4.3 文档处理模块.html 56 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658
  1. <!doctype html>
  2. <html lang="zh" class="no-js">
  3. <head>
  4. <meta charset="utf-8">
  5. <meta name="viewport" content="width=device-width,initial-scale=1">
  6. <link rel="prev" href="4.2%20%E5%9F%BA%E7%A1%80%E6%A8%A1%E5%9D%97.html">
  7. <link rel="next" href="4.4%20%E5%90%91%E9%87%8F%E5%AD%98%E5%82%A8.html">
  8. <link rel="icon" href="../img/AI.jpg">
  9. <meta name="generator" content="mkdocs-1.6.1, mkdocs-material-9.6.21">
  10. <title>4.3 文档处理模块 - 企业知识库RAG智能问答项目V1.0</title>
  11. <link rel="stylesheet" href="../assets/stylesheets/main.2a3383ac.min.css">
  12. <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
  13. <link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Roboto:300,300i,400,400i,700,700i%7CRoboto+Mono:400,400i,700,700i&amp;display=fallback">
  14. <style>:root{--md-text-font:"Roboto";--md-code-font:"Roboto Mono"}</style>
  15. <link rel="stylesheet" href="../stylesheets/extra.css">
  16. <script>__md_scope=new URL("..",location),__md_hash=e=>[...e].reduce(((e,_)=>(e<<5)-e+_.charCodeAt(0)),0),__md_get=(e,_=localStorage,t=__md_scope)=>JSON.parse(_.getItem(t.pathname+"."+e)),__md_set=(e,_,t=localStorage,a=__md_scope)=>{try{t.setItem(a.pathname+"."+e,JSON.stringify(_))}catch(e){}}</script>
  17. </head>
  18. <body dir="ltr">
  19. <input class="md-toggle" data-md-toggle="drawer" type="checkbox" id="__drawer" autocomplete="off">
  20. <input class="md-toggle" data-md-toggle="search" type="checkbox" id="__search" autocomplete="off">
  21. <label class="md-overlay" for="__drawer"></label>
  22. <div data-md-component="skip">
  23. <a href="#_1" class="md-skip">
  24. 跳转至
  25. </a>
  26. </div>
  27. <div data-md-component="announce">
  28. </div>
  29. <header class="md-header md-header--shadow" data-md-component="header">
  30. <nav class="md-header__inner md-grid" aria-label="页眉">
  31. <a href="../index.html" title="企业知识库RAG智能问答项目V1.0" class="md-header__button md-logo" aria-label="企业知识库RAG智能问答项目V1.0" data-md-component="logo">
  32. <img src="../img/AI.jpg" alt="logo">
  33. </a>
  34. <label class="md-header__button md-icon" for="__drawer">
  35. <svg xmlns="http://www.w3.org/2000/svg" viewbox="0 0 24 24"><path d="M3 6h18v2H3zm0 5h18v2H3zm0 5h18v2H3z"></path></svg>
  36. </label>
  37. <div class="md-header__title" data-md-component="header-title">
  38. <div class="md-header__ellipsis">
  39. <div class="md-header__topic">
  40. <span class="md-ellipsis">
  41. 企业知识库RAG智能问答项目V1.0
  42. </span>
  43. </div>
  44. <div class="md-header__topic" data-md-component="header-topic">
  45. <span class="md-ellipsis">
  46. 4.3 文档处理模块
  47. </span>
  48. </div>
  49. </div>
  50. </div>
  51. <script>var palette=__md_get("__palette");if(palette&&palette.color){if("(prefers-color-scheme)"===palette.color.media){var media=matchMedia("(prefers-color-scheme: light)"),input=document.querySelector(media.matches?"[data-md-color-media='(prefers-color-scheme: light)']":"[data-md-color-media='(prefers-color-scheme: dark)']");palette.color.media=input.getAttribute("data-md-color-media"),palette.color.scheme=input.getAttribute("data-md-color-scheme"),palette.color.primary=input.getAttribute("data-md-color-primary"),palette.color.accent=input.getAttribute("data-md-color-accent")}for(var[key,value]of Object.entries(palette.color))document.body.setAttribute("data-md-color-"+key,value)}</script>
