{"id":51571,"date":"2025-08-11T00:29:15","date_gmt":"2025-08-10T16:29:15","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/51571.html"},"modified":"2025-08-11T00:29:15","modified_gmt":"2025-08-10T16:29:15","slug":"langchain-unstructured-%e5%9f%ba%e7%a1%80%e4%bd%bf%e7%94%a8%ef%bc%9apdf-%e4%b8%8e-markdown-%e5%a4%84%e7%90%86%e8%a7%a3%e6%9e%90","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/51571.html","title":{"rendered":"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790"},"content":{"rendered":"<\/p>\n<h4>\u6587\u7ae0\u76ee\u5f55<\/h4>\n<ul>\n<li>LangChain-Unstructured \u57fa\u7840\u4f7f\u7528&#xff1a;PDF \u4e0e Markdown \u5904\u7406\u89e3\u6790<\/li>\n<li>\n<ul>\n<li>\u4e00\u3001\u6838\u5fc3\u4f9d\u8d56\u4e0e\u5e93\u8bf4\u660e<\/li>\n<li>\u4e8c\u3001\u6838\u5fc3\u7c7b\u4e0e\u65b9\u6cd5\u8be6\u89e3<\/li>\n<li>\n<ul>\n<li>1.UnstructuredLoader<\/li>\n<li>\n<ul>\n<li>&#xff08;1&#xff09;\u7c7b\u539f\u578b\u4e0e\u6838\u5fc3\u529f\u80fd<\/li>\n<li>&#xff08;2&#xff09;\u521d\u59cb\u5316\u53c2\u6570\u8be6\u89e3<\/li>\n<li>&#xff08;3&#xff09;\u6838\u5fc3\u65b9\u6cd5\u8be6\u89e3<\/li>\n<li>\n<ul>\n<li>\u2460 &#096;load()&#096;<\/li>\n<li>\u2461 &#096;lazy_load()&#096;<\/li>\n<li>\u2462 &#096;aload()&#096;&#xff08;\u5f02\u6b65&#xff09;<\/li>\n<li>\u2463 &#096;alazy_load()&#096;&#xff08;\u5f02\u6b65&#xff09;<\/li>\n<li>\u2464 &#096;load_and_split()&#096;<\/li>\n<\/ul>\n<\/li>\n<li>&#xff08;3&#xff09;\u4f7f\u7528\u793a\u4f8b<\/li>\n<li>\n<ul>\n<li>\u793a\u4f8b 1&#xff1a;\u89e3\u6790\u672c\u5730 PDF \u6587\u4ef6&#xff08;\u672c\u5730\u5f15\u64ce&#xff0c;\u5e26\u540e\u5904\u7406&#xff09;<\/li>\n<li>\u793a\u4f8b 2&#xff1a;\u901a\u8fc7 URL \u52a0\u8f7d\u7f51\u9875\u5185\u5bb9<\/li>\n<li>\u793a\u4f8b 3&#xff1a;\u4f7f\u7528 API \u89e3\u6790\u626b\u63cf\u578b PDF&#xff08;OCR \u8bc6\u522b&#xff09;<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<li>2. UnstructuredFileLoader<\/li>\n<li>3. &#096;UnstructuredPDFLoader&#096; \u4e0e &#096;UnstructuredMarkdownLoader&#096;<\/li>\n<li>\n<ul>\n<li>&#096;UnstructuredPDFLoader&#096;&#xff08;PDF \u4e13\u7528&#xff09;<\/li>\n<li>&#096;UnstructuredMarkdownLoader&#096;&#xff08;Markdown \u4e13\u7528&#xff09;<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<li>\u4e09\u3001PDF \u89e3\u6790\u5b9e\u6218\u793a\u4f8b<\/li>\n<li>\n<ul>\n<li>\u793a\u4f8b 1&#xff1a;\u9ad8\u7cbe\u5ea6\u89e3\u6790\u542b\u8868\u683c\u7684 PDF<\/li>\n<li>\u793a\u4f8b 2&#xff1a;\u5904\u7406\u626b\u63cf\u578b PDF&#xff08;OCR \u8bc6\u522b&#xff09;<\/li>\n<\/ul>\n<\/li>\n<li>\u56db\u3001Markdown \u89e3\u6790\u5b9e\u6218\u793a\u4f8b<\/li>\n<li>\n<ul>\n<li>\u793a\u4f8b 3&#xff1a;\u89e3\u6790\u542b\u590d\u6742\u7ed3\u6784\u7684 Markdown<\/li>\n<\/ul>\n<\/li>\n<li>\u4e94\u3001\u8fdb\u9636\u6280\u5de7&#xff1a;\u540e\u5904\u7406\u51fd\u6570\u4e0e\u5143\u7d20\u8fc7\u6ee4<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h2>LangChain-Unstructured \u57fa\u7840\u4f7f\u7528&#xff1a;PDF \u4e0e Markdown \u5904\u7406\u89e3\u6790<\/h2>\n<p>LangChain-Unstructured \u4f5c\u4e3a LangChain \u751f\u6001\u4e2d\u5904\u7406\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u6838\u5fc3\u5de5\u5177&#xff0c;\u63d0\u4f9b\u4e86\u4e30\u5bcc\u7684\u7c7b\u548c\u65b9\u6cd5\u6765\u89e3\u6790 PDF \u548c Markdown \u7b49\u683c\u5f0f\u3002\u672c\u6587\u5c06\u7cfb\u7edf\u8bb2\u89e3\u5176\u6838\u5fc3\u5e93\u3001\u5173\u952e\u7c7b\u3001\u65b9\u6cd5\u53c2\u6570\u53ca\u5b9e\u6218\u793a\u4f8b&#xff0c;\u5e2e\u52a9\u4f60\u5168\u9762\u638c\u63e1\u6587\u6863\u89e3\u6790\u6d41\u7a0b\u3002<\/p>\n<h3>\u4e00\u3001\u6838\u5fc3\u4f9d\u8d56\u4e0e\u5e93\u8bf4\u660e<\/h3>\n<p>\u5728\u4f7f\u7528 LangChain-Unstructured \u5904\u7406 PDF \u548c Markdown \u65f6&#xff0c;\u9700\u4e86\u89e3\u4ee5\u4e0b\u6838\u5fc3\u4f9d\u8d56\u53ca\u5176\u4f5c\u7528&#xff1a;<\/p>\n<table>\n<tr>\u5e93\u540d\u79f0\u529f\u80fd\u8bf4\u660e\u5fc5\u5907\u6027<\/tr>\n<tbody>\n<tr>\n<td>langchain-community<\/td>\n<td>LangChain \u793e\u533a\u7248&#xff0c;\u5305\u542b UnstructuredFileLoader \u7b49\u6838\u5fc3\u89e3\u6790\u7c7b<\/td>\n<td>\u5fc5\u9700<\/td>\n<\/tr>\n<tr>\n<td>unstructured<\/td>\n<td>\u5e95\u5c42\u6587\u6863\u89e3\u6790\u5f15\u64ce&#xff0c;\u652f\u6301 20&#043; \u6587\u6863\u683c\u5f0f\u7684\u7ed3\u6784\u5316\u8f6c\u6362<\/td>\n<td>\u5fc5\u9700<\/td>\n<\/tr>\n<tr>\n<td>unstructured[pdf]<\/td>\n<td>unstructured \u7684 PDF \u89e3\u6790\u6269\u5c55&#xff0c;\u5305\u542b pdfminer\u3001pypdf \u7b49\u89e3\u6790\u5668<\/td>\n<td>\u5904\u7406 PDF \u65f6\u5fc5\u9700<\/td>\n<\/tr>\n<tr>\n<td>unstructured[md]<\/td>\n<td>Markdown \u89e3\u6790\u6269\u5c55&#xff0c;\u652f\u6301\u6807\u9898\u3001\u5217\u8868\u3001\u4ee3\u7801\u5757\u7b49\u5143\u7d20\u8bc6\u522b<\/td>\n<td>\u5904\u7406 Markdown \u65f6\u5fc5\u9700<\/td>\n<\/tr>\n<tr>\n<td>pdf2image<\/td>\n<td>\u5c06 PDF \u9875\u9762\u8f6c\u6362\u4e3a\u56fe\u50cf&#xff08;\u7528\u4e8e\u590d\u6742\u6392\u7248\u6216\u626b\u63cf\u4ef6\u89e3\u6790&#xff09;<\/td>\n<td>\u53ef\u9009&#xff08;\u9ad8\u7cbe\u5ea6 PDF \u89e3\u6790\u65f6\u9700\u7528&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>python-multipart<\/td>\n<td>\u652f\u6301\u591a\u90e8\u5206\u6587\u6863\u89e3\u6790&#xff08;\u5982\u5305\u542b\u9644\u4ef6\u7684 PDF&#xff09;<\/td>\n<td>\u53ef\u9009<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u5b89\u88c5\u547d\u4ee4&#xff1a;<\/p>\n<p><span class=\"token comment\"># \u57fa\u7840\u5b89\u88c5&#xff08;\u542b PDF \u548c Markdown \u652f\u6301&#xff09;<\/span><br \/>\npip <span class=\"token function\">install<\/span> langchain-community unstructured<span class=\"token punctuation\">[<\/span>pdf,md<span class=\"token punctuation\">]<\/span> pdf2image<\/p>\n<h3>\u4e8c\u3001\u6838\u5fc3\u7c7b\u4e0e\u65b9\u6cd5\u8be6\u89e3<\/h3>\n<p>LangChain-Unstructured \u4e2d\u6700\u6838\u5fc3\u7684\u7c7b\u662f UnstructuredFileLoader&#xff08;\u901a\u7528\u6587\u4ef6\u52a0\u8f7d\u5668&#xff09;\u548c\u884d\u751f\u7684 UnstructuredPDFLoader\u3001UnstructuredMarkdownLoader&#xff08;\u4e13\u7528\u52a0\u8f7d\u5668&#xff09;\u3002\u4ee5\u4e0b\u8be6\u7ec6\u8bb2\u89e3\u5176\u7528\u6cd5\u3002<\/p>\n<h4>1.UnstructuredLoader<\/h4>\n<p>UnstructuredLoader \u662f langchain-unstructured \u5e93\u4e2d\u7528\u4e8e\u52a0\u8f7d\u548c\u89e3\u6790\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u6838\u5fc3\u7c7b&#xff0c;\u5b83\u63d0\u4f9b\u4e86\u66f4\u7075\u6d3b\u7684\u6587\u6863\u5904\u7406\u80fd\u529b&#xff0c;\u652f\u6301\u672c\u5730\u6587\u4ef6\u3001\u5b57\u8282\u6d41\u3001\u7f51\u9875 URL \u7b49\u591a\u79cd\u8f93\u5165\u65b9\u5f0f&#xff0c;\u5e76\u53ef\u901a\u8fc7 API \u6216\u672c\u5730\u5f15\u64ce\u89e3\u6790\u6587\u6863\u3002\u662f\u5904\u7406\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u5f3a\u5927\u5de5\u5177&#xff0c;\u5176\u6838\u5fc3\u4f18\u52bf\u5728\u4e8e&#xff1a;<\/p>\n<li>\u591a\u8f93\u5165\u652f\u6301&#xff1a;\u672c\u5730\u6587\u4ef6\u3001\u5b57\u8282\u6d41\u3001\u7f51\u9875 URL \u5168\u8986\u76d6\u3002<\/li>\n<li>\u7075\u6d3b\u89e3\u6790\u65b9\u5f0f&#xff1a;\u672c\u5730\u89e3\u6790&#xff08;\u79bb\u7ebf&#xff09;\u6216 API \u89e3\u6790&#xff08;\u9ad8\u7cbe\u5ea6&#xff0c;\u652f\u6301\u590d\u6742\u573a\u666f&#xff09;\u3002<\/li>\n<li>\u53ef\u6269\u5c55\u5904\u7406&#xff1a;\u901a\u8fc7 post_processors \u81ea\u5b9a\u4e49\u6587\u672c\u6e05\u6d17\u903b\u8f91\u3002<\/li>\n<li>\u9002\u914d\u591a\u573a\u666f&#xff1a;\u540c\u6b65\/\u5f02\u6b65\u3001\u60f0\u6027\/\u975e\u60f0\u6027\u52a0\u8f7d&#xff0c;\u6ee1\u8db3\u4e0d\u540c\u6027\u80fd\u9700\u6c42\u3002<\/li>\n<p>\u4f7f\u7528\u65f6\u9700\u6839\u636e\u6587\u6863\u7c7b\u578b&#xff08;\u6587\u672c\/\u626b\u63cf PDF\u3001\u7f51\u9875\u7b49&#xff09;\u548c\u573a\u666f&#xff08;\u79bb\u7ebf\/\u5728\u7ebf\u3001\u540c\u6b65\/\u5f02\u6b65&#xff09;\u9009\u62e9\u5408\u9002\u7684\u53c2\u6570\u914d\u7f6e&#xff0c;\u89e3\u6790\u540e\u7684 Document \u5bf9\u8c61\u53ef\u76f4\u63a5\u7528\u4e8e\u6587\u672c\u5206\u5272\u3001\u5411\u91cf\u5b58\u50a8\u7b49\u540e\u7eed\u6b65\u9aa4&#xff0c;\u4e3a LLM \u5e94\u7528&#xff08;\u5982\u95ee\u7b54\u3001\u6458\u8981&#xff09;\u63d0\u4f9b\u7ed3\u6784\u5316\u8f93\u5165\u3002\u4ee5\u4e0b\u662f\u5176\u8be6\u7ec6\u4f7f\u7528\u6307\u5357&#xff1a;<\/p>\n<h5>&#xff08;1&#xff09;\u7c7b\u539f\u578b\u4e0e\u6838\u5fc3\u529f\u80fd<\/h5>\n<p>\u7c7b\u5b9a\u4e49<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">langchain_unstructured<\/span><span class=\"token punctuation\">.<\/span>document_loaders<span class=\"token punctuation\">.