Fix Google SERP parser after results page markup changes

google_stocks/__init__.py

@@ -9,6 +9,7 @@ from datetime import datetime
 import lxml.html
 from playwright.sync_api import Page, Playwright, expect, sync_playwright
 
+
 auth_file = os.path.expanduser("~/lib/auth/google.json")
 data_loc = os.path.expanduser("~/lib/google_stocks")
 
@@ -95,9 +96,10 @@ class Index:
         accept_cookies(page)
         stay_signed_out(page)
 
-        expect(page.get_by_text("Market Summary")).to_be_visible()
+        expect(page.locator("#center_col")).to_be_visible()
 
         html = page.content()
+        context.storage_state(path=auth_file)
         filename = data_filename("serp")
         with open(filename, "w") as out:
             out.write(html)
@@ -113,36 +115,88 @@ class Index:
     def parse_html(self, html: str) -> None:
         """Parse HTML."""
         root = lxml.html.fromstring(html)
+        self._set_mapped_attributes(root)
 
-        re_percent_change = re.compile(r" *\(([0-9.]+)%\) *")
+        tag = self._find_price_tag(root)
+        tag_text = " ".join(tag.text_content().split())
+        self._set_price_values(tag, tag_text)
+        self._set_state_and_timestamp(tag_text)
 
+    def _set_mapped_attributes(self, root: lxml.html.HtmlElement) -> None:
         for attrid_tag in root.findall(".//*[@data-attrid]"):
             attrid = attrid_tag.get("data-attrid")
             if attrid not in attr_map:
                 continue
             setattr(self, attr_map[attrid], attrid_tag.text_content())
 
+    def _find_price_tag(self, root: lxml.html.HtmlElement) -> lxml.html.HtmlElement:
         tag = root.find('.//*[@data-attrid="Price"]')
         assert tag is not None
+        return tag
 
-        assert tag[0] is not None and tag[1] is not None
-        self.price = decimal.Decimal(tag[0].text_content().replace(",", "").strip())
-        percent_change_str = tag[1][0].text_content().strip()
-        percent_change_str = percent_change_str.replace("\N{MINUS SIGN}", "-")
+    @staticmethod
+    def _clean_num(s: str) -> str:
+        return s.replace(",", "").replace("\N{MINUS SIGN}", "-").strip()
 
-        assert tag[2][0][0].text and tag[2][0][1].text
+    def _extract_price_text(self, tag: lxml.html.HtmlElement, tag_text: str) -> str:
+        re_number = re.compile(r"[0-9][0-9,]*\.[0-9]+")
+        re_quote_triplet = re.compile(
+            r"([0-9][0-9,]*\.[0-9]+)\s*([+\-\N{MINUS SIGN}][0-9][0-9,]*\.[0-9]+)\s*\(([+\-\N{MINUS SIGN}]?[0-9.]+)%\)"
+        )
 
-        self.state = tag[2][0][0].text.rstrip(": ").lower()
-        self.timestamp = tag[2][0][1].text
+        for node_text in tag.xpath('.//*[@aria-live="polite"]/text()'):
+            m = re_number.search(node_text)
+            if m:
+                return m.group(0)
 
-        self.day_change = decimal.Decimal(percent_change_str)
-        m = re_percent_change.match(tag[1][1].text_content())
+        m = re_quote_triplet.search(tag_text)
         assert m
-        percent_change = decimal.Decimal(m.group(1))
-        if percent_change_str[0] == "-":
+        return m.group(1)
+
+    def _set_price_values(self, tag: lxml.html.HtmlElement, tag_text: str) -> None:
+        re_day_change = re.compile(r"([+\-\N{MINUS SIGN}][0-9][0-9,]*\.[0-9]+)")
+        re_percent = re.compile(r"\(([+\-\N{MINUS SIGN}]?[0-9.]+)%\)")
+
+        price_text = self._extract_price_text(tag, tag_text)
+        self.price = decimal.Decimal(self._clean_num(price_text))
+
+        m = re_day_change.search(tag_text)
+        assert m
+        self.day_change = decimal.Decimal(self._clean_num(m.group(1)))
+
+        m = re_percent.search(tag_text)
+        assert m
+        percent_change_text = self._clean_num(m.group(1))
+        percent_change = decimal.Decimal(percent_change_text)
+        if not percent_change_text.startswith(("+", "-")) and self.day_change < 0:
             percent_change = -percent_change
         self.percent_change = percent_change
 
+    def _set_state_and_timestamp(self, tag_text: str) -> None:
+        re_timestamp = re.compile(r"([0-9]{1,2}\s+[A-Za-z]{3},\s+[0-9]{1,2}:[0-9]{2}\s+[A-Z]{2,5})")
+        re_state_time = re.compile(
+            r"\b(Closed|Open|Market open|Market closed|Pre-market|After hours):\s*([0-9]{1,2}\s+[A-Za-z]{3},\s+[0-9]{1,2}:[0-9]{2}\s+[A-Z]{2,5})",
+            re.IGNORECASE,
+        )
+
+        m = re_state_time.search(tag_text)
+        if m:
+            self.state = m.group(1).strip().lower()
+            self.timestamp = m.group(2).strip()
+            return
+
+        tm = re_timestamp.search(tag_text)
+        if not tm:
+            return
+
+        self.timestamp = tm.group(1).strip()
+        prefix = tag_text[: tm.start()]
+        states = ("Market closed", "Market open", "Pre-market", "After hours", "Closed", "Open")
+        for state in states:
+            if prefix.endswith(state) or state in prefix:
+                self.state = state.lower()
+                break
+
     @property
     def one_line(self) -> str:
         """Index name, price and price change."""