  52. <label class="md-header__button md-icon" for="__search">
  53. <svg xmlns="http://www.w3.org/2000/svg" viewbox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"></path></svg>
  54. </label>
  55. <div class="md-search" data-md-component="search" role="dialog">
  56. <label class="md-search__overlay" for="__search"></label>
  57. <div class="md-search__inner" role="search">
  58. <form class="md-search__form" name="search">
  59. <input type="text" class="md-search__input" name="query" aria-label="搜索" placeholder="搜索" autocapitalize="off" autocorrect="off" autocomplete="off" spellcheck="false" data-md-component="search-query" required>
  60. <label class="md-search__icon md-icon" for="__search">
  61. <svg xmlns="http://www.w3.org/2000/svg" viewbox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"></path></svg>
  62. <svg xmlns="http://www.w3.org/2000/svg" viewbox="0 0 24 24"><path d="M20 11v2H8l5.5 5.5-1.42 1.42L4.16 12l7.92-7.92L13.5 5.5 8 11z"></path></svg>
  63. </label>
  64. <nav class="md-search__options" aria-label="查找">
  65. <button type="reset" class="md-search__icon md-icon" title="清空当前内容" aria-label="清空当前内容" tabindex="-1">
  66. <svg xmlns="http://www.w3.org/2000/svg" viewbox="0 0 24 24"><path d="M19 6.41 17.59 5 12 10.59 6.41 5 5 6.41 10.59 12 5 17.59 6.41 19 12 13.41 17.59 19 19 17.59 13.41 12z"></path></svg>
  67. </button>
  68. </nav>
  69. </form>
  70. <div class="md-search__output">
  71. <div class="md-search__scrollwrap" tabindex="0" data-md-scrollfix>
  72. <div class="md-search-result" data-md-component="search-result">
  73. <div class="md-search-result__meta">
  74. 正在初始化搜索引擎
  75. </div>
  76. <ol class="md-search-result__list" role="presentation"></ol>
  77. </div>
  78. </div>
  79. </div>
  80. </div>
  81. </div>
  82. </nav>
  83. </header>
  84. <div class="md-container" data-md-component="container">
  85. <main class="md-main" data-md-component="main">
  86. <div class="md-main__inner md-grid">
  87. <div class="md-sidebar md-sidebar--primary" data-md-component="sidebar" data-md-type="navigation">
  88. <div class="md-sidebar__scrollwrap">
  89. <div class="md-sidebar__inner">
  90. <nav class="md-nav md-nav--primary" aria-label="导航栏" data-md-level="0">
  91. <label class="md-nav__title" for="__drawer">
  92. <a href="../index.html" title="企业知识库RAG智能问答项目V1.0" class="md-nav__button md-logo" aria-label="企业知识库RAG智能问答项目V1.0" data-md-component="logo">
  93. <img src="../img/AI.jpg" alt="logo">
  94. </a>
  95. 企业知识库RAG智能问答项目V1.0
  96. </label>
  97. <ul class="md-nav__list" data-md-scrollfix>
  98. <li class="md-nav__item">
  99. <a href="../index.html" class="md-nav__link">
  100. <span class="md-ellipsis">
  101. 知识库内容
  102. </span>
  103. </a>
  104. </li>
  105. <li class="md-nav__item md-nav__item--nested">
  106. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_2">
  107. <label class="md-nav__link" for="__nav_2" id="__nav_2_label" tabindex="0">
  108. <span class="md-ellipsis">
  109. ① 项目概述
  110. </span>
  111. <span class="md-nav__icon md-icon"></span>
  112. </label>