<\/span>UnstructuredLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">|<\/span> Path <span class=\"token operator\">|<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">|<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>Path<span class=\"token punctuation\">]<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token operator\">*<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token builtin\">file<\/span><span class=\"token punctuation\">:<\/span> IO<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">bytes<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">|<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>IO<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">bytes<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    partition_via_api<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">bool<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">False<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    post_processors<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>Callable<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    api_key<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    client<span class=\"token punctuation\">:<\/span> UnstructuredClient <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    url<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    web_url<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token operator\">**<\/span>kwargs<span class=\"token punctuation\">:<\/span> Any<span class=\"token punctuation\">,<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">[<\/span>source<span class=\"token punctuation\">]<\/span><\/p>\n<p>\u6838\u5fc3\u529f\u80fd<\/p>\n<ul>\n<li>\u652f\u6301\u591a\u79cd\u8f93\u5165\u6e90&#xff1a;\u672c\u5730\u6587\u4ef6\u8def\u5f84\u3001\u5b57\u8282\u6d41&#xff08;IO[bytes]&#xff09;\u3001\u7f51\u9875 URL&#xff08;web_url&#xff09;\u3002<\/li>\n<li>\u53ef\u9009\u62e9\u672c\u5730\u89e3\u6790\u6216\u901a\u8fc7 Unstructured API \u89e3\u6790&#xff08;partition_via_api \u63a7\u5236&#xff09;\u3002<\/li>\n<li>\u63d0\u4f9b\u6587\u6863\u540e\u5904\u7406\u80fd\u529b&#xff08;post_processors&#xff09;&#xff0c;\u5982\u6587\u672c\u6e05\u6d17\u3001\u683c\u5f0f\u8f6c\u6362\u7b49\u3002<\/li>\n<li>\u8f93\u51fa\u7ed3\u6784\u5316\u7684 Document \u5bf9\u8c61&#xff0c;\u5305\u542b\u6587\u672c\u5185\u5bb9&#xff08;page_content&#xff09;\u548c\u5143\u6570\u636e&#xff08;metadata&#xff09;\u3002<\/li>\n<\/ul>\n<h5>&#xff08;2&#xff09;\u521d\u59cb\u5316\u53c2\u6570\u8be6\u89e3<\/h5>\n<table>\n<tr>\u53c2\u6570\u540d\u7c7b\u578b\u529f\u80fd\u8bf4\u660e\u793a\u4f8b\u503c<\/tr>\n<tbody>\n<tr>\n<td>file_path<\/td>\n<td>str \/ Path \/ \u5217\u8868&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>\u672c\u5730\u6587\u4ef6\u8def\u5f84&#xff0c;\u652f\u6301\u5355\u4e2a\u6216\u591a\u4e2a\u6587\u4ef6&#xff08;\u5217\u8868\u5f62\u5f0f&#xff09;\u3002\u4e0e file\u3001web_url \u4e92\u65a5&#xff0c;\u4e09\u9009\u4e00\u3002<\/td>\n<td>&#034;.\/report.pdf&#034; \u6216 [&#034;.\/a.pdf&#034;, &#034;.\/b.md&#034;]<\/td>\n<\/tr>\n<tr>\n<td>file<\/td>\n<td>IO[bytes] \/ \u5217\u8868&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>\u5b57\u8282\u6d41\u5bf9\u8c61&#xff08;\u5982\u6253\u5f00\u7684\u6587\u4ef6\u53e5\u67c4&#xff09;&#xff0c;\u652f\u6301\u5355\u4e2a\u6216\u591a\u4e2a\u3002\u7528\u4e8e\u76f4\u63a5\u5904\u7406\u5185\u5b58\u4e2d\u7684\u6587\u4ef6\u6570\u636e\u3002<\/td>\n<td>open(&#034;doc.pdf&#034;, &#034;rb&#034;)<\/td>\n<\/tr>\n<tr>\n<td>web_url<\/td>\n<td>str&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>\u7f51\u9875 URL&#xff0c;\u7528\u4e8e\u52a0\u8f7d\u5e76\u89e3\u6790\u7f51\u9875\u5185\u5bb9&#xff08;\u81ea\u52a8\u63d0\u53d6\u6587\u672c\u548c\u7ed3\u6784&#xff09;\u3002<\/td>\n<td>&#034;https:\/\/example.com\/article&#034;<\/td>\n<\/tr>\n<tr>\n<td>partition_via_api<\/td>\n<td>bool&#xff08;\u9ed8\u8ba4 False&#xff09;<\/td>\n<td>\u662f\u5426\u901a\u8fc7 Unstructured API \u89e3\u6790\u6587\u6863&#xff08;\u9700\u914d\u7f6e api_key&#xff09;\u3002False \u5219\u4f7f\u7528\u672c\u5730\u5f15\u64ce\u89e3\u6790\u3002<\/td>\n<td>True&#xff08;\u4f7f\u7528 API&#xff09;\u3001False&#xff08;\u672c\u5730\u89e3\u6790&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>post_processors<\/td>\n<td>\u5217\u8868[Callable[[str], str]]&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>\u6587\u672c\u540e\u5904\u7406\u51fd\u6570\u5217\u8868&#xff0c;\u6bcf\u4e2a\u51fd\u6570\u63a5\u6536\u5b57\u7b26\u4e32\u5e76\u8fd4\u56de\u5904\u7406\u540e\u7684\u5b57\u7b26\u4e32&#xff08;\u5982\u53bb\u9664\u591a\u4f59\u7a7a\u683c\u3001\u66ff\u6362\u7279\u6b8a\u5b57\u7b26&#xff09;\u3002<\/td>\n<td>[lambda x: x.strip(), lambda x: x.replace(&#034;\\\\n&#034;, &#034; &#034;)]<\/td>\n<\/tr>\n<tr>\n<td>api_key<\/td>\n<td>str&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>Unstructured API \u7684\u5bc6\u94a5&#xff0c;\u5f53 partition_via_api&#061;True \u65f6\u5fc5\u586b\u3002\u53ef\u901a\u8fc7\u73af\u5883\u53d8\u91cf UNSTRUCTURED_API_KEY \u914d\u7f6e\u3002<\/td>\n<td>&#034;your_api_key_here&#034;<\/td>\n<\/tr>\n<tr>\n<td>client<\/td>\n<td>UnstructuredClient&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>\u81ea\u5b9a\u4e49\u7684 UnstructuredClient \u5b9e\u4f8b&#xff0c;\u7528\u4e8e\u9ad8\u7ea7 API \u914d\u7f6e&#xff08;\u5982\u8d85\u65f6\u8bbe\u7f6e\u3001\u4ee3\u7406\u7b49&#xff09;\u3002<\/td>\n<td>UnstructuredClient(api_key&#061;&#034;key&#034;, timeout&#061;30)<\/td>\n<\/tr>\n<tr>\n<td>url<\/td>\n<td>str&#xff08;\u53ef\u9009&#xff09;<\/td>\n<td>Unstructured API \u7684\u670d\u52a1\u5730\u5740&#xff0c;\u9ed8\u8ba4\u4f7f\u7528\u5b98\u65b9\u5730\u5740\u3002\u4ec5\u5f53 partition_via_api&#061;True \u65f6\u6709\u6548\u3002<\/td>\n<td>&#034;https:\/\/custom-unstructured-api.com&#034;<\/td>\n<\/tr>\n<tr>\n<td>**kwargs<\/td>\n<td>\u4efb\u610f\u952e\u503c\u5bf9<\/td>\n<td>\u4f20\u9012\u7ed9\u5e95\u5c42\u89e3\u6790\u5f15\u64ce\u7684\u53c2\u6570&#xff0c;\u5982 strategy&#061;&#034;hi_res&#034;&#xff08;\u9ad8\u7cbe\u5ea6\u89e3\u6790&#xff09;\u3001ocr_languages&#061;&#034;chi_sim&#034;&#xff08;OCR \u8bed\u8a00&#xff09;\u7b49\u3002<\/td>\n<td>strategy&#061;&#034;hi_res&#034;, pdf_infer_table_structure&#061;True<\/td>\n<\/tr>\n<tr>\n<td>coordinates<\/td>\n<td>bool\u7c7b\u578b<\/td>\n<td>\u914d\u7f6e\u662f\u5426\u5728\u5143\u6570\u636e\u4e2d\u4fdd\u7559\u5143\u7d20\u7684\u5750\u6807\u4fe1\u606f&#xff1a; \u2460\u542f\u7528\u540e&#xff0c;\u89e3\u6790\u7ed3\u679c\u7684 metadata \u4f1a\u5305\u542b\u6bcf\u4e2a\u5143\u7d20&#xff08;\u5982\u6807\u9898\u3001\u6bb5\u843d&#xff09;\u5728\u6587\u6863\u4e2d\u7684\u4f4d\u7f6e\u5750\u6807&#xff08;\u5982\u50cf\u7d20\u5750\u6807\u3001\u9875\u9762\u8303\u56f4&#xff09;&#xff1b; \u2461\u9002\u7528\u4e8e\u9700\u8981\u5b9a\u4f4d\u6587\u672c\u5728\u539f\u59cb\u6587\u6863\u4e2d\u7269\u7406\u4f4d\u7f6e\u7684\u573a\u666f&#xff08;\u5982\u9ad8\u4eae\u663e\u793a\u3001\u6587\u6863\u5bf9\u6bd4&#xff0c;\u50cf\u6709\u8868\u683c\u63d0\u53d6\u89e3\u6790\u9700\u6c42\u5219\u907f\u5f00&#xff09;\u3002<\/td>\n<td>True<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h5>&#xff08;3&#xff09;\u6838\u5fc3\u65b9\u6cd5\u8be6\u89e3<\/h5>\n<h6>\u2460 load()<\/h6>\n<p>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> \u2192 <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">]<\/span><\/p>\n<ul>\n<li>\u529f\u80fd&#xff1a;\u540c\u6b65\u52a0\u8f7d\u5e76\u89e3\u6790\u6587\u6863&#xff0c;\u8fd4\u56de Document \u5bf9\u8c61\u5217\u8868\u3002<\/li>\n<li>\u9002\u7528\u573a\u666f&#xff1a;\u4e2d\u5c0f\u578b\u6587\u4ef6&#xff0c;\u9700\u4e00\u6b21\u6027\u83b7\u53d6\u6240\u6709\u7ed3\u679c\u3002<\/li>\n<\/ul>\n<h6>\u2461 lazy_load()<\/h6>\n<p>lazy_load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> \u2192 Iterator<span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">]<\/span><\/p>\n<ul>\n<li>\u529f\u80fd&#xff1a;\u540c\u6b65\u60f0\u6027\u52a0\u8f7d&#xff0c;\u8fd4\u56de\u8fed\u4ee3\u5668&#xff0c;\u9010\u6b21\u751f\u6210 Document \u5bf9\u8c61\u3002<\/li>\n<li>\u9002\u7528\u573a\u666f&#xff1a;\u5927\u578b\u6587\u4ef6\u6216\u6279\u91cf\u5904\u7406&#xff0c;\u51cf\u5c11\u5185\u5b58\u5360\u7528\u3002<\/li>\n<\/ul>\n<h6>\u2462 aload()&#xff08;\u5f02\u6b65&#xff09;<\/h6>\n<p><span class=\"token keyword\">async<\/span> aload<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> \u2192 <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">]<\/span><\/p>\n<ul>\n<li>\u529f\u80fd&#xff1a;\u5f02\u6b65\u52a0\u8f7d\u5e76\u89e3\u6790\u6587\u6863&#xff0c;\u8fd4\u56de Document \u5217\u8868\u3002\u9002\u7528\u4e8e\u5f02\u6b65\u4ee3\u7801\u73af\u5883&#xff08;\u5982 asyncio&#xff09;\u3002<\/li>\n<\/ul>\n<h6>\u2463 alazy_load()&#xff08;\u5f02\u6b65&#xff09;<\/h6>\n<p><span class=\"token keyword\">async<\/span> alazy_load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> \u2192 AsyncIterator<span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">]<\/span><\/p>\n<ul>\n<li>\u529f\u80fd&#xff1a;\u5f02\u6b65\u60f0\u6027\u52a0\u8f7d&#xff0c;\u8fd4\u56de\u5f02\u6b65\u8fed\u4ee3\u5668&#xff0c;\u901a\u8fc7 async for \u9010\u6b21\u83b7\u53d6 Document\u3002<\/li>\n<\/ul>\n<h6>\u2464 load_and_split()<\/h6>\n<p>load_and_split<span class=\"token punctuation\">(<\/span>text_splitter<span class=\"token punctuation\">:<\/span> TextSplitter <span class=\"token operator\">|<\/span> <span class=\"token boolean\">None<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">)<\/span> \u2192 <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">]<\/span><\/p>\n<ul>\n<li>\u529f\u80fd&#xff1a;\u52a0\u8f7d\u6587\u6863\u540e\u6309\u6307\u5b9a\u7684 TextSplitter \u62c6\u5206\u5185\u5bb9&#xff0c;\u8fd4\u56de\u62c6\u5206\u540e\u7684 Document \u7247\u6bb5&#xff08;\u7528\u4e8e RAG \u7b49\u573a\u666f&#xff09;\u3002<\/li>\n<li>\u53c2\u6570&#xff1a;text_splitter \u4e3a\u6587\u672c\u5206\u5272\u5668&#xff0c;\u9ed8\u8ba4\u4f7f\u7528 RecursiveCharacterTextSplitter\u3002<\/li>\n<\/ul>\n<h5>&#xff08;3&#xff09;\u4f7f\u7528\u793a\u4f8b<\/h5>\n<h6>\u793a\u4f8b 1&#xff1a;\u89e3\u6790\u672c\u5730 PDF \u6587\u4ef6&#xff08;\u672c\u5730\u5f15\u64ce&#xff0c;\u5e26\u540e\u5904\u7406&#xff09;<\/h6>\n<p><span class=\"token keyword\">from<\/span> langchain_unstructured<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredLoader<\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u540e\u5904\u7406\u51fd\u6570&#xff1a;\u53bb\u9664\u7a7a\u767d\u5e76\u66ff\u6362\u6362\u884c<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">clean_text<\/span><span class=\"token punctuation\">(<\/span>text<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> text<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>replace<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;\\\\n&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034; &#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316\u52a0\u8f7d\u5668<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/example.