4x-tech/hf-test
2
1
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

update parsing markdown

Browse Source
This commit is contained in:
haitaoxing
2025-07-22 18:24:00 +08:00
parent 58f91ae65e
commit d2f2240a6c
2 changed files with 179 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files

138
docext-test/docext_api_test.py
View File

@@ -77,7 +77,7 @@ markdown_text = """ Page 1 of 1
<td>5.98</td> <td>5.98</td>
</tr> </tr>
<tr> <tr>
<td>*计算机配套产品*金骏芦宝 24V1A电源适配器1000mA适用于按摩器甩脂机瘦身腰带LED台灯吸尘器扫地机器人加湿器充电器电源线</td> <td>*24V2A电源适配器1000mA</td>
<td></td> <td></td>
<td></td> <td></td>
<td></td> <td></td>
@@ -116,12 +116,15 @@ markdown_text = """ Page 1 of 1
def extract_invoice_info(markdown_text): def extract_invoice_info(markdown_text):
try: try:
# 提取发票号码 # 提取发票号码
invoice_match = re.search(r'发票号码:\s*(\d+)', markdown_text) invoice_match = re.search(r'发票号码:\s*(\d+)', markdown_text)
if not invoice_match: if not invoice_match:
raise ValueError("未找到发票号码信息") raise ValueError("未找到发票号码信息")
invoice_number = invoice_match.group(1) invoice_number = invoice_match.group(1)
print("invoice_number:", invoice_number) # print("invoice_number:", invoice_number)
# 提取销售方名称 # 提取销售方名称
seller_section = markdown_text.split('销售方信息') seller_section = markdown_text.split('销售方信息')
if len(seller_section) < 2: if len(seller_section) < 2:
@@ -130,49 +133,101 @@ def extract_invoice_info(markdown_text):
if not seller_match: if not seller_match:
raise ValueError("未找到销售方名称") raise ValueError("未找到销售方名称")
seller_name = seller_match.group(1).strip() seller_name = seller_match.group(1).strip()
print("seller_name:", seller_name) # print("seller_name:", seller_name)
# 修正金额正则表达式（移除$符号） # 修正金额正则表达式（移除$符号）
# amount_match = re.search(r'小写\s*￥(\d+\.\d+)', markdown_text) # amount_match = re.search(r'小写\s*￥(\d+\.\d+)', markdown_text)
amount_match = re.search(r'\(小写\)\s*￥(\d+\.\d+)', markdown_text) amount_match = re.search(r'\(小写\)\s*￥(\d+\.\d+)', markdown_text)
if not amount_match: if not amount_match:
raise ValueError("未找到金额信息") raise ValueError("未找到金额信息")
amount = amount_match.group(1) amount = amount_match.group(1)
print("amount:", amount)
# 构建基础数据
invoice_data = {
"invoice_number": invoice_number,
"seller_name": seller_name,
"total_amount": amount,
"items": []
}
# print("amount:", amount)
# 提取商品明细 # 提取商品明细
item_section = markdown_text.split('<tbody>') item_section = markdown_text.split('<tbody>')
if len(item_section) < 2: if len(item_section) < 2:
raise ValueError("未找到商品明细部分") raise ValueError("未找到商品明细部分")
print("发票号码:", invoice_number)
print("销售方名称:", seller_name)
print("金额:", amount)
# 修正表格解析逻辑 # 修正表格解析逻辑
table_rows = re.findall(r'<tr>.*?</tr>', item_section[1],re.DOTALL) table_rows = re.findall(r'<tr>.*?</tr>', item_section[1],re.DOTALL)
if len(table_rows) < 2: if len(table_rows) < 3:
raise ValueError("商品明细数据不完整") raise ValueError("商品明细数据不完整")
for row in table_rows[:-2]:
# print("row:", row)
item_name_match = re.search(r'<td>(.*?)</td>', row)
model_match = re.search(r'<td[^>]*>.*?</td>\s*<td[^>]*>(.*?)</td>', row)
quantity_match = re.search(r'<td>(\d+)</td>', row)
# print("item_name_match:", item_name_match)
# print("model_match:", model_match)
# print("quantity_match:", quantity_match)
# if not all([item_name_match, model_match, quantity_match]):
# raise ValueError("商品信息解析失败")
if item_name_match is not None:
print("项目名称:", item_name_match.group(1))
item_name = item_name_match.group(1)
else:
print("项目名称:", "无")
item_name = "无"
if model_match is not None:
print("规格型号:", model_match.group(1))
model = model_match.group(1)
else:
print("规格型号:", "无")
model = "无"
if quantity_match is not None:
print("数量:", quantity_match.group(1))
quantity = quantity_match.group(1)
else:
print("数量:", "无")
quantity = "无"
item_data = {
"name": item_name,
"model": model,
"quantity": quantity
}
invoice_data["items"].append(item_data)
# 解析第一行数据 # # 解析第一行数据
first_row = table_rows[1] # 跳过表头 # first_row = table_rows[0] # 跳过表头
item_name_match = re.search(r'<td>(.*?)</td>', first_row) # # print("first_row:", first_row)
model_match = re.search(r'<td>([^<]*)</td>', first_row) # item_name_match = re.search(r'<td>(.*?)</td>', first_row)
quantity_match = re.search(r'<td>(\d+)</td>', first_row) # model_match = re.search(r'<td[^>]*>.*?</td>\s*<td[^>]*>(.*?)</td>', first_row)
# quantity_match = re.search(r'<td>(\d+)</td>', first_row)
if not all([item_name_match, model_match, quantity_match]): # # print("item_name_match:", item_name_match)
raise ValueError("商品信息解析失败") # # print("model_match:", model_match)
# # print("quantity_match:", quantity_match)
return { # if not all([item_name_match, model_match, quantity_match]):
"发票号码": invoice_number, # raise ValueError("商品信息解析失败")
"销售方名称": seller_name,
"项目名称": item_name_match.group(1).strip(), # print("发票号码:", invoice_number)
"规格型号": model_match.group(1).strip(), # print("销售方名称:", seller_name)
"数量": quantity_match.group(1), # print("项目名称:", item_name_match.group(1))
"金额": amount # print("规格型号:", model_match.group(1))
} # print("数量:", quantity_match.group(1))
# print("金额:", amount)
return invoice_data
except Exception as e: except Exception as e:
print(f"解析发票信息时出错: {str(e)}") print(f"解析发票信息时出错: {str(e)}")
return None return None
def convert_pdf_to_markdown( def convert_pdf_to_markdown(
file_paths: list[str], file_paths: list[str],
client client
@@ -214,24 +269,27 @@ def get_pdf_files(directory):
if __name__ == "__main__": if __name__ == "__main__":
extract_invoice_info(markdown_text) # # test extract_invoice_info function
# info = extract_invoice_info(markdown_text)
# print("Extracted invoice info:", info)
# Example usage # Example usage
# client url can be the local host or the public url like `https://6986bdd23daef6f7eb.gradio.live/` # client url can be the local host or the public url like `https://6986bdd23daef6f7eb.gradio.live/`
# CLIENT_URL = "https://fceec28e477468b094.gradio.live/" CLIENT_URL = "https://fceec28e477468b094.gradio.live/"
# client = Client(CLIENT_URL, auth=("admin", "admin")) client = Client(CLIENT_URL, auth=("admin", "admin"))
# pdf_directory = "pdfs" pdf_directory = "pdfs"
# pdf_files = get_pdf_files(pdf_directory) pdf_files = get_pdf_files(pdf_directory)
# for pdf_file in pdf_files: print(pdf_files)
# print(f"Found PDF file: {pdf_file}") for pdf_file in pdf_files:
print(f"Found PDF file: {pdf_file}")
# # Single image conversion # Single image conversion
# markdown_content = convert_pdf_to_markdown( markdown_content = convert_pdf_to_markdown(
# [pdf_file],client [pdf_file],client
# ) )
# # print(markdown_content) # print(markdown_content)
# invoice_info = extract_invoice_info(markdown_content) invoice_info = extract_invoice_info(markdown_content)
# print(f"Extracted invoice info: {invoice_info}") print(f"Extracted invoice info: {invoice_info}")

