fix bs compare

周伟奇
Showing 3 changed files with 123 additions and 34 deletions
src/apps/doc/consts.py
src/apps/doc/ocr/wb.py
src/celery_compare/tasks.py
--- a/src/apps/doc/consts.py
View file @50de32a
+++ b/src/apps/doc/consts.py
View file @50de32a
@@ -1410,7 +1410,7 @@ SE_STAMP_VALUE = '有'
 SE_LAYOUT_VALUE = '旧版-旧打印、新版-新打印'
 SE_GB_NEW_FIELD = ['vinNo']
 SE_GB_USED_FIELD = ['customerName', 'idNum', 'date']
-SE_BS_FIELD = ['户名', '打印日期', '流水日期', '(担保人)户名', '(担保人)打印日期', '(担保人)流水日期']
+SE_BS_FIELD = ['户名', '打印日期', '流水日期', '(担保人1)户名', '(担保人1)打印日期', '(担保人1)流水日期', '(担保人2)户名', '(担保人2)打印日期', '(担保人2)流水日期']
 SE_HMH_FIELD = ['借款人/承租人姓名', '借款人/承租人证件号', '申请号', '渠道', '签字']
 SE_BD_FIELD = ['被保险人姓名', '被保险人证件号码', '车架号', '机动车损失保险金额', '第三者责任保险金额', '绝对免赔率', '保险起始日期', '保险截止日期', '保单章', '第一受益人', '保险费合计']
 JDMPV_VALUE = ['-', '--', '0%', '0.00', '/', '0']
@@ -1781,11 +1781,14 @@ BD_COMPARE_LOGIC = {
 BS_COMPARE_LOGIC = {
    '户名': ('role', 'se_one_compare', {}, '主共借人未提供银行流水'),
-    '(担保人)户名': ('role', 'se_one_compare', {}, '担保人未提供银行流水'),
+    '(担保人1)户名': ('role', 'se_name_compare', {}, '担保人1未提供银行流水'),
+    '(担保人2)户名': ('role', 'se_name_compare', {}, '担保人2未提供银行流水'),
    '打印日期': ('print_time', 'se_bs_print_date_compare', {}, '主共借人银行流水打印日期超过15天'),
-    '(担保人)打印日期': ('print_time', 'se_bs_print_date_compare', {}, '担保人银行流水打印日期超过15天'),
+    '(担保人1)打印日期': ('print_time', 'se_bs_print_date_compare', {}, '担保人1银行流水打印日期超过15天'),
+    '(担保人2)打印日期': ('print_time', 'se_bs_print_date_compare', {}, '担保人2银行流水打印日期超过15天'),
    '流水日期': ('date', 'se_bs_date_compare', {}, '主共借人银行流水日期不满足3个月'),
-    '(担保人)流水日期': ('date', 'se_bs_date_compare', {}, '担保人银行流水日期不满足3个月'),
+    '(担保人1)流水日期': ('date', 'se_bs_date_compare', {}, '担保人1银行流水日期不满足3个月'),
+    '(担保人2)流水日期': ('date', 'se_bs_date_compare', {}, '担保人2银行流水日期不满足3个月'),
 }
 SPECIAL_REASON = '主共借人未提供银行流水，含担保人需人工查看直系亲属关系'
--- a/src/apps/doc/ocr/wb.py
View file @50de32a
+++ b/src/apps/doc/ocr/wb.py
View file @50de32a
@@ -312,21 +312,21 @@ class BSWorkbook(Workbook):
                            (ws.title, idx_list[i] + min_row, idx_list[i + 1] + min_row - 1, day_mean))
    def build_metadata_rows(self, confidence, code, verify_list, print_time, start_date, end_date, res_count_tuple, is_verify_classify):
-        metadata_rows = [
+        metadata_rows = [('流水识别置信度', confidence)]
-            ('流水识别置信度', confidence),
-            ('图片总数', res_count_tuple[0]),
-            ('识别成功', res_count_tuple[1]),
-            self.blank_row,
-        ]
-        verify_highlight_row = []
        if is_verify_classify:
            verify_res = '疑似伪造' if len(verify_list) > 0 else '正常'
+        else:
+            verify_res = ''
        metadata_rows.append(('流水检测结果', verify_res))
-            if len(verify_list) > 0:
+        metadata_rows.append(('图片总数', res_count_tuple[0]))
+        metadata_rows.append(('识别成功', res_count_tuple[1]))
+        metadata_rows.append(self.blank_row)
+        verify_highlight_row = []
+        if is_verify_classify and len(verify_list) > 0:
            metadata_rows.append(self.verify_header)
            metadata_rows.extend(verify_list)
-            for r in range(5, len(metadata_rows)+1):
+            for r in range(6, len(metadata_rows)+1):
                verify_highlight_row.append(r)
        metadata_rows.append(self.blank_row)
@@ -355,14 +355,18 @@ class BSWorkbook(Workbook):
        ms = self.create_sheet('{0}{1}({2})'.format(self.meta_sheet_title, role_name, card))
        for row in metadata_rows:
            ms.append(row)
+        if len(verify_highlight_row) > 0:
+            for cell in ms[2]:
+                cell.fill = self.amount_fill
        for row in verify_highlight_row:
            for cell in ms[row]:
                cell.fill = self.amount_fill
        if res_count_tuple[0] != res_count_tuple[1]:
-            for cell in ms[2]:
-                cell.fill = self.amount_fill
            for cell in ms[3]:
                cell.fill = self.amount_fill
+            for cell in ms[4]:
+                cell.fill = self.amount_fill
        return ms
    @staticmethod
--- a/src/celery_compare/tasks.py
View file @50de32a
+++ b/src/celery_compare/tasks.py
View file @50de32a
