feature: add external pdf extract operator by using mineru (#36)

* feature: add UnstructuredFormatter

* feature: add UnstructuredFormatter in db

* feature: add unstructured[docx]==0.18.15

* feature: support doc

* feature: add mineru

* feature: add external pdf extract operator by using mineru

* feature: mineru docker install bugfix

---------

Co-authored-by: Startalker <438747480@qq.com>

runtime/ops/formatter/external_pdf_formatter/__init__.py

@@ -0,0 +1,6 @@
+# -*- coding: utf-8 -*-
+
+from datamate.core.base_op import OPERATORS
+
+OPERATORS.register_module(module_name='ExternalPDFFormatter',
+                          module_path="ops.formatter.external_pdf_formatter.process")

runtime/ops/formatter/external_pdf_formatter/metadata.yml

@@ -0,0 +1,16 @@
+name: '外部PDF文本抽取'
+name_en: 'External PDF Text Extraction'
+description: '基于外部API，抽取PDF中的文本。'
+description_en: 'Extracts text from PDF files based on external APIs.'
+language: 'python'
+vendor: 'huawei'
+raw_id: 'ExternalPDFFormatter'
+version: '1.0.0'
+types:
+  - 'collect'
+modal: 'text'
+effect:
+  before: ''
+  after: ''
+inputs: 'text'
+outputs: 'text'

runtime/ops/formatter/external_pdf_formatter/process.py

@@ -0,0 +1,38 @@
+#!/user/bin/python
+# -*- coding: utf-8 -*-
+
+"""
+Description: 外部PDF文本抽取
+Create: 2025/10/29 17:24
+"""
+import json
+import os
+import time
+from loguru import logger
+from typing import Dict, Any
+
+from datamate.core.base_op import Mapper
+from datamate.common.utils.rest_client import http_request
+
+
+class ExternalPDFFormatter(Mapper):
+    """基于外部API，抽取PDF中的文本"""
+
+    def __init__(self, *args, **kwargs):
+        super(ExternalPDFFormatter, self).__init__(*args, **kwargs)
+        self.base_url = os.getenv("EXTERNAL_PDF_BASE_URL", "http://datamate-mineru:9001")
+        self.pdf_extract_url = f"{self.base_url}/api/pdf-extract"
+
+    def execute(self, sample: Dict[str, Any]) -> Dict[str, Any]:
+        start = time.time()
+        filename = sample[self.filename_key]
+        try:
+            data = {"source_path": sample[self.filepath_key], "export_path": sample[self.export_path_key]}
+            response = http_request(method="POST", url=self.pdf_extract_url, data=data)
+            sample[self.text_key] = json.loads(response.text).get("result")
+            logger.info(
+                f"fileName: {filename}, method: ExternalPDFFormatter costs {(time.time() - start):6f} s")
+        except UnicodeDecodeError as err:
+            logger.exception(f"fileName: {filename}, method: ExternalPDFFormatter causes decode error: {err}")
+            raise
+        return sample