add qrs ocr

@@ -1042,6 +1042,10 @@ JYPZ_FIELD_ORDER = (("type", "标题"),
 CONTRACT_CN_NAME = '合同'
 CONTRACT_CLASSIFY = 41
 
+# 合同-送达地址确认书
+CONTRACT_QRS_CN_NAME = '送达地址确认书'
+CONTRACT_QRS_CLASSIFY = 49
+
 # 合同编号: 每页
 
 HIL_CONTRACT_1_CN_NAME = '售后回租合同'
@@ -1053,13 +1057,14 @@ HIL_CONTRACT_2_CLASSIFY = 44
 HIL_CONTRACT_3_CN_NAME = '车辆处置协议'
 HIL_CONTRACT_3_CLASSIFY = 45
 
-CONTRACT_SET = {CONTRACT_CLASSIFY, HIL_CONTRACT_1_CLASSIFY, HIL_CONTRACT_2_CLASSIFY, HIL_CONTRACT_3_CLASSIFY}
+CONTRACT_SET = {CONTRACT_QRS_CLASSIFY, CONTRACT_CLASSIFY, HIL_CONTRACT_1_CLASSIFY, HIL_CONTRACT_2_CLASSIFY, HIL_CONTRACT_3_CLASSIFY}
 
 CONTRACT_MAP = {
     HIL_CONTRACT_1_CLASSIFY: HIL_CONTRACT_1_CN_NAME,
     HIL_CONTRACT_2_CLASSIFY: HIL_CONTRACT_2_CN_NAME,
     HIL_CONTRACT_3_CLASSIFY: HIL_CONTRACT_3_CN_NAME,
     CONTRACT_CLASSIFY: CONTRACT_CN_NAME,
+    CONTRACT_QRS_CLASSIFY: CONTRACT_QRS_CN_NAME,
 }
 
 # 保单
@@ -1203,6 +1208,7 @@ DDA_OCR_FIELD = 'bs_ocr'
 HMH_OCR_FIELD = 'hmh_ocr'
 JYPZ_OCR_FIELD = 'jypz_ocr'
 HT_FIELD = 'ht_ocr'
+QRS_FIELD = 'qrs_ocr'
 BD_FIELD = 'bd_ocr'
 BS_FIELD = 'bss_ocr'
 HIL_CONTRACT_1_FIELD = 'hil_contract_1_ocr'
@@ -1226,6 +1232,7 @@ RESULT_MAPPING = {
     HMH_CLASSIFY: HMH_OCR_FIELD,
     JYPZ_CLASSIFY: JYPZ_OCR_FIELD,
     CONTRACT_CLASSIFY: HT_FIELD,
+    CONTRACT_QRS_CLASSIFY: QRS_FIELD,
     INSURANCE_CLASSIFY: BD_FIELD,
     BS_CLASSIFY: BS_FIELD,
     HIL_CONTRACT_1_CLASSIFY: HIL_CONTRACT_1_FIELD,
@@ -1479,6 +1486,10 @@ AFC_CON_FIELD_ORDER_LTGT = (
     ('标准利率', '标准利率'),
 )
 
+SE_AFC_CON_QRS_MAP = {
+    '合同编号': (1, '合同编号'),
+}
+
 SE_AFC_CON_MAP = {
     '合同编号-每页': (None, None, '合同编号', None),
     '所购车辆价格-小写-重要条款': (1, 1, '所购车辆价格', None),
@@ -2308,6 +2319,7 @@ FILE_NAME_PREFIX_MAP = {
 ECONTRACT_KEYWORDS_MAP = {
     AFC_PREFIX: [
         ('抵押贷款合同', CONTRACT_CLASSIFY),
+        ('送达地址确认书', CONTRACT_QRS_CLASSIFY),
         # ('电子签署-抵押登记豁免函', HMH_CLASSIFY, 0),
     ],
     HIL_PREFIX: [

@@ -987,6 +987,15 @@ class Command(BaseCommand, LoggerMixin):
                 #     }
                 # }
                 license_summary[classify] = [res]
+            elif classify == consts.CONTRACT_QRS_CLASSIFY:
+                res = {}
+                for key, (pno, key1) in consts.SE_AFC_CON_QRS_MAP.items():
+                    res[key] = page_info_dict.get(str(pno), {}).get(key1, '')
+                    res.setdefault(consts.IMG_PATH_KEY, dict())[key] = page_info_dict.get(str(pno), {}).get(
+                        consts.IMG_PATH_KEY, '')
+                    res.setdefault(consts.ALL_POSITION_KEY, dict())[key] = page_info_dict.get(str(pno), {}).get(
+                        consts.ALL_POSITION_KEY, {}).get(key1, [])
+                license_summary[classify] = [res]
             else:
                 res = {}
                 for key, (pno1, pno2, end_idx, key1, key2) in consts.SE_HIL_CON_MAP[classify].items():
@@ -1474,6 +1483,16 @@ class Command(BaseCommand, LoggerMixin):
                                         'page_num': page_num,
                                         'page_info': page_info
                                     }
+                        elif classify_1_str == str(consts.CONTRACT_QRS_CLASSIFY):
+                            ocr_result = afc_predict(pdf_handler.pdf_info, is_qrs=True)
+                            page_num = 'page_1'
+                            page_res = {
+                                page_num: {
+                                    'classify': int(classify_1_str),
+                                    'page_num': page_num,
+                                    'page_info': ocr_result.pop(page_num, {}) 
+                                }
+                            }
                         else:
                             file_type_1 = consts.HIL_CONTRACT_TYPE_MAP.get(classify_1_str)
                             ocr_result_1 = hil_predict(pdf_handler.pdf_info, file_type_1)