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    partition_via_api<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">False<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u672c\u5730\u89e3\u6790<\/span><br \/>\n    post_processors<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>clean_text<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5e94\u7528\u6587\u672c\u6e05\u6d17<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;hi_res&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u9ad8\u7cbe\u5ea6\u89e3\u6790<\/span><br \/>\n    pdf_infer_table_structure<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span>  <span class=\"token comment\"># \u89e3\u6790\u8868\u683c\u7ed3\u6784<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u52a0\u8f7d\u6587\u6863<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8f93\u51fa\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u89e3\u6790\u5f97\u5230 <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\"> \u4e2a\u6587\u6863\u5143\u7d20&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u7b2c\u4e00\u4e2a\u5143\u7d20\u7c7b\u578b&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;category&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5185\u5bb9\u9884\u89c8&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">200]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>\u89e3\u6790\u5f97\u5230 15 \u4e2a\u6587\u6863\u5143\u7d20<br \/>\n\u7b2c\u4e00\u4e2a\u5143\u7d20\u7c7b\u578b&#xff1a;Title<br \/>\n\u5185\u5bb9\u9884\u89c8&#xff1a;2024 \u5e74\u6280\u672f\u8d8b\u52bf\u62a5\u544a \u672c\u62a5\u544a\u603b\u7ed3\u4e86 2024 \u5e74\u4eba\u5de5\u667a\u80fd\u3001\u533a\u5757\u94fe\u7b49\u9886\u57df\u7684\u6838\u5fc3\u8d8b\u52bf&#xff0c;\u5305\u542b\u5e02\u573a\u6570\u636e\u548c\u6848\u4f8b\u5206\u6790\u3002\u4ee5\u4e0b\u4e3a\u4e3b\u8981\u5185\u5bb9&#xff1a;&#8230;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u672c\u5730\u89e3\u6790\u6a21\u5f0f\u65e0\u9700 API \u5bc6\u94a5&#xff0c;\u9002\u5408\u79bb\u7ebf\u573a\u666f\u3002<\/li>\n<li>post_processors \u6210\u529f\u53bb\u9664\u4e86\u6587\u672c\u4e2d\u7684\u591a\u4f59\u7a7a\u767d\u548c\u6362\u884c&#xff0c;\u4f7f\u5185\u5bb9\u66f4\u6574\u6d01\u3002<\/li>\n<li>strategy&#061;&#034;hi_res&#034; \u786e\u4fdd\u4e86\u590d\u6742\u6392\u7248&#xff08;\u5982\u8868\u683c&#xff09;\u7684\u51c6\u786e\u89e3\u6790&#xff0c;\u5143\u6570\u636e category \u6807\u8bb0\u4e86\u5143\u7d20\u7c7b\u578b&#xff08;\u6807\u9898\u3001\u6bb5\u843d\u7b49&#xff09;\u3002<\/li>\n<\/ul>\n<h6>\u793a\u4f8b 2&#xff1a;\u901a\u8fc7 URL \u52a0\u8f7d\u7f51\u9875\u5185\u5bb9<\/h6>\n<p><span class=\"token keyword\">from<\/span> langchain_unstructured<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredLoader<\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316\u52a0\u8f7d\u5668&#xff08;\u89e3\u6790\u7f51\u9875&#xff09;<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredLoader<span class=\"token punctuation\">(<\/span><br \/>\n    web_url<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;https:\/\/example.com\/ai-trends&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u7f51\u9875 URL<\/span><br \/>\n    partition_via_api<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">False<\/span>  <span class=\"token comment\"># \u672c\u5730\u89e3\u6790\u7f51\u9875<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u60f0\u6027\u52a0\u8f7d&#xff08;\u9010\u5143\u7d20\u83b7\u53d6&#xff09;<\/span><br \/>\n<span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> loader<span class=\"token punctuation\">.<\/span>lazy_load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5143\u7d20\u7c7b\u578b&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;category&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5185\u5bb9&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">150]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">\\\\n&#8212;&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>\u5143\u7d20\u7c7b\u578b&#xff1a;Title<br \/>\n\u5185\u5bb9&#xff1a;2024 \u5e74\u4eba\u5de5\u667a\u80fd\u53d1\u5c55\u8d8b\u52bf<br \/>\n&#8212;<br \/>\n\u5143\u7d20\u7c7b\u578b&#xff1a;NarrativeText<br \/>\n\u5185\u5bb9&#xff1a;\u968f\u7740\u5927\u6a21\u578b\u6280\u672f\u7684\u6210\u719f&#xff0c;2024 \u5e74\u4eba\u5de5\u667a\u80fd\u5728\u5404\u884c\u5404\u4e1a\u7684\u5e94\u7528\u8fdb\u4e00\u6b65\u6df1\u5316&#xff0c;\u5c24\u5176\u662f\u5728\u533b\u7597\u3001\u6559\u80b2\u548c\u5236\u9020\u4e1a&#8230;<br \/>\n&#8212;<br \/>\n\u5143\u7d20\u7c7b\u578b&#xff1a;ListItem<br \/>\n\u5185\u5bb9&#xff1a;- \u591a\u6a21\u6001\u6a21\u578b\u6210\u4e3a\u4e3b\u6d41&#xff0c;\u8de8\u5a92\u4f53\u7406\u89e3\u80fd\u529b\u663e\u8457\u63d0\u5347<br \/>\n&#8212;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>web_url \u53c2\u6570\u76f4\u63a5\u52a0\u8f7d\u7f51\u9875\u5e76\u63d0\u53d6\u7ed3\u6784\u5316\u5143\u7d20&#xff08;\u6807\u9898\u3001\u6bb5\u843d\u3001\u5217\u8868\u7b49&#xff09;&#xff0c;\u5143\u6570\u636e category \u6807\u8bb0\u4e86\u5143\u7d20\u7c7b\u578b\u3002<\/li>\n<li>lazy_load() \u9010\u6b21\u8fd4\u56de\u5143\u7d20&#xff0c;\u9002\u5408\u5904\u7406\u957f\u7f51\u9875&#xff0c;\u907f\u514d\u4e00\u6b21\u6027\u52a0\u8f7d\u5360\u7528\u8fc7\u591a\u5185\u5b58\u3002<\/li>\n<\/ul>\n<h6>\u793a\u4f8b 3&#xff1a;\u4f7f\u7528 API \u89e3\u6790\u626b\u63cf\u578b PDF&#xff08;OCR \u8bc6\u522b&#xff09;<\/h6>\n<p><span class=\"token keyword\">from<\/span> langchain_unstructured<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredLoader<br \/>\n<span class=\"token keyword\">import<\/span> os<\/p>\n<p><span class=\"token comment\"># \u914d\u7f6e API \u5bc6\u94a5&#xff08;\u4e5f\u53ef\u901a\u8fc7\u73af\u5883\u53d8\u91cf\u8bbe\u7f6e&#xff09;<\/span><br \/>\nos<span class=\"token punctuation\">.<\/span>environ<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;UNSTRUCTURED_API_KEY&#034;<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;your_api_key&#034;<\/span><\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316\u52a0\u8f7d\u5668&#xff08;API \u89e3\u6790 &#043; OCR&#xff09;<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/scan_pdf.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u626b\u63cf\u578b PDF&#xff08;\u56fe\u7247\u5185\u5bb9&#xff09;<\/span><br \/>\n    partition_via_api<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u4f7f\u7528 API \u89e3\u6790<\/span><br \/>\n    ocr_languages<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;eng&#043;chi_sim&#034;<\/span>  <span class=\"token comment\"># \u652f\u6301\u4e2d\u82f1\u53cc\u8bed OCR<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5f02\u6b65\u52a0\u8f7d&#xff08;\u9002\u5408 API \u8c03\u7528&#xff0c;\u907f\u514d\u963b\u585e&#xff09;<\/span><br \/>\n<span class=\"token keyword\">import<\/span> asyncio<br \/>\n<span class=\"token keyword\">async<\/span> <span class=\"token keyword\">def<\/span> <span class=\"token function\">main<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    docs <span class=\"token operator\">&#061;<\/span> <span class=\"token keyword\">await<\/span> loader<span class=\"token punctuation\">.<\/span>aload<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;OCR \u89e3\u6790\u7ed3\u679c&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>docs<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">300]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>asyncio<span class=\"token punctuation\">.