se agrega integración con SAP HANA, manejo de errores en el demonio y actualización de documentación

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

pdf_utilities.py

@@ -134,55 +134,69 @@ def _extract_factura_details(text: str, data: dict) -> None:
     if table_start_index != -1:
         items_lines = lines[table_start_index + 1:]
         current_item = None
+        pre_item_buffer = []
         for line in items_lines:
             line = line.strip()
             if not line: continue
             if re.search(r"^(?:SUB|TOTAL|IGV|SON:|OP\.|RES|CUO|DET|BANCO|Representación|Podrá|Autorizado|FORMA|CUOTA|GRAVADO|I\.G\.V\.|DESEMBOLSO|OBSERVACION)", line, re.IGNORECASE):
                 break
-            amounts = re.findall(r"(?<![A-Z0-9-])(\d+(?:,\d{3})*(?:\.\d{2,}))(?![A-Z0-9-])", line)
+            amounts = re.findall(r"(?<![A-Z0-9-])(\d+(?:,\d{3})*(?:\.\d{2,}))(?:(?=\s)|(?=$)|(?=[A-Z][a-z]))", line)
             if len(amounts) >= 2:
                 if current_item: data["items"].append(current_item)
                 qty = amounts[0]
-                total = amounts[-1]
-                p_unit = amounts[-2]
-                if len(amounts) >= 4:
-                    p_unit = amounts[1]
+                p_unit = amounts[-2].replace(",", "")
+                total = amounts[-1].replace(",", "")
                 desc_line = line
-                for amt in [qty, p_unit, total]:
-                    desc_line = desc_line.replace(amt, "", 1)
+                seen = set()
+                for amt in amounts:
+                    if amt not in seen:
+                        desc_line = desc_line.replace(amt, "", 1)
+                        seen.add(amt)
                 desc_parts = [p.strip() for p in desc_line.split() if p.strip()]
-                codigo = None
-                for i in range(min(3, len(desc_parts))):
-                    if re.search(r"[A-Z].*\d|\d.*[A-Z]|[A-Z]{2,}-\d+", desc_parts[i]):
-                        codigo = desc_parts[i]
-                        desc_parts.pop(i)
+                cod = None
+                for j in range(min(3, len(desc_parts))):
+                    if re.search(r"[A-Z].*\d|\d.*[A-Z]|[A-Z]{2,}-\d+", desc_parts[j]):
+                        cod = desc_parts[j]
+                        desc_parts.pop(j)
                         break
                 if desc_parts and desc_parts[0].isdigit() and len(desc_parts[0]) <= 2:
                     desc_parts.pop(0)
                 descripcion = " ".join(desc_parts)
                 current_item = {
-                    "codigo": codigo,
+                    "codigo": cod,
                     "cantidad": qty,
                     "descripcion": descripcion,
-                    "precio_unitario": p_unit.replace(",", ""),
-                    "total": total.replace(",", ""),
+                    "precio_unitario": p_unit,
+                    "total": total,
                     "guias_remision": [],
                     "guia_transportista": None,
                     "bolsas": None,
                     "peso": None,
                     "unidad_peso": None,
                 }
+                if pre_item_buffer:
+                    current_item["descripcion"] = " ".join(pre_item_buffer) + " " + descripcion
+                    pre_item_buffer = []
             elif current_item and len(line) > 5:
                 current_item["descripcion"] += " " + line
+            elif not current_item and len(line) > 5:
+                pre_item_buffer.append(line)
         if current_item: data["items"].append(current_item)
 
+    # Extraer guías de cada ítem usando los mismos patrones que sunat_schema
+    GUIAS_PDF_PATTERNS = [
+        (r"(?:GRE\s+REMITENTE|GRR|G\.R\.R)\s*[:.]?\s*([TVE][A-Z0-9]{3,4})\s*-?\s*(\d+)", "09"),
+        (r"\b(?:GRT|G\.R\.T)\s*[:.]?\s*([TVE][A-Z0-9]{3,4})\s*-?\s*(\d+)", "31"),
+    ]
     for item in data["items"]:
         desc = item["descripcion"]
-        guias_remision_matches = re.findall(r"\b(?:GRR\w*|GR\b|G\.R\.R\.?|GUIA REMISI.N)\s*[:/\s]*((?:[TVE][A-Z0-9]{3})-\d+)", desc, re.IGNORECASE)
-        item["guias_remision"] = [m.strip() for m in guias_remision_matches] if guias_remision_matches else []
-        guia_transportista_match = re.search(r"\b(?:GRT\w*|GR\b|G\.R\.T\.?|GUIA TRANSPORTISTA)\s*[:/\s]*((?:[TVE][A-Z0-9]{3})-\d+)", desc, re.IGNORECASE)
-        if guia_transportista_match:
-            item["guia_transportista"] = guia_transportista_match.group(1).strip()
+        for patron, tipo in GUIAS_PDF_PATTERNS:
+            for m in re.finditer(patron, desc, re.IGNORECASE):
+                numero = m.group(2).lstrip("0") or "0"
+                if tipo == "09":
+                    item["guias_remision"].append(f"{m.group(1).upper()}-{numero}")
+                else:
+                    item["guia_transportista"] = f"{m.group(1).upper()}-{numero}"
         bolsas_match = re.search(r"(\d+)\s*(?:BOLSAS|BLS|BOLS|SACOS)\b", desc, re.IGNORECASE)
         if bolsas_match:
             item["bolsas"] = bolsas_match.group(1)