81
mk.md Normal file
View File

@@ -0,0 +1,81 @@
Page 1 of 1
| | |
| --- | --- |
| [QR Code] | |
**发票信息**
电子发票(普通发票)
<signature>国家税务总局厦门市税务局</signature>
<signature>厦门市税务局</signature>
发票号码: 25947000000028179639
开票日期: 2025年05月20日
**购买方信息**
名称: 集美大学
统一社会信用代码/纳税人识别号: 12350000426600329N
**销售方信息**
名称: 厦门京东东和贸易有限公司
统一社会信用代码/纳税人识别号: 91350212MA34A9L25L
<table>
<thead>
<tr>
<th>项目名称</th>
<th>规格型号</th>
<th>单位</th>
<th>数量</th>
<th>单价</th>
<th>金额</th>
<th>税率/征收率</th>
<th>税额</th>
</tr>
</thead>
<tbody>
<tr>
<td>*计算机配套产品*金骏芦宝 24V1A电源适配器1000mA适用于按摩器甩脂机瘦身腰带LED台灯吸尘器扫地机器人加湿器充电器电源线</td>
<td>ROSE-240100C</td>
<td>个</td>
<td>2</td>
<td>23.01</td>
<td>46.02</td>
<td>13%</td>
<td>5.98</td>
</tr>
<tr>
<td>*计算机配套产品*金骏芦宝 24V1A电源适配器1000mA适用于按摩器甩脂机瘦身腰带LED台灯吸尘器扫地机器人加湿器充电器电源线</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td>-7.18</td>
<td>13%</td>
<td>-0.93</td>
</tr>
<tr>
<td>合计</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td>￥38.84</td>
<td></td>
<td>￥5.05</td>
</tr>
<tr>
<td>价税合计(大写)</td>
<td>肆拾叁圆捌角玖分</td>
<td></td>
<td></td>
<td></td>
<td>(小写) ￥43.89</td>
<td></td>
<td></td>
</tr>
</tbody>
</table>
备 注 订单号:316584139470
开票人: 王梅