<\/span>run<span class=\"token punctuation\">(<\/span>main<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>OCR \u89e3\u6790\u7ed3\u679c&#xff1a;\u8fd9\u662f\u4e00\u4efd\u626b\u63cf\u7684\u62a5\u544a&#xff0c;\u5305\u542b\u4e2d\u82f1\u6587\u5185\u5bb9\u30022024 \u5e74\u5168\u7403 AI \u5e02\u573a\u89c4\u6a21\u9884\u8ba1\u8fbe\u5230 2 \u4e07\u4ebf\u7f8e\u5143&#xff0c;\u540c\u6bd4\u589e\u957f 40%\u3002Key points: 1. \u751f\u6210\u5f0f AI \u5546\u4e1a\u5316\u52a0\u901f&#xff1b;2. \u8fb9\u7f18 AI \u8bbe\u5907\u666e\u53ca&#8230;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>partition_via_api&#061;True \u542f\u7528 Unstructured API&#xff0c;\u7ed3\u5408 OCR \u6280\u672f\u6210\u529f\u8bc6\u522b\u626b\u63cf\u578b PDF \u4e2d\u7684\u6587\u672c\u3002<\/li>\n<li>ocr_languages \u6307\u5b9a\u8bc6\u522b\u8bed\u8a00&#xff0c;\u786e\u4fdd\u4e2d\u82f1\u53cc\u8bed\u5185\u5bb9\u51c6\u786e\u63d0\u53d6\u3002<\/li>\n<li>\u5f02\u6b65\u65b9\u6cd5 aload() \u9002\u5408 API \u8c03\u7528\u573a\u666f&#xff0c;\u63d0\u9ad8\u4ee3\u7801\u6548\u7387\u3002<\/li>\n<\/ul>\n<h4>2. UnstructuredFileLoader<\/h4>\n<p>UnstructuredFileLoader \u662f LangChain-Unstructured \u4e2d\u8bbe\u8ba1\u6700\u901a\u7528\u7684\u672c\u5730\u6587\u4ef6\u52a0\u8f7d\u5668&#xff0c;\u5176\u6838\u5fc3\u5b9a\u4f4d\u662f\u4f5c\u4e3a\u5404\u7c7b\u975e\u7ed3\u6784\u5316\u6587\u6863&#xff08;\u5305\u62ec\u4f46\u4e0d\u9650\u4e8e PDF\u3001Markdown\u3001Word\u3001HTML\u3001TXT \u7b49\u672c\u5730\u6587\u4ef6&#xff09;\u7684\u7edf\u4e00\u89e3\u6790\u5165\u53e3\u3002\u5b83\u7684\u5e95\u5c42\u901a\u8fc7\u8c03\u7528 unstructured \u5e93\u7684\u591a\u683c\u5f0f\u89e3\u6790\u80fd\u529b&#xff0c;\u80fd\u591f\u6839\u636e\u6587\u4ef6\u7684\u540e\u7f00\u540d\u81ea\u52a8\u5339\u914d\u5bf9\u5e94\u7684\u89e3\u6790\u5f15\u64ce&#xff08;\u4f8b\u5982\u5bf9 .pdf \u6587\u4ef6\u542f\u7528 PDF \u89e3\u6790\u5668&#xff0c;\u5bf9 .md \u6587\u4ef6\u542f\u7528 Markdown \u89e3\u6790\u5668&#xff09;&#xff0c;\u65e0\u9700\u5f00\u53d1\u8005\u624b\u52a8\u6307\u5b9a\u6587\u4ef6\u7c7b\u578b\u3002<\/p>\n<p>\u7c7b\u539f\u578b<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">UnstructuredFileLoader<\/span><span class=\"token punctuation\">(<\/span>BaseLoader<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">def<\/span> <span class=\"token function\">__init__<\/span><span class=\"token punctuation\">(<\/span><br \/>\n        self<span class=\"token punctuation\">,<\/span><br \/>\n        file_path<span class=\"token punctuation\">:<\/span> Union<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">,<\/span> Path<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        mode<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;single&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        post_processors<span class=\"token punctuation\">:<\/span> Optional<span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Callable<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        <span class=\"token operator\">**<\/span>unstructured_kwargs<span class=\"token punctuation\">:<\/span> Any<span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;<br \/>\n        \u53c2\u6570\u8bf4\u660e&#xff1a;<br \/>\n        &#8211; file_path: \u6587\u4ef6\u8def\u5f84&#xff08;\u5b57\u7b26\u4e32\u6216 Path \u5bf9\u8c61&#xff09;<br \/>\n        &#8211; mode: \u89e3\u6790\u6a21\u5f0f&#xff08;&#034;single&#034; \u5408\u5e76\u4e3a\u5355\u4e2a\u6587\u6863&#xff1b;&#034;elements&#034; \u6309\u5143\u7d20\u62c6\u5206&#xff09;<br \/>\n        &#8211; post_processors: \u89e3\u6790\u540e\u5904\u7406\u51fd\u6570\u5217\u8868&#xff08;\u5982\u8fc7\u6ee4\u7a7a\u6587\u672c&#xff09;<br \/>\n        -** unstructured_kwargs: \u4f20\u9012\u7ed9 unstructured \u5e93\u7684\u539f\u751f\u53c2\u6570<br \/>\n        &#034;&#034;&#034;<\/span><\/p>\n<p>\u6838\u5fc3\u53c2\u6570\u89e3\u6790<\/p>\n<table>\n<tr>\u53c2\u6570\u540d\u7c7b\u578b\u529f\u80fd\u8be6\u89e3\u793a\u4f8b\u503c<\/tr>\n<tbody>\n<tr>\n<td>file_path<\/td>\n<td>str\/Path<\/td>\n<td>\u5f85\u89e3\u6790\u6587\u4ef6\u7684\u672c\u5730\u8def\u5f84&#xff08;\u5fc5\u586b&#xff09;<\/td>\n<td>\u201c.\/docs\/report.pdf\u201d<\/td>\n<\/tr>\n<tr>\n<td>mode<\/td>\n<td>str<\/td>\n<td>\u89e3\u6790\u540e\u6587\u6863\u7684\u7ec4\u7ec7\u65b9\u5f0f&#xff1a;\u2460 \u201csingle\u201d&#xff1a;\u6240\u6709\u5185\u5bb9\u5408\u5e76\u4e3a\u4e00\u4e2a Document \u5bf9\u8c61\u2461 \u201celements\u201d&#xff1a;\u6309\u5143\u7d20(\u6807\u9898\u3001\u6bb5\u843d\u7b49)\u62c6\u5206\u4e3a\u591a\u4e2a Document \u5bf9\u8c61<\/td>\n<td>\u201celements\u201d<\/td>\n<\/tr>\n<tr>\n<td>post_processors<\/td>\n<td>List[Callable]<\/td>\n<td>\u5bf9\u89e3\u6790\u540e\u7684\u5143\u7d20\u5217\u8868\u8fdb\u884c\u5904\u7406\u7684\u51fd\u6570&#xff08;\u5982\u8fc7\u6ee4\u7a7a\u767d\u3001\u5408\u5e76\u77ed\u6587\u672c&#xff09;<\/td>\n<td>[lambda els: [e for e in els if e.text.strip()]]<\/td>\n<\/tr>\n<tr>\n<td>strategy<\/td>\n<td>str<\/td>\n<td>\u89e3\u6790\u7b56\u7565&#xff08;\u901a\u8fc7 unstructured_kwargs \u4f20\u9012&#xff09;&#xff1a;\u2460 \u201cfast\u201d&#xff1a;\u5feb\u901f\u89e3\u6790(\u4f18\u5148\u6587\u672c\u63d0\u53d6)\u2461 \u201chi_res\u201d&#xff1a;\u9ad8\u7cbe\u5ea6\u89e3\u6790(\u4fdd\u7559\u590d\u6742\u7ed3\u6784)\u2462 \u201cocr_only\u201d&#xff1a;\u5f3a\u5236 OCR \u8bc6\u522b&#xff08;\u626b\u63cf\u4ef6&#xff09;<\/td>\n<td>\u201chi_res\u201d<\/td>\n<\/tr>\n<tr>\n<td>pdf_infer_table_structure<\/td>\n<td>bool<\/td>\n<td>\u662f\u5426\u89e3\u6790 PDF \u8868\u683c\u7ed3\u6784&#xff08;\u4ec5 PDF \u6709\u6548&#xff09;&#xff0c;\u8bbe\u4e3a True \u65f6\u8868\u683c\u8f6c\u4e3a Markdown \u683c\u5f0f<\/td>\n<td>True<\/td>\n<\/tr>\n<tr>\n<td>md_start_tag\/md_end_tag<\/td>\n<td>str<\/td>\n<td>Markdown \u89e3\u6790\u65f6\u7684\u8d77\u59cb\/\u7ed3\u675f\u6807\u8bb0&#xff08;\u7528\u4e8e\u63d0\u53d6\u6587\u6863\u7247\u6bb5&#xff09;<\/td>\n<td>\u201c\u201d, \u201c\u201d<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u65b9\u6cd5&#xff1a;load()<\/p>\n<p>\u529f\u80fd&#xff1a;\u6267\u884c\u89e3\u6790\u5e76\u8fd4\u56de\u6587\u6863\u5217\u8868 \u8fd4\u56de\u503c&#xff1a;List[Document]&#xff08;\u6bcf\u4e2a Document \u542b page_content \u6587\u672c\u548c metadata \u5143\u6570\u636e&#xff09;<\/p>\n<p>\u7b80\u5355\u4f7f\u7528\u793a\u4f8b<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredFileLoader<\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316\u52a0\u8f7d\u5668&#xff0c;\u89e3\u6790\u4e00\u4e2a\u7b80\u5355\u7684TXT\u6587\u4ef6&#xff08;\u901a\u7528\u52a0\u8f7d\u5668\u652f\u6301\u591a\u683c\u5f0f&#xff09;<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredFileLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/sample.txt&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u4efb\u610f\u683c\u5f0f\u6587\u4ef6\u8def\u5f84<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;single&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5408\u5e76\u4e3a\u5355\u4e2a\u6587\u6863<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;fast&#034;<\/span>  <span class=\"token comment\"># \u5feb\u901f\u89e3\u6790\u6a21\u5f0f<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6267\u884c\u89e3\u6790<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8f93\u51fa\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u89e3\u6790\u5f97\u5230\u7684\u6587\u6863\u6570\u91cf&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6587\u6863\u5185\u5bb9\u9884\u89c8&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">200]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6587\u6863\u5143\u6570\u636e&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8fd0\u884c\u7ed3\u679c<\/p>\n<p>\u89e3\u6790\u5f97\u5230\u7684\u6587\u6863\u6570\u91cf&#xff1a;1<br \/>\n\u6587\u6863\u5185\u5bb9\u9884\u89c8&#xff1a;\u8fd9\u662f\u4e00\u4e2a\u793a\u4f8b\u6587\u672c\u6587\u4ef6&#xff0c;\u5305\u542b\u7b80\u5355\u7684\u6bb5\u843d\u5185\u5bb9\u3002<br \/>\n\u5b83\u7528\u4e8e\u6f14\u793aUnstructuredFileLoader\u7684\u57fa\u672c\u529f\u80fd&#xff0c;<br \/>\n\u8be5\u52a0\u8f7d\u5668\u80fd\u591f\u81ea\u52a8\u8bc6\u522b\u6587\u4ef6\u7c7b\u578b\u5e76\u63d0\u53d6\u6587\u672c\u3002<br \/>\n\u6587\u6863\u5143\u6570\u636e&#xff1a;{&#039;source&#039;: &#039;.