = Fundamentals(self._data, self.ticker)
 
@@ -194,6 +196,13 @@ def get_mutualfund_holders(self, proxy=None, as_dict=False):
                 return data.to_dict()
             return data
 
+    def get_major_holdings(self, proxy=None, as_dict=False):
+        self._holdings.proxy = proxy
+        data = self._holdings.major
+        if as_dict:
+            return data.to_dict()
+        return data
+
     def get_insider_purchases(self, proxy=None, as_dict=False):
         self._holders.proxy = proxy or self.proxy
         data = self._holders.insider_purchases

yfinance/const.py

@@ -152,4 +152,5 @@
     "sectorTrend",
     "recommendationTrend",
     "futuresChain",
+    "topHoldings"
 )

yfinance/scrapers/holdings.py

@@ -0,0 +1,74 @@
+import pandas as pd
+from bs4 import BeautifulSoup
+
+from yfinance.data import YfData
+
+class Holdings:
+    _SCRAPE_URL_ = 'https://finance.yahoo.com/quote'
+
+    def __init__(self, data: YfData, symbol: str, proxy=None):
+        self._data = data
+        self._symbol = symbol
+        self.proxy = proxy
+        self._major = None
+
+    @property
+    def major(self) -> pd.DataFrame:
+        if self._major is None:
+            self._scrape(self.proxy)
+        return self._major
+
+    def _scrape(self, proxy):
+        ticker_url = "{}/{}".format(self._SCRAPE_URL_, self._symbol)
+        try:
+            resp = self._data.cache_get(ticker_url + '/holdings', proxy)
+            if "/holdings" not in resp.url:
+                raise Exception(f'{self.ticker}: does not have holdings')
+
+            # Manually parse, because while pandas.read_html gets tables, 
+            # it doesn't get table names.
+            soup = BeautifulSoup(resp.text, 'html.parser')
+            h3 = None
+            tables = {}
+            for element in soup.descendants:
+                if element.name == 'h3':
+                    h3 = element.get_text(strip=True)
+                elif element.name == 'table' and h3:
+                    try:
+                        df_list = pd.read_html(str(element))
+                        if df_list:
+                            tables[h3] = df_list[0]
+                    except ValueError:
+                        pass
+
+            # Prettify tables, convert types
+            for k in tables.keys():
+                d = tables[k]
+
+                d = d.set_index(d.columns[0])
+                d.index.name = None
+                d.columns = [k]
+
+                d[k] = d[k].str.replace('--', '')
+                f_pct = d[k].str.contains('%')
+                if (f_pct|(d[k]=='')).all():
+                    d[k] = d[k].str.replace('%', '')
+                try:
+                    d[k] = pd.to_numeric(d[k])
+                except ValueError:
+                    pass
+
+                tables[k] = d
+
+            # print("------------------------")
+            # for k in tables.keys():
+            #     print(k)
+            #     t = tables[k]
+            #     print(t)
+            #     print(t[t.columns[0]].dtype)
+            #     print("------------------------")
+
+            self._major_holdings = tables
+
+        except Exception:
+            self._major_holdings = []

yfinance/ticker.py

@@ -280,3 +280,7 @@ def earnings_forecasts(self) -> _pd.DataFrame:
     @property
     def history_metadata(self) -> dict:
         return self.get_history_metadata()
+
+    @property
+    def major_holdings(self):
+        return self.get_major_holdings()