Update app.py

app.py

@@ -1522,6 +1522,70 @@ def _merge_split_rows(grid):
     return new_grid
 
 
+def _extract_fused_desc_amount(grid):
+    """
+    Fix OCR artifact where a trailing money amount gets fused into the
+    description cell instead of its own debit/credit column.
+
+    Example (Hardin County Bank):
+      OCR: ['CHASE CREDIT CRD EPAY 8319249882 500.00', '', '', '04/11/25', '16,699.04']
+      Fix: ['CHASE CREDIT CRD EPAY 8319249882', '500.00', '', '04/11/25', '16,699.04']
+
+    Guard conditions (all must hold):
+      - Description ends with a space + money amount (NNN.NN or N,NNN.NN).
+      - ALL debit AND credit columns are empty for this row.
+      - Date column is non-empty (confirms it is a complete data row).
+    """
+    if not grid or len(grid) < 2:
+        return grid
+
+    ncols = len(grid[0])
+    header = [str(c or "").strip().upper() for c in grid[0]]
+
+    date_col = desc_col = None
+    debit_cols = []
+    credit_cols = []
+    for ci, h in enumerate(header):
+        if re.search(r"\bDATE\b", h) and date_col is None:
+            date_col = ci
+        if re.search(r"\b(DESCRIPTION|DETAILS?|NARRATION|PARTICULARS?|TRANSACTION)\b", h) and desc_col is None:
+            desc_col = ci
+        if re.search(r"\b(DEBIT|DEBITS|DR|WITHDRAWAL|SUBTRACTIONS?)\b", h):
+            debit_cols.append(ci)
+        if re.search(r"\b(CREDIT|CREDITS|CR|DEPOSIT|ADDITIONS?)\b", h):
+            credit_cols.append(ci)
+
+    if date_col is None or desc_col is None or not debit_cols:
+        return grid
+
+    money_cols = debit_cols + credit_cols
+    _TRAILING_AMT = re.compile(r"^(.+?)\s+(-?\d{1,3}(?:,\d{3})*\.\d{2})$")
+    debit_col = debit_cols[0]
+
+    new_grid = [grid[0]]
+    for row in grid[1:]:
+        cells = [str(c or "").strip() for c in row]
+        while len(cells) < ncols:
+            cells.append("")
+
+        desc = cells[desc_col]
+        no_money = all(not cells[c] for c in money_cols)
+        has_date = bool(cells[date_col])
+
+        if desc and no_money and has_date:
+            m = _TRAILING_AMT.match(desc)
+            if m:
+                cells = list(cells)
+                cells[desc_col] = m.group(1).strip()
+                cells[debit_col] = m.group(2)
+                new_grid.append(cells)
+                continue
+
+        new_grid.append(cells)
+
+    return new_grid
+
+
 def normalize_html_tables(text: str) -> str:
     """
     For every <table>...</table>:
@@ -1566,6 +1630,10 @@ def normalize_html_tables(text: str) -> str:
             # (Hardin County Bank and similar monospace statement formats)
             grid = _merge_split_rows(grid)
 
+            # Fix 7: extract trailing amount fused into description cell
+            # (Hardin County Bank: "CHASE CREDIT CRD EPAY 8319249882 500.00")
+            grid = _extract_fused_desc_amount(grid)
+
             # Fix 3: fused key-value rows in summary sections (no-space guard — safe on all PDFs)
             grid = _fix_fused_keyvalue_rows(grid)