  113. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_2_label" aria-expanded="false">
  114. <label class="md-nav__title" for="__nav_2">
  115. <span class="md-nav__icon md-icon"></span>
  116. ① 项目概述
  117. </label>
  118. <ul class="md-nav__list" data-md-scrollfix>
  119. <li class="md-nav__item">
  120. <a href="../%E2%91%A0%20%E9%A1%B9%E7%9B%AE%E6%A6%82%E8%BF%B0/1.1%20%E9%A1%B9%E7%9B%AE%E6%9E%B6%E6%9E%84.html" class="md-nav__link">
  121. <span class="md-ellipsis">
  122. 1.1 项目架构
  123. </span>
  124. </a>
  125. </li>
  126. <li class="md-nav__item">
  127. <a href="../%E2%91%A0%20%E9%A1%B9%E7%9B%AE%E6%A6%82%E8%BF%B0/1.2%20%E9%A1%B9%E7%9B%AE%E7%8E%AF%E5%A2%83%28window%29.html" class="md-nav__link">
  128. <span class="md-ellipsis">
  129. 1.2 项目环境(window)
  130. </span>
  131. </a>
  132. </li>
  133. <li class="md-nav__item">
  134. <a href="../%E2%91%A0%20%E9%A1%B9%E7%9B%AE%E6%A6%82%E8%BF%B0/1.3%20%E9%A1%B9%E7%9B%AE%E7%8E%AF%E5%A2%83%28mac%29.html" class="md-nav__link">
  135. <span class="md-ellipsis">
  136. 1.3 项目环境(mac)
  137. </span>
  138. </a>
  139. </li>
  140. </ul>
  141. </nav>
  142. </li>
  143. <li class="md-nav__item md-nav__item--nested">
  144. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_3">
  145. <label class="md-nav__link" for="__nav_3" id="__nav_3_label" tabindex="0">
  146. <span class="md-ellipsis">
  147. ② 项目工具
  148. </span>
  149. <span class="md-nav__icon md-icon"></span>
  150. </label>
  151. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_3_label" aria-expanded="false">
  152. <label class="md-nav__title" for="__nav_3">
  153. <span class="md-nav__icon md-icon"></span>
  154. ② 项目工具
  155. </label>
  156. <ul class="md-nav__list" data-md-scrollfix>
  157. <li class="md-nav__item">
  158. <a href="../%E2%91%A1%20%E9%A1%B9%E7%9B%AE%E5%B7%A5%E5%85%B7/2.1%20Ollama%E5%A4%A7%E6%A8%A1%E5%9E%8B%E9%AB%98%E6%95%88%E7%AE%A1%E7%90%86%E5%B7%A5%E5%85%B7.html" class="md-nav__link">
  159. <span class="md-ellipsis">
  160. 2.1 Ollama大模型高效管理工具
  161. </span>
  162. </a>
  163. </li>
  164. <li class="md-nav__item">
  165. <a href="../%E2%91%A1%20%E9%A1%B9%E7%9B%AE%E5%B7%A5%E5%85%B7/2.2%20LangChain%E5%9F%BA%E7%A1%80%E7%9F%A5%E8%AF%86.md.html" class="md-nav__link">
  166. <span class="md-ellipsis">
  167. 2.2 LangChain基础知识.md
  168. </span>
  169. </a>
  170. </li>
  171. <li class="md-nav__item">
  172. <a href="../%E2%91%A1%20%E9%A1%B9%E7%9B%AE%E5%B7%A5%E5%85%B7/2.3%20%E5%90%91%E9%87%8F%E7%B4%A2%E5%BC%95%EF%BC%9AHNSW%E3%80%81IVF%E4%B8%8EIVF_PQ.html" class="md-nav__link">
  173. <span class="md-ellipsis">
  174. 2.3 向量索引:HNSW、IVF与IVF_PQ
  175. </span>
  176. </a>
  177. </li>
  178. <li class="md-nav__item">
  179. <a href="../%E2%91%A1%20%E9%A1%B9%E7%9B%AE%E5%B7%A5%E5%85%B7/2.4%20Milvus%E5%90%91%E9%87%8F%E6%95%B0%E6%8D%AE%E5%BA%93.html" class="md-nav__link">
  180. <span class="md-ellipsis">
  181. 2.4 Milvus向量数据库
  182. </span>
  183. </a>
  184. </li>
  185. </ul>
  186. </nav>
  187. </li>
  188. <li class="md-nav__item md-nav__item--nested">
  189. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_4">
  190. <label class="md-nav__link" for="__nav_4" id="__nav_4_label" tabindex="0">
  191. <span class="md-ellipsis">
  192. ③ 基于Mysql库的问答系统
  193. </span>