@@ -328,6 +328,7 @@ class AFCOCRResult(models.Model):
     hil_contract_1_ocr = models.TextField(null=True, verbose_name="HIL合同1")
     hil_contract_2_ocr = models.TextField(null=True, verbose_name="HIL合同2")
     hil_contract_3_ocr = models.TextField(null=True, verbose_name="HIL合同3")
+    qrs_ocr = models.TextField(null=True, verbose_name="AFC合同确认书")
 
     update_time = models.DateTimeField(auto_now=True, verbose_name='修改时间')
     create_time = models.DateTimeField(auto_now_add=True, verbose_name='创建时间')
@@ -363,6 +364,7 @@ class HILOCRResult(models.Model):
     hil_contract_1_ocr = models.TextField(null=True, verbose_name="HIL合同1")
     hil_contract_2_ocr = models.TextField(null=True, verbose_name="HIL合同2")
     hil_contract_3_ocr = models.TextField(null=True, verbose_name="HIL合同3")
+    qrs_ocr = models.TextField(null=True, verbose_name="AFC合同确认书")
 
     update_time = models.DateTimeField(auto_now=True, verbose_name='修改时间')
     create_time = models.DateTimeField(auto_now_add=True, verbose_name='创建时间')
@@ -397,6 +399,7 @@ class AFCSEOCRResult(models.Model):
     hil_contract_1_ocr = models.TextField(null=True, verbose_name="HIL合同1")
     hil_contract_2_ocr = models.TextField(null=True, verbose_name="HIL合同2")
     hil_contract_3_ocr = models.TextField(null=True, verbose_name="HIL合同3")
+    qrs_ocr = models.TextField(null=True, verbose_name="AFC合同确认书")
 
     update_time = models.DateTimeField(auto_now=True, verbose_name='修改时间')
     create_time = models.DateTimeField(auto_now_add=True, verbose_name='创建时间')
@@ -432,6 +435,7 @@ class HILSEOCRResult(models.Model):
     hil_contract_1_ocr = models.TextField(null=True, verbose_name="HIL合同1")
     hil_contract_2_ocr = models.TextField(null=True, verbose_name="HIL合同2")
     hil_contract_3_ocr = models.TextField(null=True, verbose_name="HIL合同3")
+    qrs_ocr = models.TextField(null=True, verbose_name="AFC合同确认书")
 
     update_time = models.DateTimeField(auto_now=True, verbose_name='修改时间')
     create_time = models.DateTimeField(auto_now_add=True, verbose_name='创建时间')

@@ -9,7 +9,21 @@ from .get_char import Finder
 import numpy as np
 
 
-def predict(pdf_info):
+def extract_info(ocr_results):
+    contract_no = {
+        "words": None,
+        "position": None
+    } 
+    for bbox, text in ocr_results.get('0', {}).values():
+        if text.startswith('CH-B'):
+            contract_no['words'] = text 
+            contract_no['position'] = [bbox[0], bbox[1], bbox[2], bbox[-1]] 
+            break
+
+    return {'page_1': {'合同编号': contract_no}}
+
+
+def predict(pdf_info, is_qrs=False):
     ocr_results = {}
     for pno in pdf_info:
         ocr_results[pno] = {}
@@ -32,9 +46,12 @@ def predict(pdf_info):
         keys = list(range(len(ocr_result)))
         ocr_result = dict(zip(keys, ocr_result))
         ocr_results[pno] = ocr_result
-    # 输入是整个 PDF 中的信息
-    f = Finder(pdf_info, ocr_results=ocr_results)
-    results = f.get_info()
+    if is_qrs:
+        results = extract_info(ocr_results)
+    else:
+        # 输入是整个 PDF 中的信息
+        f = Finder(pdf_info, ocr_results=ocr_results)
+        results = f.get_info()
     return results
 
 

+import pyodbc
+
+hil_sql = """
+    ALTER TABLE hil_ocr_result ADD qrs_ocr nvarchar(max);
+    ALTER TABLE hil_se_ocr_result ADD qrs_ocr nvarchar(max);
+"""
+
+afc_sql = """
+    ALTER TABLE afc_ocr_result ADD qrs_ocr nvarchar(max);
+    ALTER TABLE afc_se_ocr_result ADD qrs_ocr nvarchar(max);
+"""
+
+hil_cnxn = pyodbc.connect('DRIVER={ODBC Driver 17 for SQL Server};', autocommit=True)
+
+hil_cursor = hil_cnxn.cursor()
+hil_cursor.execute(hil_sql)
+
+hil_cursor.close()
+hil_cnxn.close()
+
+afc_cnxn = pyodbc.connect('DRIVER={ODBC Driver 17 for SQL Server};', autocommit=True)
+
+afc_cursor = afc_cnxn.cursor()
+afc_cursor.execute(afc_sql)
+
+afc_cursor.close()
+afc_cnxn.close()