fix: DDMA claim checkbox, eligibility patient name, and ID+DOB patient lookup

- Claim step2: robust member checkbox selection with page-wide fallback
- Eligibility: take patient name from 'Benefits for' heading, not subscriber name
- Processor: update the patient just created/matched by ID+DOB, not first ID match

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

apps/SeleniumService/selenium_DDMA_eligibilityCheckWorker.py

@@ -542,9 +542,21 @@ class AutomationDeltaDentalMAEligibilityCheck:
                         return False
                     return True
 
+                # Preferred: the "Benefits for JOSHUA ZHANG" heading names the actual
+                # patient. "Subscriber name:" is the policy holder (e.g. a parent), so
+                # it is only used below if this heading is missing.
+                for line in page_lines:
+                    heading = re.match(r'^benefits\s+for\s+(.+)$', line, re.IGNORECASE)
+                    if heading and looks_like_name(heading.group(1)):
+                        patientName = heading.group(1).strip()
+                        print(f"[DDMA step2] Extracted name from 'Benefits for' heading: '{patientName}'")
+                        break
+
                 # Scan lines for the "Name" label.  When "Name" appears alone we scan
                 # forward past accessibility text to the first plausible name value.
                 for i, line in enumerate(page_lines):
+                    if patientName:
+                        break
                     # Case 1: "Name : Value" or "Name: Value" on the same line
                     # (DDMA's detail page labels this "Subscriber name:", so that
                     # prefix must be accepted rather than excluded)