  194. <span class="md-nav__icon md-icon"></span>
  195. </label>
  196. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_4_label" aria-expanded="false">
  197. <label class="md-nav__title" for="__nav_4">
  198. <span class="md-nav__icon md-icon"></span>
  199. ③ 基于Mysql库的问答系统
  200. </label>
  201. <ul class="md-nav__list" data-md-scrollfix>
  202. <li class="md-nav__item">
  203. <a href="../%E2%91%A2%20%E5%9F%BA%E4%BA%8EMysql%E5%BA%93%E7%9A%84%E9%97%AE%E7%AD%94%E7%B3%BB%E7%BB%9F/3.1%20Python%E6%97%A5%E5%BF%97%E4%BB%8B%E7%BB%8D%E4%B8%8E%E5%BA%94%E7%94%A8.html" class="md-nav__link">
  204. <span class="md-ellipsis">
  205. 3.1 Python日志介绍与应用
  206. </span>
  207. </a>
  208. </li>
  209. <li class="md-nav__item">
  210. <a href="../%E2%91%A2%20%E5%9F%BA%E4%BA%8EMysql%E5%BA%93%E7%9A%84%E9%97%AE%E7%AD%94%E7%B3%BB%E7%BB%9F/3.2%20Redis%E6%95%B0%E6%8D%AE%E5%BA%93%E7%AE%80%E4%BB%8B%E4%B8%8E%E4%BD%BF%E7%94%A8.html" class="md-nav__link">
  211. <span class="md-ellipsis">
  212. 3.2 Redis数据库简介与使用
  213. </span>
  214. </a>
  215. </li>
  216. <li class="md-nav__item">
  217. <a href="../%E2%91%A2%20%E5%9F%BA%E4%BA%8EMysql%E5%BA%93%E7%9A%84%E9%97%AE%E7%AD%94%E7%B3%BB%E7%BB%9F/3.3%20%E5%9F%BA%E4%BA%8EMysql%E6%95%B0%E6%8D%AE%E5%BA%93%E5%AE%9E%E7%8E%B0%E9%97%AE%E7%AD%94%E7%B3%BB%E7%BB%9F.html" class="md-nav__link">
  218. <span class="md-ellipsis">
  219. 3.3 基于Mysql数据库实现问答系统
  220. </span>
  221. </a>
  222. </li>
  223. <li class="md-nav__item">
  224. <a href="../%E2%91%A2%20%E5%9F%BA%E4%BA%8EMysql%E5%BA%93%E7%9A%84%E9%97%AE%E7%AD%94%E7%B3%BB%E7%BB%9F/3.4%20BM25%E7%AE%97%E6%B3%95%E7%AE%80%E4%BB%8B%E4%B8%8E%E5%BA%94%E7%94%A8.html" class="md-nav__link">
  225. <span class="md-ellipsis">
  226. 3.4 BM25算法简介与应用
  227. </span>
  228. </a>
  229. </li>
  230. </ul>
  231. </nav>
  232. </li>
  233. <li class="md-nav__item md-nav__item--active md-nav__item--nested">
  234. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_5" checked>
  235. <label class="md-nav__link" for="__nav_5" id="__nav_5_label" tabindex="0">
  236. <span class="md-ellipsis">
  237. ④ 基于Milvus库的问答系统
  238. </span>
  239. <span class="md-nav__icon md-icon"></span>
  240. </label>
  241. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_5_label" aria-expanded="true">
  242. <label class="md-nav__title" for="__nav_5">
  243. <span class="md-nav__icon md-icon"></span>
  244. ④ 基于Milvus库的问答系统
  245. </label>
  246. <ul class="md-nav__list" data-md-scrollfix>
  247. <li class="md-nav__item">
  248. <a href="4.1%20%E6%95%B4%E4%BD%93%E6%9E%B6%E6%9E%84%E4%B8%8E%E5%B7%A5%E7%A8%8B%E6%B5%81%E7%A8%8B.html" class="md-nav__link">
  249. <span class="md-ellipsis">
  250. 4.1 整体架构与工程流程
  251. </span>
  252. </a>
  253. </li>
  254. <li class="md-nav__item">
  255. <a href="4.2%20%E5%9F%BA%E7%A1%80%E6%A8%A1%E5%9D%97.html" class="md-nav__link">
  256. <span class="md-ellipsis">
  257. 4.2 基础模块
  258. </span>
  259. </a>
  260. </li>
  261. <li class="md-nav__item md-nav__item--active">
  262. <input class="md-nav__toggle md-toggle" type="checkbox" id="__toc">
  263. <label class="md-nav__link md-nav__link--active" for="__toc">
  264. <span class="md-ellipsis">
  265. 4.3 文档处理模块
  266. </span>
  267. <span class="md-nav__icon md-icon"></span>
  268. </label>
