feat: MassHealth member detail demographics, wait for name/plan before PDF

- Extract gender, address (street/city/state/zip) and home phone from the MassHealth Member Detail page and save to the patient; phone is only filled when the patient has none - Shared extract_eligibility_row polls up to 30s until the matching results row has both name and plan before the PDF is printed (AngularJS fills cell text after the rows appear); used by all three MH eligibility workers - Fall back to the Member Detail heading for the name when the results table never filled it in Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-30 15:53:34 -04:00
parent 13be6e97d5
commit ebc5fba2c9
8 changed files with 219 additions and 232 deletions
@@ -0,0 +1,80 @@
+import time
+
+from selenium.webdriver.common.by import By
+
+# Cell texts of every row in the Eligible / Ineligible results tables.
+# innerText (not .text) so cells rendered off-screen still return their text.
+_ROWS_JS = """
+const out = [];
+for (const label of ['Eligible', 'Ineligible']) {
+  const rows = document.evaluate(
+    `//h4[normalize-space(text())='${label}']/following::table[1]/tbody/tr`,
+    document, null, XPathResult.ORDERED_NODE_SNAPSHOT_TYPE, null);
+  for (let i = 0; i < rows.snapshotLength; i++) {
+    const cells = Array.from(rows.snapshotItem(i).querySelectorAll('td'))
+      .map(td => (td.innerText || td.textContent || '').trim());
+    out.push([label, cells]);
+  }
+}
+return out;
+"""
+
+
+def _normalize_id(s):
+    return "".join(c for c in str(s) if c.isalnum()).lower()
+
+
+def extract_eligibility_row(driver, member_id, timeout=30):
+    """Wait for the MassHealth results row matching member_id and return
+    {eligibility, firstName, lastName, insurance}.
+
+    AngularJS adds the rows first and fills the cell text afterwards — the
+    member ID can appear before the name and plan do. Keep polling until the
+    matching row has both a name and a plan, so callers never print the PDF
+    before the data is captured. After the timeout, returns whatever was
+    found ({"eligibility": None} if no row matched).
+    """
+    norm_self = _normalize_id(member_id)
+    best = {"eligibility": None}
+    deadline = time.time() + timeout
+    attempt = 0
+
+    while True:
+        attempt += 1
+        try:
+            rows = driver.execute_script(_ROWS_JS) or []
+        except Exception as e:
+            print(f"[extraction] reading rows failed: {e}")
+            rows = []
+
+        for label, cells in rows:
+            if len(cells) < 3:
+                continue
+            norm_cell = _normalize_id(cells[2])
+            if not (norm_self and norm_cell and (norm_self in norm_cell or norm_cell in norm_self)):
+                continue
+
+            # name in cells[4], plan in cells[6] (fallback to last cell)
+            full_name = cells[4] if len(cells) > 4 else ""
+            plan_name = cells[6] if len(cells) > 6 else (cells[-1] if len(cells) > 4 else "")
+            name_parts = full_name.split()
+            best = {
+                "eligibility": "Y" if label == "Eligible" else "N",
+                "firstName": name_parts[0] if name_parts else "",
+                "lastName": " ".join(name_parts[1:]),
+                "insurance": plan_name,
+            }
+            if full_name and plan_name:
+                print(f"[extraction] MATCHED {label} → name='{full_name}' plan='{plan_name}' (attempt {attempt})")
+                return best
+            break
+
+        if time.time() >= deadline:
+            break
+        if attempt % 5 == 1:
+            state = "name/plan still empty" if best["eligibility"] else "no matching row yet"
+            print(f"[extraction] {state} for memberId='{member_id}' (attempt {attempt}), waiting...")
+        time.sleep(1)
+
+    print(f"[extraction] gave up after {timeout}s for memberId='{member_id}' → {best}")
+    return best