@@ -880,7 +880,7 @@ def get_se_cms_compare_info(last_obj, application_entity, detect_list):
    company_info_list = []
    # 个人信息证件---------------------------------------------------------------------------------------------------------
-    province = cms_info.get('province', '')
+    # province = cms_info.get('province', '')
    for individual_info in cms_info.get('applicantInformation', []):
        all_id_num = []
@@ -1082,12 +1082,19 @@ def get_se_cms_compare_info(last_obj, application_entity, detect_list):
        dbr_bs_role_list = []
        for dbr_bs_role, _, _ in main_role_info.get(consts.APPLICANT_TYPE_ORDER[2], []):
            dbr_bs_role_list.append(dbr_bs_role)
-        if len(dbr_bs_role_list) > 0:
+        if len(dbr_bs_role_list) >= 1:
            bs_field_input.extend([
-                (consts.SE_BS_FIELD[3], dbr_bs_role_list),
+                (consts.SE_BS_FIELD[3], dbr_bs_role_list[0]),
                (consts.SE_BS_FIELD[4], first_submission_date),
                (consts.SE_BS_FIELD[5], '90天'),
            ])
+            if len(dbr_bs_role_list) >= 2:
+                bs_field_input.extend([
+                    (consts.SE_BS_FIELD[6], dbr_bs_role_list[1]),
+                    (consts.SE_BS_FIELD[7], first_submission_date),
+                    (consts.SE_BS_FIELD[8], '90天'),
+                ])
        bs_info[consts.BS_EN] = bs_field_input
        compare_info['Bank Statement'] = bs_info
@@ -1522,7 +1529,8 @@ def se_bs_compare(license_en, ocr_res_dict, strip_list):
    if ocr_res_str is not None:
        pre_field_list = strip_list[:3]
-        suf_field_list = strip_list[3:]
+        dbr1_field_list = strip_list[3:6]
+        dbr2_field_list = strip_list[6:]
        ocr_res_list = json.loads(ocr_res_str)
        # length = len(ocr_res_list)
@@ -1558,14 +1566,46 @@ def se_bs_compare(license_en, ocr_res_dict, strip_list):
                max_correct_count = correct_count
                pre_best_res = pre_tmp_res_part
-        # 担保人
+        # 担保人1
-        suf_best_res = {}
+        dbr1_best_res = {}
-        if len(suf_field_list) > 0:
+        if len(dbr1_field_list) > 0:
+            max_correct_count = 0
+            for ocr_res in ocr_res_list:
+                correct_count = 0
+                dbr1_tmp_res_part = {}
+                for idx, (name, value) in enumerate(dbr1_field_list):
+                    ocr_str_or_list = ocr_res.get(compare_logic[name][0])
+                    if isinstance(ocr_str_or_list, str) or isinstance(ocr_str_or_list, list):
+                        result = getattr(cp, compare_logic[name][1])(value, ocr_str_or_list, **compare_logic[name][2])
+                        if isinstance(ocr_str_or_list, list):
+                            ocr_str = json.dumps(ocr_str_or_list, ensure_ascii=False)
+                        else:
+                            ocr_str = ocr_str_or_list
+                        reason = compare_logic[name][3]
+                    else:
+                        result = consts.RESULT_N
+                        ocr_str = empty_str
+                        reason = compare_logic[name][3]
+                    if idx == 0 and result == consts.RESULT_N:
+                        break
+                    if result == consts.RESULT_Y:
+                        correct_count += 1
+                    dbr1_tmp_res_part[name] = (result, ocr_str, reason)
+                if correct_count > 0 and correct_count >= max_correct_count:
+                    max_correct_count = correct_count
+                    dbr1_best_res = dbr1_tmp_res_part
+        # 担保人2
+        dbr2_best_res = {}
+        if len(dbr1_field_list) > 0:
            max_correct_count = 0
            for ocr_res in ocr_res_list:
                correct_count = 0
-                suf_tmp_res_part = {}
+                dbr2_tmp_res_part = {}
-                for idx, (name, value) in enumerate(suf_field_list):
+                for idx, (name, value) in enumerate(dbr2_field_list):
                    ocr_str_or_list = ocr_res.get(compare_logic[name][0])
                    if isinstance(ocr_str_or_list, str) or isinstance(ocr_str_or_list, list):