\/sample.txt&#039;, &#039;filename&#039;: &#039;sample.txt&#039;}<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u793a\u4f8b\u4e2d\u4f7f\u7528 UnstructuredFileLoader \u89e3\u6790 TXT \u6587\u4ef6&#xff0c;\u65e0\u9700\u989d\u5916\u914d\u7f6e\u683c\u5f0f\u53c2\u6570&#xff0c;\u4f53\u73b0\u4e86\u5176\u201c\u901a\u7528\u201d\u7279\u6027&#xff1b;<\/li>\n<li>mode&#061;&#034;single&#034; \u6a21\u5f0f\u4e0b&#xff0c;\u6240\u6709\u6587\u672c\u88ab\u5408\u5e76\u4e3a\u4e00\u4e2a Document \u5bf9\u8c61&#xff0c;\u9002\u5408\u9700\u8981\u5b8c\u6574\u6587\u672c\u7684\u573a\u666f&#xff1b;<\/li>\n<li>\u5143\u6570\u636e\u4e2d\u5305\u542b\u6587\u4ef6\u6765\u6e90&#xff08;source&#xff09;\u548c\u6587\u4ef6\u540d&#xff08;filename&#xff09;&#xff0c;\u4fbf\u4e8e\u8ffd\u8e2a\u6587\u6863\u539f\u59cb\u4f4d\u7f6e\u3002<\/li>\n<\/ul>\n<h4>3. UnstructuredPDFLoader \u4e0e UnstructuredMarkdownLoader<\/h4>\n<p>\u662f UnstructuredLoader \u7684\u5b50\u7c7b&#xff0c;\u9884\u914d\u7f6e\u4e86\u5bf9\u5e94\u683c\u5f0f\u7684\u89e3\u6790\u53c2\u6570&#xff0c;\u4f7f\u7528\u66f4\u7b80\u6d01\u3002<\/p>\n<h5>UnstructuredPDFLoader&#xff08;PDF \u4e13\u7528&#xff09;<\/h5>\n<p>UnstructuredPDFLoader \u662f\u9488\u5bf9 PDF \u683c\u5f0f\u4f18\u5316\u7684\u4e13\u7528\u52a0\u8f7d\u5668&#xff0c;\u7ee7\u627f\u81ea UnstructuredFileLoader\u3002\u5b83\u5728\u901a\u7528\u52a0\u8f7d\u5668\u7684\u57fa\u7840\u4e0a&#xff0c;\u9884\u8bbe\u4e86 PDF \u89e3\u6790\u7684\u4e13\u5c5e\u53c2\u6570\u548c\u5f15\u64ce&#xff08;\u5982\u9ed8\u8ba4\u542f\u7528 pdfminer \u89e3\u6790\u5668&#xff09;&#xff0c;\u80fd\u591f\u66f4\u7cbe\u51c6\u5730\u5904\u7406 PDF \u7279\u6709\u7684\u7ed3\u6784&#xff08;\u5982\u9875\u9762\u5e03\u5c40\u3001\u5b57\u4f53\u6837\u5f0f\u3001\u5d4c\u5165\u8868\u683c\u7b49&#xff09;\u3002\u5bf9\u4e8e\u5305\u542b\u590d\u6742\u6392\u7248&#xff08;\u5982\u591a\u5217\u6587\u672c\u3001\u8de8\u9875\u8868\u683c\u3001\u6c34\u5370&#xff09;\u7684 PDF \u6587\u6863&#xff0c;\u8be5\u52a0\u8f7d\u5668\u80fd\u901a\u8fc7\u5185\u7f6e\u7684\u683c\u5f0f\u9002\u914d\u903b\u8f91\u51cf\u5c11\u6587\u672c\u9519\u4e71\u95ee\u9898\u3002\u6b64\u5916&#xff0c;\u5b83\u539f\u751f\u652f\u6301 PDF \u7279\u6709\u7684\u89e3\u6790\u53c2\u6570&#xff08;\u5982 pdf_infer_table_structure \u7528\u4e8e\u8868\u683c\u8bc6\u522b&#xff09;&#xff0c;\u65e0\u9700\u901a\u8fc7 unstructured_kwargs \u95f4\u63a5\u4f20\u9012&#xff0c;\u63d0\u5347\u4e86\u4ee3\u7801\u7684\u53ef\u8bfb\u6027\u3002<\/p>\n<p>\u7c7b\u539f\u578b<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">UnstructuredPDFLoader<\/span><span class=\"token punctuation\">(<\/span>UnstructuredFileLoader<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">def<\/span> <span class=\"token function\">__init__<\/span><span class=\"token punctuation\">(<\/span><br \/>\n        self<span class=\"token punctuation\">,<\/span><br \/>\n        file_path<span class=\"token punctuation\">:<\/span> Union<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">,<\/span> Path<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        mode<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;single&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        post_processors<span class=\"token punctuation\">:<\/span> Optional<span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Callable<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        <span class=\"token operator\">**<\/span>unstructured_kwargs<span class=\"token punctuation\">:<\/span> Any<span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token builtin\">super<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>__init__<span class=\"token punctuation\">(<\/span>file_path<span class=\"token punctuation\">,<\/span> mode<span class=\"token punctuation\">,<\/span> post_processors<span class=\"token punctuation\">,<\/span><span class=\"token operator\">**<\/span> unstructured_kwargs<span class=\"token punctuation\">)<\/span><\/p>\n<p>\u6838\u5fc3\u53c2\u6570\u89e3\u6790&#xff08;\u4e0e\u7236\u7c7b\u4e00\u81f4&#xff0c;\u91cd\u70b9\u5f3a\u8c03 PDF \u9002\u914d\u7279\u6027&#xff09;<\/p>\n<table>\n<tr>\u53c2\u6570\u540d\u7c7b\u578b\u529f\u80fd\u8be6\u89e3&#xff08;PDF \u573a\u666f\u4e0b&#xff09;<\/tr>\n<tbody>\n<tr>\n<td>file_path<\/td>\n<td>str\/Path<\/td>\n<td>\u5f85\u89e3\u6790 PDF \u6587\u4ef6\u7684\u672c\u5730\u8def\u5f84&#xff08;\u5fc5\u586b&#xff09;&#xff0c;\u9700\u786e\u4fdd\u6587\u4ef6\u540e\u7f00\u4e3a .pdf<\/td>\n<\/tr>\n<tr>\n<td>mode<\/td>\n<td>str<\/td>\n<td>\u89e3\u6790\u6a21\u5f0f&#xff1a;- \u201csingle\u201d&#xff1a;\u5408\u5e76\u6240\u6709\u9875\u9762\u5185\u5bb9\u4e3a\u5355\u4e2a\u6587\u6863- \u201celements\u201d&#xff1a;\u6309 PDF \u4e2d\u7684\u5143\u7d20&#xff08;\u6807\u9898\u3001\u8868\u683c\u7b49&#xff09;\u62c6\u5206<\/td>\n<\/tr>\n<tr>\n<td>post_processors<\/td>\n<td>List[Callable]<\/td>\n<td>\u9488\u5bf9 PDF \u89e3\u6790\u7ed3\u679c\u7684\u540e\u5904\u7406\u51fd\u6570&#xff08;\u5982\u8fc7\u6ee4\u9875\u7709\u9875\u811a\u3001\u5408\u5e76\u8de8\u9875\u6bb5\u843d&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>**unstructured_kwargs<\/td>\n<td>\u4efb\u610f\u952e\u503c\u5bf9<\/td>\n<td>PDF \u4e13\u7528\u89e3\u6790\u53c2\u6570&#xff1a;- strategy&#xff1a;\u89e3\u6790\u7b56\u7565&#xff08;\u201chi_res\u201d \u9002\u5408\u590d\u6742\u6392\u7248 PDF&#xff09;- pdf_infer_table_structure&#xff1a;\u662f\u5426\u89e3\u6790\u8868\u683c\u7ed3\u6784- page_numbers&#xff1a;\u6307\u5b9a\u89e3\u6790\u7684\u9875\u7801\u5217\u8868&#xff08;\u5982 [1,3,5]&#xff09;<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<ul>\n<li>\u4e0e\u901a\u7528\u52a0\u8f7d\u5668\u7684\u533a\u522b&#xff1a;\u9ed8\u8ba4\u542f\u7528 PDF \u4e13\u7528\u89e3\u6790\u5f15\u64ce&#xff0c;\u65e0\u9700\u989d\u5916\u6307\u5b9a\u6587\u4ef6\u7c7b\u578b\u3002<\/li>\n<\/ul>\n<p>\u7b80\u5355\u4f7f\u7528\u793a\u4f8b<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredPDFLoader<\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316PDF\u4e13\u7528\u52a0\u8f7d\u5668&#xff0c;\u89e3\u6790\u542b\u6807\u9898\u548c\u6bb5\u843d\u7684PDF<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredPDFLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/report.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;elements&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u6309\u5143\u7d20\u62c6\u5206<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;fast&#034;<\/span>  <span class=\"token comment\"># \u5feb\u901f\u89e3\u6790<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6267\u884c\u89e3\u6790<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8f93\u51fa\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u89e3\u6790\u5f97\u5230\u7684\u5143\u7d20\u6570\u91cf&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token comment\"># \u6253\u5370\u7b2c\u4e00\u4e2a\u5143\u7d20\u7684\u7c7b\u578b\u548c\u5185\u5bb9<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u7b2c\u4e00\u4e2a\u5143\u7d20\u7c7b\u578b&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;element_type&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u7b2c\u4e00\u4e2a\u5143\u7d20\u5185\u5bb9&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">150]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8fd0\u884c\u7ed3\u679c<\/p>\n<p>\u89e3\u6790\u5f97\u5230\u7684\u5143\u7d20\u6570\u91cf&#xff1a;8<br \/>\n\u7b2c\u4e00\u4e2a\u5143\u7d20\u7c7b\u578b&#xff1a;Title<br \/>\n\u7b2c\u4e00\u4e2a\u5143\u7d20\u5185\u5bb9&#xff1a;2024\u5e74\u4eba\u5de5\u667a\u80fd\u53d1\u5c55\u62a5\u544a<br \/>\n&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;&#061;<br \/>\n\u53d1\u5e03\u65e5\u671f&#xff1a;2024\u5e741\u6708<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u4e13\u7528\u52a0\u8f7d\u5668\u81ea\u52a8\u8bc6\u522b PDF \u4e2d\u7684\u6807\u9898\u5143\u7d20&#xff08;element_type&#061;&#034;Title&#034;&#xff09;&#xff0c;\u4f53\u73b0\u4e86\u5bf9 PDF \u683c\u5f0f\u7684\u9488\u5bf9\u6027\u4f18\u5316&#xff1b;<\/li>\n<li>mode&#061;&#034;elements&#034; \u6a21\u5f0f\u4e0b&#xff0c;\u6587\u6863\u6309\u8bed\u4e49\u5355\u5143\u62c6\u5206&#xff0c;\u4fbf\u4e8e\u540e\u7eed\u6309\u5143\u7d20\u7c7b\u578b&#xff08;\u6807\u9898\/\u6bb5\u843d\u7b49&#xff09;\u7b5b\u9009\u5185\u5bb9&#xff1b;<\/li>\n<li>\u65e0\u9700\u624b\u52a8\u6307\u5b9a\u6587\u4ef6\u7c7b\u578b&#xff0c;\u52a0\u8f7d\u5668\u9ed8\u8ba4\u4f7f\u7528 PDF \u89e3\u6790\u5f15\u64ce&#xff0c;\u7b80\u5316\u4e86\u914d\u7f6e\u3002<\/li>\n<\/ul>\n<h5>UnstructuredMarkdownLoader&#xff08;Markdown \u4e13\u7528&#xff09;<\/h5>\n<p>UnstructuredMarkdownLoader \u662f\u4e13\u4e3a Markdown \u683c\u5f0f\u8bbe\u8ba1\u7684\u52a0\u8f7d\u5668&#xff0c;\u7ee7\u627f\u81ea UnstructuredFileLoader\u3002\u5b83\u6df1\u5ea6\u9002\u914d Markdown \u7684\u8bed\u6cd5\u89c4\u5219&#xff0c;\u80fd\u591f\u7cbe\u51c6\u8bc6\u522b\u6807\u9898\u5c42\u7ea7&#xff08;# \u81f3 ######&#xff09;\u3001\u5217\u8868&#xff08;\u6709\u5e8f\/\u65e0\u5e8f&#xff09;\u3001\u4ee3\u7801\u5757&#xff08;&#096;&#096;&#096;\u5305\u88f9&#xff09;\u3001\u5f15\u7528&#xff08;&gt; \u5f00\u5934&#xff09;\u7b49\u7279\u6709\u5143\u7d20&#xff0c;\u5e76\u5728\u5143\u6570\u636e\u4e2d\u6807\u8bb0\u5bf9\u5e94\u7684 element_type&#xff08;\u5982 Header\u3001ListItem\u3001Code&#xff09;\u3002\u5bf9\u4e8e\u5305\u542b\u5d4c\u5957\u7ed3\u6784&#xff08;\u5982\u5217\u8868\u4e2d\u5d4c\u5957\u4ee3\u7801\u5757&#xff09;\u7684\u590d\u6742 Markdown \u6587\u6863&#xff0c;\u8be5\u52a0\u8f7d\u5668\u80fd\u4fdd\u6301\u5143\u7d20\u95f4\u7684\u5c42\u7ea7\u5173\u7cfb&#xff0c;\u907f\u514d\u89e3\u6790\u9519\u4e71\u3002\u4e0e\u901a\u7528\u52a0\u8f7d\u5668\u76f8\u6bd4&#xff0c;\u5b83\u65e0\u9700\u81ea\u52a8\u5224\u65ad\u6587\u4ef6\u7c7b\u578b&#xff0c;\u76f4\u63a5\u8c03\u7528 Markdown \u4e13\u7528\u89e3\u6790\u903b\u8f91&#xff0c;\u89e3\u6790\u6548\u7387\u548c\u7cbe\u5ea6\u66f4\u9ad8\u3002<\/p>\n<p>\u7c7b\u539f\u578b<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">UnstructuredMarkdownLoader<\/span><span class=\"token punctuation\">(<\/span>UnstructuredFileLoader<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">def<\/span> <span class=\"token function\">__init__<\/span><span class=\"token punctuation\">(<\/span><br \/>\n        self<span class=\"token punctuation\">,<\/span><br \/>\n        