Update app.py

app.py

@@ -834,8 +834,9 @@ def _promote_misplaced_header_row(grid):
 # Matches a value suffix fused onto a label with no space.
 # e.g. "Beginning Interest Rate0.00%" → label + "0.00%"
 #      "Number of days in this Period31" → label + "31"
+# Strict: no \s* between groups so label trailing-space detects space-separated values
 _FUSED_KV_RE = re.compile(
-    r"^(.+?)\s*(-?\d{1,3}(?:,\d{3})*(?:\.\d+)?%?)$",
+    r"^(.+?)(-?\d{1,3}(?:,\d{3})*(?:\.\d+)?%?)$",
     re.DOTALL,
 )
 
@@ -846,14 +847,28 @@ def _fix_fused_keyvalue_rows(grid):
     cells empty — common in Interest Summary / Fee Summary sections appended
     to a transaction table by OCR.
 
+    e.g. "Beginning Interest Rate0.00%"  → label="Beginning Interest Rate"  value="0.00%"
+         "Number of days in this Period31" → label="..."  value="31"
+
     Guard conditions (no-op if not met — safe for all other tables):
       - All cells except first must be empty.
       - A non-space character must immediately precede the digit run (fused).
       - Both label and value parts must be non-empty after splitting.
+      - The cell must NOT contain an account number (7+ digit run) — prevents
+        incorrectly splitting "STREAMLINED CHECKING #208479667" or
+        "Charges debited from account #208479667".
+      - The cell must NOT contain a proper dollar amount with decimal — prevents
+        incorrectly splitting metadata like "Beginning Balance:$20.48$3.46".
     """
     if not grid or len(grid) < 2:
         return grid
 
+    # Matches a proper dollar/currency amount with decimal point — these appear
+    # in legitimate label cells and must not trigger the fused-KV split.
+    _DOLLAR_AMOUNT_RE = re.compile(r"\$\s*\d{1,3}(?:,\d{3})*\.\d{2}")
+    # Matches a long account/reference number (7+ consecutive digits)
+    _LONG_NUMBER_RE = re.compile(r"\d{7,}")
+
     ncols = len(grid[0])
     new_grid = [grid[0]]
 
@@ -870,8 +885,15 @@ def _fix_fused_keyvalue_rows(grid):
             new_grid.append(row)
             continue
 
-        # Require no-space immediately before the digit run
-        if not re.search(r"\S\d", first):
+        # Guard: skip rows containing a proper dollar amount with decimal
+        # (these are metadata/balance rows, not fused key-value summary rows)
+        if _DOLLAR_AMOUNT_RE.search(first):
+            new_grid.append(row)
+            continue
+
+        # Guard: skip rows containing a long account/reference number (7+ digits)
+        # (e.g. "STREAMLINED CHECKING #208479667", "account #208479667")
+        if _LONG_NUMBER_RE.search(first):
             new_grid.append(row)
             continue
 
@@ -880,9 +902,16 @@ def _fix_fused_keyvalue_rows(grid):
             new_grid.append(row)
             continue
 
-        label = m.group(1).strip()
+        label_raw = m.group(1)   # trailing space means value was space-separated, not fused
         value = m.group(2).strip()
 
+        # Space before value means NOT fused — "Interest Rate 0.00%" is clean
+        if label_raw != label_raw.rstrip():
+            new_grid.append(row)
+            continue
+
+        label = label_raw.strip()
+
         if not label or not value:
             new_grid.append(row)
             continue