                        result = getattr(cp, compare_logic[name][1])(value, ocr_str_or_list, **compare_logic[name][2])
@@ -1584,11 +1624,30 @@ def se_bs_compare(license_en, ocr_res_dict, strip_list):
                    if result == consts.RESULT_Y:
                        correct_count += 1
-                    suf_tmp_res_part[name] = (result, ocr_str, reason)
+                    dbr2_tmp_res_part[name] = (result, ocr_str, reason)
                if correct_count > 0 and correct_count >= max_correct_count:
                    max_correct_count = correct_count
-                    suf_best_res = suf_tmp_res_part
+                    dbr2_best_res = dbr2_tmp_res_part
+        dbr_ok = False
+        # 有担保人
+        if len(dbr1_field_list) > 0:
+            # 有担保人12
+            if len(dbr2_field_list) > 0:
+                if len(dbr1_best_res) > 0 and len(dbr2_best_res) > 0:
+                    dbr_ok = True
+            # 有担保人1
+            else:
+                if len(dbr1_best_res) > 0:
+                    dbr_ok = True
+        # 无担保人
+        # else:
+        #     pass
+        best_res_empty = False
+        if len(pre_best_res) == 0 and len(dbr1_best_res) == 0 and len(dbr2_best_res) == 0:
+            best_res_empty = True
        for name, value in pre_field_list:
            if len(pre_best_res) > 0:
@@ -1596,7 +1655,9 @@ def se_bs_compare(license_en, ocr_res_dict, strip_list):
            else:
                result = consts.RESULT_N
                ocr_str = empty_str
-                if len(suf_field_list) > 0 and len(suf_best_res) > 0:
+                if best_res_empty:
+                    reason = consts.SPECIAL_REASON_3
+                elif dbr_ok:  # 有担保人且担保人都提供了流水
                    reason = consts.SPECIAL_REASON
                else:
                    reason = compare_logic[pre_field_list[0][0]][3]
@@ -1606,17 +1667,38 @@ def se_bs_compare(license_en, ocr_res_dict, strip_list):
                value = json.dumps(value, ensure_ascii=False)
            result_field_list.append((name, value, result, ocr_str, img_path, error_type, reason))
-        if len(suf_field_list) > 0:
+        if len(dbr1_field_list) > 0:
-            for name, value in suf_field_list:
+            for name, value in dbr1_field_list:
-                if len(suf_best_res) > 0:
+                if len(dbr1_best_res) > 0:
-                    result, ocr_str, reason = suf_best_res[name]
+                    result, ocr_str, reason = dbr1_best_res[name]
                else:
                    result = consts.RESULT_N
                    ocr_str = empty_str
-                    if len(pre_best_res) > 0:
+                    if best_res_empty:
+                        reason = consts.SPECIAL_REASON_3
+                    elif len(pre_best_res) > 0:
+                        reason = consts.SPECIAL_REASON_2
+                    else:
+                        reason = compare_logic[dbr1_field_list[0][0]][3]
+                img_path = empty_str
+                error_type = empty_error_type if result == consts.RESULT_Y else ErrorType.OCR.value
+                if isinstance(value, list):
+                    value = json.dumps(value, ensure_ascii=False)
+                result_field_list.append((name, value, result, ocr_str, img_path, error_type, reason))
+        if len(dbr2_field_list) > 0:
+            for name, value in dbr2_field_list:
+                if len(dbr2_best_res) > 0:
+                    result, ocr_str, reason = dbr2_best_res[name]
+                else:
+                    result = consts.RESULT_N
+                    ocr_str = empty_str
+                    if best_res_empty:
+                        reason = consts.SPECIAL_REASON_3
+                    elif len(pre_best_res) > 0:
                        reason = consts.SPECIAL_REASON_2
                    else:
-                        reason = compare_logic[suf_field_list[0][0]][3]
+                        reason = compare_logic[dbr2_field_list[0][0]][3]
                img_path = empty_str
                error_type = empty_error_type if result == consts.RESULT_Y else ErrorType.OCR.value
                if isinstance(value, list):