file_path<span class=\"token punctuation\">:<\/span> Union<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">,<\/span> Path<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        mode<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;single&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        post_processors<span class=\"token punctuation\">:<\/span> Optional<span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Callable<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span>List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> List<span class=\"token punctuation\">[<\/span>Element<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        <span class=\"token operator\">**<\/span>unstructured_kwargs<span class=\"token punctuation\">:<\/span> Any<span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> <span class=\"token boolean\">None<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token builtin\">super<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>__init__<span class=\"token punctuation\">(<\/span>file_path<span class=\"token punctuation\">,<\/span> mode<span class=\"token punctuation\">,<\/span> post_processors<span class=\"token punctuation\">,<\/span><span class=\"token operator\">**<\/span> unstructured_kwargs<span class=\"token punctuation\">)<\/span><\/p>\n<p>\u6838\u5fc3\u53c2\u6570\u89e3\u6790&#xff08;\u4e0e\u7236\u7c7b\u4e00\u81f4&#xff0c;\u91cd\u70b9\u5f3a\u8c03 Markdown \u9002\u914d\u7279\u6027&#xff09;<\/p>\n<table>\n<tr>\u53c2\u6570\u540d\u7c7b\u578b\u529f\u80fd\u8be6\u89e3&#xff08;Markdown \u573a\u666f\u4e0b&#xff09;<\/tr>\n<tbody>\n<tr>\n<td>file_path<\/td>\n<td>str\/Path<\/td>\n<td>\u5f85\u89e3\u6790 Markdown \u6587\u4ef6\u7684\u672c\u5730\u8def\u5f84&#xff08;\u5fc5\u586b&#xff09;&#xff0c;\u9700\u786e\u4fdd\u6587\u4ef6\u540e\u7f00\u4e3a .md \u6216 .markdown<\/td>\n<\/tr>\n<tr>\n<td>mode<\/td>\n<td>str<\/td>\n<td>\u89e3\u6790\u6a21\u5f0f&#xff1a;\u2460 \u201csingle\u201d&#xff1a;\u5408\u5e76\u6240\u6709\u5185\u5bb9\u4e3a\u5355\u4e2a\u6587\u6863\u2461 \u201celements\u201d&#xff1a;\u6309 Markdown \u5143\u7d20&#xff08;\u6807\u9898\u3001\u5217\u8868\u7b49&#xff09;\u62c6\u5206<\/td>\n<\/tr>\n<tr>\n<td>post_processors<\/td>\n<td>List[Callable]<\/td>\n<td>\u9488\u5bf9 Markdown \u89e3\u6790\u7ed3\u679c\u7684\u540e\u5904\u7406\u51fd\u6570&#xff08;\u5982\u8fc7\u6ee4\u6ce8\u91ca\u3001\u5408\u5e76\u76f8\u90bb\u6587\u672c&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>**unstructured_kwargs<\/td>\n<td>\u4efb\u610f\u952e\u503c\u5bf9<\/td>\n<td>Markdown \u4e13\u7528\u89e3\u6790\u53c2\u6570&#xff1a;\u2460 md_start_tag\/md_end_tag&#xff1a;\u63d0\u53d6\u6307\u5b9a\u6807\u8bb0\u5305\u88f9\u7684\u5185\u5bb9\u2461 include_headers_footers&#xff1a;\u662f\u5426\u5305\u542b\u6587\u6863\u9996\u5c3e\u7684\u6ce8\u91ca\u5185\u5bb9<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<ul>\n<li>\u7279\u6027&#xff1a;\u81ea\u52a8\u8bc6\u522b Markdown \u4e2d\u7684\u6807\u9898&#xff08;#&#xff09;\u3001\u5217\u8868&#xff08;-\/*&#xff09;\u3001\u4ee3\u7801\u5757&#xff08;&#096;&#096;&#096;&#xff09;\u7b49\u5143\u7d20\u3002<\/li>\n<\/ul>\n<p>\u7b80\u5355\u4f7f\u7528\u793a\u4f8b<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredMarkdownLoader<\/p>\n<p><span class=\"token comment\"># \u521d\u59cb\u5316Markdown\u4e13\u7528\u52a0\u8f7d\u5668&#xff0c;\u89e3\u6790\u542b\u5217\u8868\u7684\u6587\u6863<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredMarkdownLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/guide.md&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;elements&#034;<\/span>  <span class=\"token comment\"># \u6309\u5143\u7d20\u62c6\u5206<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6267\u884c\u89e3\u6790<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u7b5b\u9009\u5217\u8868\u9879\u5143\u7d20<\/span><br \/>\nlist_items <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>doc <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> documents <span class=\"token keyword\">if<\/span> doc<span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;element_type&#034;<\/span><span class=\"token punctuation\">]<\/span> <span class=\"token operator\">&#061;&#061;<\/span> <span class=\"token string\">&#034;ListItem&#034;<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u8f93\u51fa\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u89e3\u6790\u5f97\u5230\u7684\u5217\u8868\u9879\u6570\u91cf&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>list_items<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u7b2c\u4e00\u4e2a\u5217\u8868\u9879\u5185\u5bb9&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>list_items<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8fd0\u884c\u7ed3\u679c<\/p>\n<p>\u89e3\u6790\u5f97\u5230\u7684\u5217\u8868\u9879\u6570\u91cf&#xff1a;3<br \/>\n\u7b2c\u4e00\u4e2a\u5217\u8868\u9879\u5185\u5bb9&#xff1a;- \u5b89\u88c5\u4f9d\u8d56&#xff1a;&#096;pip install langchain-community unstructured&#096;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u4e13\u7528\u52a0\u8f7d\u5668\u7cbe\u51c6\u8bc6\u522b Markdown \u4e2d\u7684\u5217\u8868\u9879&#xff08;element_type&#061;&#034;ListItem&#034;&#xff09;&#xff0c;\u4fdd\u7559\u4e86\u539f\u59cb\u683c\u5f0f&#xff08;\u5305\u62ec &#8211; \u7b26\u53f7\u548c\u4ee3\u7801\u5757\u6807\u8bb0&#xff09;&#xff1b;<\/li>\n<li>\u5bf9\u4e8e\u6280\u672f\u6587\u6863\u7b49\u5bcc\u542b\u7ed3\u6784\u5316\u5143\u7d20\u7684 Markdown \u6587\u4ef6&#xff0c;\u8fd9\u79cd\u9488\u5bf9\u6027\u89e3\u6790\u80fd\u6709\u6548\u5206\u79bb\u4e0d\u540c\u7c7b\u578b\u7684\u5185\u5bb9&#xff0c;\u4fbf\u4e8e\u540e\u7eed\u5904\u7406&#xff08;\u5982\u5355\u72ec\u63d0\u53d6\u4ee3\u7801\u5757&#xff09;\u3002<\/li>\n<\/ul>\n<h3>\u4e09\u3001PDF \u89e3\u6790\u5b9e\u6218\u793a\u4f8b<\/h3>\n<h4>\u793a\u4f8b 1&#xff1a;\u9ad8\u7cbe\u5ea6\u89e3\u6790\u542b\u8868\u683c\u7684 PDF<\/h4>\n<p>\u89e3\u6790\u5305\u542b\u6807\u9898\u3001\u6bb5\u843d\u548c\u8868\u683c\u7684\u590d\u6742 PDF&#xff0c;\u4fdd\u7559\u8868\u683c\u7ed3\u6784\u3002<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredPDFLoader<\/p>\n<p><span class=\"token comment\"># 1. \u521d\u59cb\u5316\u52a0\u8f7d\u5668&#xff08;\u9ad8\u7cbe\u5ea6\u6a21\u5f0f &#043; \u89e3\u6790\u8868\u683c&#xff09;<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredPDFLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/example_pdf.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;elements&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u6309\u5143\u7d20\u62c6\u5206<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;hi_res&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u9ad8\u7cbe\u5ea6\u89e3\u6790<\/span><br \/>\n    pdf_infer_table_structure<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span>  <span class=\"token comment\"># \u89e3\u6790\u8868\u683c\u4e3a Markdown \u683c\u5f0f<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 2. \u6267\u884c\u89e3\u6790<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 3. \u8f93\u51fa\u89e3\u6790\u7ed3\u679c&#xff08;\u7b5b\u9009\u5173\u952e\u5143\u7d20&#xff09;<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5171\u89e3\u6790\u51fa <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\"> \u4e2a\u5143\u7d20&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> documents<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token number\">4<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">:<\/span>  <span class=\"token comment\"># \u6253\u5370\u524d4\u4e2a\u5143\u7d20<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\\\\n\u7c7b\u578b&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;element_type&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u9875\u7801&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;page_number&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5185\u5bb9&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">150]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#8230;&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u9884\u89c8\u524d150\u5b57\u7b26<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa(\u8fd9\u91cc\u5177\u4f53\u5185\u5bb9\u6839\u636e\u6587\u4ef6\u5185\u5bb9\u6765\u5b9a&#xff0c;\u4e0b\u9762\u8f93\u51fa\u4ec5\u4f5c\u53c2\u8003)<\/p>\n<p>\u5171\u89e3\u6790\u51fa 12 \u4e2a\u5143\u7d20<\/p>\n<p>\u7c7b\u578b&#xff1a;Title<br \/>\n\u9875\u7801&#xff1a;1<br \/>\n\u5185\u5bb9&#xff1a;2023 \u4eba\u5de5\u667a\u80fd\u884c\u4e1a\u62a5\u544a&#8230;<\/p>\n<p>\u7c7b\u578b&#xff1a;NarrativeText<br \/>\n\u9875\u7801&#xff1a;1<br \/>\n\u5185\u5bb9&#xff1a;\u4eba\u5de5\u667a\u80fd&#xff08;AI&#xff09;\u884c\u4e1a\u5728 2023 \u5e74\u8fce\u6765\u7206\u53d1\u5f0f\u589e\u957f&#xff0c;\u5168\u7403\u5e02\u573a\u89c4\u6a21\u7a81\u7834 1.5 \u4e07\u4ebf\u7f8e\u5143&#xff0c;\u540c\u6bd4\u589e\u957f 37%&#8230;<\/p>\n<p>\u7c7b\u578b&#xff1a;Table<br \/>\n\u9875\u7801&#xff1a;2<br \/>\n\u5185\u5bb9&#xff1a;| \u9886\u57df       | \u5e02\u573a\u89c4\u6a21&#xff08;\u4ebf\u7f8e\u5143&#xff09; | \u589e\u957f\u7387 |<br \/>\n|&#8212;&#8212;&#8212;&#8212;|&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;|&#8212;&#8212;&#8211;|<br \/>\n| \u8ba1\u7b97\u673a\u89c6\u89c9   | 4200             | 45%    |<br \/>\n| \u81ea\u7136\u8bed\u8a00\u5904\u7406 | 3800             | 52%    |&#8230;<\/p>\n<p>\u7c7b\u578b&#xff1a;ListItem<br \/>\n\u9875\u7801&#xff1a;3<br \/>\n\u5185\u5bb9&#xff1a;1. \u6280\u672f\u8d8b\u52bf&#xff1a;\u591a\u6a21\u6001\u6a21\u578b\u6210\u4e3a\u4e3b\u6d41&#xff0c;\u8de8\u9886\u57df\u878d\u5408\u80fd\u529b\u663e\u8457\u63d0\u5347&#8230;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u89e3\u6790\u7ed3\u679c\u6309\u5143\u7d20\u7c7b\u578b&#xff08;\u6807\u9898\u3001\u6b63\u6587\u3001\u8868\u683c\u3001\u5217\u8868\u9879&#xff09;\u62c6\u5206&#xff0c;\u5143\u6570\u636e\u5305\u542b\u9875\u7801\u548c\u7c7b\u578b&#xff0c;\u4fbf\u4e8e\u540e\u7eed\u7b5b\u9009\u3002<\/li>\n<li>\u8868\u683c\u88ab\u8f6c\u4e3a Markdown \u683c\u5f0f&#xff08;| \u5b57\u6bb5 | \u503c |&#xff09;&#xff0c;\u4fdd\u7559\u4e86\u539f\u59cb\u7ed3\u6784&#xff0c;\u53ef\u76f4\u63a5\u7528\u4e8e LLM \u5904\u7406\u6216\u8f6c\u6362\u4e3a DataFrame\u3002<\/li>\n<\/ul>\n<h4>\u793a\u4f8b 2&#xff1a;\u5904\u7406\u626b\u63cf\u578b PDF&#xff08;OCR \u8bc6\u522b&#xff09;<\/h4>\n<p>\u5bf9\u65e0\u6cd5\u76f4\u63a5\u590d\u5236\u6587\u672c\u7684\u626b\u63cf\u4ef6 PDF&#xff0c;\u4f7f\u7528 OCR \u7b56\u7565\u89e3\u6790\u3002<\/p>\n<p>loader <span class=\"token operator\">&#061;<\/span> UnstructuredPDFLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/scan_pdf.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u626b\u63cf\u578b PDF&#xff08;\u56fe\u7247\u5185\u5bb9&#xff09;<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;single&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5408\u5e76\u4e3a\u5355\u4e2a\u6587\u6863<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;ocr_only&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5f3a\u5236 OCR \u8bc6\u522b<\/span><br \/>\n    ocr_languages<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;eng&#043;chi_sim&#034;<\/span>  <span class=\"token comment\"># \u652f\u6301\u82f1\u6587\u548c\u4e2d\u6587<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p>documents <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;OCR \u89e3\u6790\u7ed3\u679c&#xff1a;\\\\n<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>documents<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">500]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>OCR \u89e3\u6790\u7ed3\u679c&#xff1a;<br \/>\n\u8fd9\u662f\u4e00\u4efd\u626b\u63cf\u7684\u4e2d\u6587\u62a5\u544a&#xff0c;\u5305\u542b\u4e2d\u82f1\u6587\u6df7\u5408\u5185\u5bb9\u3002<br \/>\n\u62a5\u544a\u6458\u8981&#xff1a;<br \/>\n2023 \u5e74\u5168\u7403 AI \u4e13\u5229\u7533\u8bf7\u91cf\u8fbe 12 \u4e07\u4ef6&#xff0c;\u5176\u4e2d\u4e2d\u56fd\u5360\u6bd4 45%&#xff0c;\u4f4d\u5c45\u5168\u7403\u7b2c\u4e00\u3002<br \/>\nKey Findings:<br \/>\n&#8211; AI \u82af\u7247\u6027\u80fd\u63d0\u5347 300%&#xff0c;\u6210\u672c\u4e0b\u964d 50%&#8230;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>strategy&#061;&#034;ocr_only&#034; \u5f3a\u5236\u542f\u7528 OCR \u5f15\u64ce&#xff0c;\u5373\u4f7f PDF \u662f\u56fe\u7247\u4e5f\u80fd\u63d0\u53d6\u6587\u672c\u3002<\/li>\n<li>ocr_languages \u6307\u5b9a\u8bc6\u522b\u8bed\u8a00&#xff08;chi_sim \u4e3a\u7b80\u4f53\u4e2d\u6587&#xff09;&#xff0c;\u786e\u4fdd\u591a\u8bed\u8a00\u5185\u5bb9\u51c6\u786e\u89e3\u6790\u3002<\/li>\n<\/ul>\n<h3>\u56db\u3001Markdown \u89e3\u6790\u5b9e\u6218\u793a\u4f8b<\/h3>\n<h4>\u793a\u4f8b 3&#xff1a;\u89e3\u6790\u542b\u590d\u6742\u7ed3\u6784\u7684 Markdown<\/h4>\n<p>\u89e3\u6790\u5305\u542b\u6807\u9898\u3001\u4ee3\u7801\u5757\u3001\u516c\u5f0f\u7684 Markdown \u6587\u4ef6\u3002<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredMarkdownLoader<\/p>\n<p><span class=\"token comment\"># 1. \u521d\u59cb\u5316\u52a0\u8f7d\u5668<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredMarkdownLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/example_md.md&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;elements&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u6309\u5143\u7d20\u62c6\u5206<\/span><br \/>\n    md_start_tag<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;&lt;CONTENT&gt;&#034;<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u53ef\u9009&#xff1a;\u4ec5\u89e3\u6790\u6807\u7b7e\u5185\u7684\u5185\u5bb9<\/span><br \/>\n    md_end_tag<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;&lt;\/CONTENT&gt;&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 2. \u6267\u884c\u89e3\u6790<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 3. \u8f93\u51fa\u89e3\u6790\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> documents<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\\\\n\u7c7b\u578b&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;element_type&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5185\u5bb9&#xff1a;<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">200]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#8230;&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>\u7c7b\u578b&#xff1a;Header<br \/>\n\u5185\u5bb9&#xff1a;# \u673a\u5668\u5b66\u4e60\u5de5\u4f5c\u6d41\u8be6\u89e3&#8230;<\/p>\n<p>\u7c7b\u578b&#xff1a;Paragraph<br \/>\n\u5185\u5bb9&#xff1a;\u673a\u5668\u5b66\u4e60\u5de5\u4f5c\u6d41\u5305\u542b\u6570\u636e\u6536\u96c6\u3001\u9884\u5904\u7406\u3001\u6a21\u578b\u8bad\u7ec3\u3001\u8bc4\u4f30\u56db\u4e2a\u6838\u5fc3\u9636\u6bb5&#xff0c;\u5404\u9636\u6bb5\u9700\u7d27\u5bc6\u914d\u5408&#8230;<\/p>\n<p>\u7c7b\u578b&#xff1a;Code<br \/>\n\u5185\u5bb9&#xff1a;&#096;&#096;&#096;python<br \/>\n# \u6570\u636e\u9884\u5904\u7406\u793a\u4f8b<br \/>\nimport pandas as pd<br \/>\ndf &#061; pd.read_csv(&#034;data.csv&#034;)<br \/>\ndf &#061; df.dropna()  # \u53bb\u9664\u7f3a\u5931\u503c<br \/>\n&#096;&#096;&#096;&#8230;<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>Markdown \u4e2d\u7684\u5143\u7d20\u88ab\u7cbe\u51c6\u8bc6\u522b&#xff1a;Header&#xff08;\u6807\u9898&#xff09;\u3001Paragraph&#xff08;\u6bb5\u843d&#xff09;\u3001Code&#xff08;\u4ee3\u7801\u5757&#xff09;\u3002<\/li>\n<li>\u901a\u8fc7 md_start_tag \u548c md_end_tag \u53ef\u63d0\u53d6\u6587\u6863\u4e2d\u7684\u7279\u5b9a\u7247\u6bb5&#xff08;\u5982\u4ec5\u89e3\u6790 &lt;CONTENT&gt; \u6807\u7b7e\u5305\u88f9\u7684\u5185\u5bb9&#xff09;&#xff0c;\u9002\u5408\u5904\u7406\u5927\u578b\u6587\u6863\u3002<\/li>\n<\/ul>\n<h3>\u4e94\u3001\u8fdb\u9636\u6280\u5de7&#xff1a;\u540e\u5904\u7406\u51fd\u6570\u4e0e\u5143\u7d20\u8fc7\u6ee4<\/h3>\n<p>\u89e3\u6790\u540e\u7684\u5143\u7d20\u53ef\u80fd\u5305\u542b\u7a7a\u767d\u6587\u672c\u6216\u5197\u4f59\u5185\u5bb9&#xff0c;\u53ef\u901a\u8fc7 post_processors \u4f18\u5316\u3002<\/p>\n<p><span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_loaders <span class=\"token keyword\">import<\/span> UnstructuredFileLoader<\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u540e\u5904\u7406\u51fd\u6570&#xff1a;\u8fc7\u6ee4\u7a7a\u767d\u6587\u672c &#043; \u5408\u5e76\u77ed\u6587\u672c<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">filter_and_merge<\/span><span class=\"token punctuation\">(<\/span>elements<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># 1. \u8fc7\u6ee4\u7a7a\u767d\u5143\u7d20<\/span><br \/>\n    filtered <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>e <span class=\"token keyword\">for<\/span> e <span class=\"token keyword\">in<\/span> elements <span class=\"token keyword\">if<\/span> e<span class=\"token punctuation\">.<\/span>text<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">and<\/span> <span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>e<span class=\"token punctuation\">.<\/span>text<span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&gt;<\/span> <span class=\"token number\">50<\/span><span class=\"token punctuation\">]<\/span><br \/>\n    <span class=\"token comment\"># 2. \u5408\u5e76\u77ed\u6587\u672c&#xff08;\u5c0f\u4e8e200\u5b57\u7b26\u7684\u76f8\u90bb\u5143\u7d20&#xff09;<\/span><br \/>\n    merged <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><br \/>\n    current <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> e <span class=\"token keyword\">in<\/span> filtered<span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token keyword\">if<\/span> current <span class=\"token keyword\">and<\/span> <span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034; &#034;<\/span><span class=\"token punctuation\">.<\/span>join<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span>c<span class=\"token punctuation\">.<\/span>text <span class=\"token keyword\">for<\/span> c <span class=\"token keyword\">in<\/span> current<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#043;<\/span> <span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>e<span class=\"token punctuation\">.