Skip to content

周伟奇 / bmw-ocr

Go to a project
Toggle navigation
Toggle navigation pinning
3b1e6657 by 周伟奇
Options

fix folder

1 parent 6e9f7b32
Expand all
Inline Side-by-side
Showing 3 changed files with 20 additions and 6 deletions
src/apps/doc/management/commands/folder_wsc_process.py
......@@ -48,6 +48,7 @@ class Finder:
"其他约定与条件英文": "",
"其他约定与条件中文": "",
}
def get_line(self, ocr_results, key_string):
# 根据指定关键词, 找出与关键词同处一行的字符
top, bottom = -1, -1
......@@ -69,6 +70,7 @@ class Finder:
line_text = sorted(line_text, key=lambda x: x[0][0], reverse=False)
lines = ''.join([i[1] for i in line_text])
return lines
def page_predict(self, ocr_results, page_template):
classes = []
for pno in ocr_results:
......@@ -82,6 +84,7 @@ class Finder:
classes.append([pno, score])
pred = sorted(classes, key=lambda x: x[1], reverse=True)[0]
return pred
def get_top_key(self, ocr_results, key_string): # 加入过滤词典
"""找到与 key_string 最匹配的字段的 key
"""
......@@ -90,6 +93,7 @@ class Finder:
ratio_list = [[fuzz.ratio(key_string, ocr_results[key][1]), key] for key in ocr_results]
top_key = sorted(ratio_list, key=lambda x: x[0])[-1]
return top_key
def get_top_iou(self, ocr_results, poly):
"""求最大IoU
"""
......@@ -108,6 +112,7 @@ class Finder:
return -1, -1
top_iou = sorted(iou_list, key=lambda x: x[0])[-1]
return top_iou
def get_key_value(self, ocr_results, key_string):
"""根据 key 查找 value
"""
......@@ -139,6 +144,7 @@ class Finder:
else:
value = words
return value
def get_contract_No(self):
"""提取左上角的合同编号字段
"""
......@@ -153,6 +159,7 @@ class Finder:
# TODO！！！
contract_No_list.append(contract_No)
return contract_No_list
def get_info_in_page_3(self):
"""提取第三页上的经销商名称,和经销商统一社会信用代码或公司注册号
"""
......@@ -178,6 +185,7 @@ class Finder:
words = self.get_key_value(self.ocr_results[pno], '统一社会信用代码或公司注册号')
dealer_No = words.replace('O', '0')
return dealer_name, dealer_No
def get_info_in_page_38(self):
"""提取第38页上的经销商名称
"""
......@@ -195,6 +203,7 @@ class Finder:
words = re.sub(r'[(（）)盖章《]', "", words)
dealer_name = words
return dealer_name
def get_guarantor(self):
"""提取第10页上保证人段落,所见即所得
"""
......@@ -210,6 +219,7 @@ class Finder:
words = words.replace('【', '[').replace('】', ']').replace(',', '，').replace('(', '（').replace(')', '）')
guarantor = words
return guarantor
def get_info_in_page_39(self):
"""提取综合授信合同上的一些字段
"""
......@@ -291,6 +301,7 @@ class Finder:
deposit_chn = f'{words}%'
return amount_eng, amount_chn, term_start_eng, term_end_eng, \
term_start_chn, term_end_chn, deposit_eng, deposit_chn
def get_other_arrangements_and_conditions(self):
"""获取其它约定与条件文本段落
"""
......@@ -311,6 +322,7 @@ class Finder:
words = searchObj.group(1)
other_arrangements_and_conditions_chn = words
return other_arrangements_and_conditions_eng, other_arrangements_and_conditions_chn
def get_info(self):
# 按照文档页码返回一个合同编号列表,依次表示每一页上识别到的合同编号
contract_No_list = self.get_contract_No()
......@@ -337,6 +349,7 @@ class Finder:
self.init_result["其他约定与条件中文"] = words_chn
return self.init_result
class TIFFHandler:
def __init__(self, path, img_save_path):
......@@ -568,6 +581,7 @@ class Command(BaseCommand, LoggerMixin):
if len(true_file_set) == 0 and len(os_error_filename_set) > 0:
true_file_set.add(os_error_filename_set.pop())
for name in true_file_set:
time.sleep(10) # 防止文件较大时，读取到不完整文件
path = os.path.join(input_dir, name)
try:
......
src/celery_compare/tasks.py
......@@ -989,7 +989,7 @@ def get_se_cms_compare_info_auto(last_obj, application_entity):
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[0], hmh_name))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[1], hmh_id))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[2], first_submission_date))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
# vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[4], consts.SE_FPL_VALUE))
bhsj = float(amount) / 1.13
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[5], consts.SPLIT_STR.join([
......@@ -1464,7 +1464,7 @@ def get_se_cms_compare_info(last_obj, application_entity, detect_list):
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[0], hmh_name))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[1], hmh_id))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[2], first_submission_date))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
# vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[4], consts.SE_FPL_VALUE))
bhsj = float(amount) / 1.13
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[5], consts.SPLIT_STR.join([
......@@ -1499,7 +1499,7 @@ def get_se_cms_compare_info(last_obj, application_entity, detect_list):
gb34_field_input.append((consts.SE_GB_USED_FIELD[1], main_num))
gb34_field_input.append((consts.SE_GB_USED_FIELD[2], first_submission_date))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[2], first_submission_date))
vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
# vehicle_field_input.append((consts.SE_NEW_ADD_FIELD[3], consts.SE_STAMP_VALUE))
jypz_field_input.append(('vinNo', vin_no))
jypz_field_input.append(('vehicleTransactionAmount', amount))
jypz_field_input.append((consts.SE_GB_USED_FIELD[-1], first_submission_date))
......@@ -2683,9 +2683,9 @@ def se_compare_process(compare_info, ocr_res_dict, is_gsyh):
for i in cn_reason_list:
if i in tmp_set:
continue
elif i in consts.BS_REASON:
tmp_set.add(i)
bs_cn_reason_list.append(i)
# elif i in consts.BS_REASON:
# tmp_set.add(i)
# bs_cn_reason_list.append(i)
else:
tmp_set.add(i)
last_cn_reason_list.append(i)
......
Styling with Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!