<\/span>text<span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&lt;<\/span> <span class=\"token number\">200<\/span><span class=\"token punctuation\">:<\/span><br \/>\n            current<span class=\"token punctuation\">.<\/span>append<span class=\"token punctuation\">(<\/span>e<span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token keyword\">else<\/span><span class=\"token punctuation\">:<\/span><br \/>\n            <span class=\"token keyword\">if<\/span> current<span class=\"token punctuation\">:<\/span><br \/>\n                merged<span class=\"token punctuation\">.<\/span>append<span class=\"token punctuation\">(<\/span>current<span class=\"token punctuation\">)<\/span><br \/>\n            current <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>e<span class=\"token punctuation\">]<\/span><br \/>\n    <span class=\"token keyword\">if<\/span> current<span class=\"token punctuation\">:<\/span><br \/>\n        merged<span class=\"token punctuation\">.<\/span>append<span class=\"token punctuation\">(<\/span>current<span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token comment\"># 3. \u8f6c\u6362\u4e3a\u5408\u5e76\u540e\u7684\u5143\u7d20<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">[<\/span>merged_elem<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>merge<span class=\"token punctuation\">(<\/span>merged_elem<span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">:<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">for<\/span> merged_elem <span class=\"token keyword\">in<\/span> merged<span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u5e94\u7528\u540e\u5904\u7406\u51fd\u6570<\/span><br \/>\nloader <span class=\"token operator\">&#061;<\/span> UnstructuredFileLoader<span class=\"token punctuation\">(<\/span><br \/>\n    file_path<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;.\/example.pdf&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;elements&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    post_processors<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>filter_and_merge<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u542f\u7528\u540e\u5904\u7406<\/span><br \/>\n    strategy<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;fast&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p>documents <span class=\"token operator\">&#061;<\/span> loader<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u540e\u5904\u7406\u540e\u5269\u4f59 <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\"> \u4e2a\u5143\u7d20&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>\u9884\u671f\u8f93\u51fa<\/p>\n<p>\u540e\u5904\u7406\u540e\u5269\u4f59 8 \u4e2a\u5143\u7d20<\/p>\n<p>\u7ed3\u679c\u5206\u6790<\/p>\n<ul>\n<li>\u540e\u5904\u7406\u51fd\u6570\u8fc7\u6ee4\u4e86\u7a7a\u767d\u6587\u672c\u548c\u8fc7\u77ed\u5143\u7d20&#xff08;&lt;50\u5b57\u7b26&#xff09;&#xff0c;\u5e76\u5408\u5e76\u4e86\u76f8\u90bb\u7684\u77ed\u6587\u672c&#xff0c;\u51cf\u5c11\u4e86\u5143\u7d20\u6570\u91cf&#xff0c;\u4fbf\u4e8e\u540e\u7eed\u5904\u7406\u3002<\/li>\n<li>element.merge() \u65b9\u6cd5\u4fdd\u7559\u4e86\u5408\u5e76\u5143\u7d20\u7684\u5143\u6570\u636e&#xff08;\u5982\u6700\u65e9\u7684\u9875\u7801&#xff09;&#xff0c;\u786e\u4fdd\u6eaf\u6e90\u6027\u3002<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb1.1k\u6b21\uff0c\u70b9\u8d5e35\u6b21\uff0c\u6536\u85cf6\u6b21\u3002\u6458\u8981\uff1a LangChain-Unstructured\u662fLangChain\u751f\u6001\u4e2d\u5904\u7406\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u6838\u5fc3\u5de5\u5177\uff0c\u652f\u6301PDF\u548cMarkdown\u7b49\u591a\u79cd\u683c\u5f0f\u89e3\u6790\u3002\u672c\u6587\u4ecb\u7ecd\u4e86\u5176\u6838\u5fc3\u4f9d\u8d56\u5e93\uff08\u5982langchain-community\u3001unstructured\u7b49\uff09\u548c\u4e3b\u8981\u7c7b\u65b9\u6cd5\u3002\u91cd\u70b9\u89e3\u6790\u4e86UnstructuredLoader\u7c7b\u7684\u529f\u80fd\u4e0e\u53c2\u6570\u914d\u7f6e\uff0c\u5305\u62ec\u591a\u8f93\u5165\u6e90\u652f\u6301\u3001\u672c\u5730\/API\u89e3\u6790\u65b9\u5f0f\u9009\u62e9\u3001\u6587\u672c\u540e\u5904\u7406\u7b49\u529f\u80fd\u3002\u8be5\u5de5\u5177\u901a\u8fc7\u7075\u6d3b\u7684\u53c2\u6570\u8bbe\u7f6e\u53ef\u6ee1\u8db3\u4e0d\u540c\u573a\u666f\u9700\u6c42\uff0c\u8f93\u51fa\u7ed3\u6784\u5316\u6587\u6863\u5bf9\u8c61\uff0c\u4e3a\u540e\u7eed\u6587\u672c\u5904\u7406\u63d0\u4f9b\u57fa\u7840\u652f\u6301\u3002\u5b89\u88c5\u7b80\u5355\uff0c\u53ea\u9700\u57fa\u7840Python\u73af\u5883<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[997,2150,132,5113,224],"topic":[],"class_list":["post-51571","post","type-post","status-publish","format-standard","hentry","category-server","tag-langchain","tag-pdf","tag-132","tag-5113","tag-224"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/51571.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb1.1k\u6b21\uff0c\u70b9\u8d5e35\u6b21\uff0c\u6536\u85cf6\u6b21\u3002\u6458\u8981\uff1a LangChain-Unstructured\u662fLangChain\u751f\u6001\u4e2d\u5904\u7406\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u6838\u5fc3\u5de5\u5177\uff0c\u652f\u6301PDF\u548cMarkdown\u7b49\u591a\u79cd\u683c\u5f0f\u89e3\u6790\u3002\u672c\u6587\u4ecb\u7ecd\u4e86\u5176\u6838\u5fc3\u4f9d\u8d56\u5e93\uff08\u5982langchain-community\u3001unstructured\u7b49\uff09\u548c\u4e3b\u8981\u7c7b\u65b9\u6cd5\u3002\u91cd\u70b9\u89e3\u6790\u4e86UnstructuredLoader\u7c7b\u7684\u529f\u80fd\u4e0e\u53c2\u6570\u914d\u7f6e\uff0c\u5305\u62ec\u591a\u8f93\u5165\u6e90\u652f\u6301\u3001\u672c\u5730\/API\u89e3\u6790\u65b9\u5f0f\u9009\u62e9\u3001\u6587\u672c\u540e\u5904\u7406\u7b49\u529f\u80fd\u3002\u8be5\u5de5\u5177\u901a\u8fc7\u7075\u6d3b\u7684\u53c2\u6570\u8bbe\u7f6e\u53ef\u6ee1\u8db3\u4e0d\u540c\u573a\u666f\u9700\u6c42\uff0c\u8f93\u51fa\u7ed3\u6784\u5316\u6587\u6863\u5bf9\u8c61\uff0c\u4e3a\u540e\u7eed\u6587\u672c\u5904\u7406\u63d0\u4f9b\u57fa\u7840\u652f\u6301\u3002\u5b89\u88c5\u7b80\u5355\uff0c\u53ea\u9700\u57fa\u7840Python\u73af\u5883\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/51571.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2025-08-10T16:29:15+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"10 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/51571.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/51571.html\",\"name\":\"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2025-08-10T16:29:15+00:00\",\"dateModified\":\"2025-08-10T16:29:15+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/51571.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/51571.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/51571.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/51571.html","og_locale":"zh_CN","og_type":"article","og_title":"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb1.1k\u6b21\uff0c\u70b9\u8d5e35\u6b21\uff0c\u6536\u85cf6\u6b21\u3002\u6458\u8981\uff1a LangChain-Unstructured\u662fLangChain\u751f\u6001\u4e2d\u5904\u7406\u975e\u7ed3\u6784\u5316\u6587\u6863\u7684\u6838\u5fc3\u5de5\u5177\uff0c\u652f\u6301PDF\u548cMarkdown\u7b49\u591a\u79cd\u683c\u5f0f\u89e3\u6790\u3002\u672c\u6587\u4ecb\u7ecd\u4e86\u5176\u6838\u5fc3\u4f9d\u8d56\u5e93\uff08\u5982langchain-community\u3001unstructured\u7b49\uff09\u548c\u4e3b\u8981\u7c7b\u65b9\u6cd5\u3002\u91cd\u70b9\u89e3\u6790\u4e86UnstructuredLoader\u7c7b\u7684\u529f\u80fd\u4e0e\u53c2\u6570\u914d\u7f6e\uff0c\u5305\u62ec\u591a\u8f93\u5165\u6e90\u652f\u6301\u3001\u672c\u5730\/API\u89e3\u6790\u65b9\u5f0f\u9009\u62e9\u3001\u6587\u672c\u540e\u5904\u7406\u7b49\u529f\u80fd\u3002\u8be5\u5de5\u5177\u901a\u8fc7\u7075\u6d3b\u7684\u53c2\u6570\u8bbe\u7f6e\u53ef\u6ee1\u8db3\u4e0d\u540c\u573a\u666f\u9700\u6c42\uff0c\u8f93\u51fa\u7ed3\u6784\u5316\u6587\u6863\u5bf9\u8c61\uff0c\u4e3a\u540e\u7eed\u6587\u672c\u5904\u7406\u63d0\u4f9b\u57fa\u7840\u652f\u6301\u3002\u5b89\u88c5\u7b80\u5355\uff0c\u53ea\u9700\u57fa\u7840Python\u73af\u5883","og_url":"https:\/\/www.wsisp.com\/helps\/51571.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2025-08-10T16:29:15+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"10 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/51571.html","url":"https:\/\/www.wsisp.com\/helps\/51571.html","name":"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2025-08-10T16:29:15+00:00","dateModified":"2025-08-10T16:29:15+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/51571.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/51571.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/51571.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"LangChain-Unstructured \u57fa\u7840\u4f7f\u7528\uff1aPDF \u4e0e Markdown \u5904\u7406\u89e3\u6790"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/51571","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=51571"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/51571\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=51571"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=51571"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=51571"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=51571"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}