  269. <a href="4.3%20%E6%96%87%E6%A1%A3%E5%A4%84%E7%90%86%E6%A8%A1%E5%9D%97.html" class="md-nav__link md-nav__link--active">
  270. <span class="md-ellipsis">
  271. 4.3 文档处理模块
  272. </span>
  273. </a>
  274. <nav class="md-nav md-nav--secondary" aria-label="目录">
  275. <label class="md-nav__title" for="__toc">
  276. <span class="md-nav__icon md-icon"></span>
  277. 目录
  278. </label>
  279. <ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
  280. <li class="md-nav__item">
  281. <a href="#_1" class="md-nav__link">
  282. <span class="md-ellipsis">
  283. 学习目标
  284. </span>
  285. </a>
  286. </li>
  287. <li class="md-nav__item">
  288. <a href="#_2" class="md-nav__link">
  289. <span class="md-ellipsis">
  290. 一、文档解析
  291. </span>
  292. </a>
  293. <nav class="md-nav" aria-label="一、文档解析">
  294. <ul class="md-nav__list">
  295. <li class="md-nav__item">
  296. <a href="#1" class="md-nav__link">
  297. <span class="md-ellipsis">
  298. 1 基本介绍
  299. </span>
  300. </a>
  301. </li>
  302. <li class="md-nav__item">
  303. <a href="#2" class="md-nav__link">
  304. <span class="md-ellipsis">
  305. 2 代码实现
  306. </span>
  307. </a>
  308. </li>
  309. </ul>
  310. </nav>
  311. </li>
  312. <li class="md-nav__item">
  313. <a href="#_3" class="md-nav__link">
  314. <span class="md-ellipsis">
  315. 二、说明
  316. </span>
  317. </a>
  318. </li>
  319. <li class="md-nav__item">
  320. <a href="#_4" class="md-nav__link">
  321. <span class="md-ellipsis">
  322. 本章小结
  323. </span>
  324. </a>
  325. </li>
  326. </ul>
  327. </nav>
  328. </li>
  329. <li class="md-nav__item">
  330. <a href="4.4%20%E5%90%91%E9%87%8F%E5%AD%98%E5%82%A8.html" class="md-nav__link">
  331. <span class="md-ellipsis">
  332. 4.4 向量存储
  333. </span>
  334. </a>
  335. </li>
  336. <li class="md-nav__item"><a href="4.5%20BERT%E5%9F%BA%E7%A1%80%E4%B8%8E%E4%BB%BB%E5%8A%A1%E5%BE%AE%E8%B0%83.html" class="md-nav__link"><span class="md-ellipsis">4.5 BERT基础与任务微调</span></a></li><li class="md-nav__item">
  337. <a href="4.6%20prompts%E8%AE%BE%E8%AE%A1%E4%B8%8Equery%E6%84%8F%E5%9B%BE%E8%AF%86%E5%88%AB.html" class="md-nav__link">
  338. <span class="md-ellipsis">
  339. 4.6 prompts设计与query意图识别
  340. </span>
  341. </a>
  342. </li>
  343. <li class="md-nav__item">
  344. <a href="4.7%20%E6%A3%80%E7%B4%A2%E7%AD%96%E7%95%A5%E4%B8%8ERAG%E7%B3%BB%E7%BB%9F%E8%AE%BE%E8%AE%A1.html" class="md-nav__link">
  345. <span class="md-ellipsis">
  346. 4.7 检索策略与RAG系统设计
  347. </span>
  348. </a>
  349. </li>
  350. <li class="md-nav__item">
  351. <a href="4.8%20RAG%E7%B3%BB%E7%BB%9F%E8%BF%90%E8%A1%8C.html" class="md-nav__link">
  352. <span class="md-ellipsis">
  353. 4.8 RAG系统运行
  354. </span>
  355. </a>
  356. </li>
  357. <li class="md-nav__item">
  358. <a href="4.9%20%E6%96%87%E6%A1%A3%E8%A7%A3%E6%9E%90%E5%B7%A5%E5%85%B7%28%E6%89%A9%E5%B1%95%E8%B5%84%E6%96%99%29.html" class="md-nav__link">
  359. <span class="md-ellipsis">
  360. 4.9 文档解析工具(扩展资料)
  361. </span>
  362. </a>
  363. </li>
  364. <li class="md-nav__item">
  365. <a href="4.10%20RAG%E4%B8%AD%E7%9A%84Query%E6%94%B9%E5%86%99%28%E6%89%A9%E5%B1%95%E8%B5%84%E6%96%99%29.html" class="md-nav__link">
  366. <span class="md-ellipsis">
  367. 4.10 RAG中的Query改写(扩展资料)
  368. </span>
  369. </a>
  370. </li>
  371. <li class="md-nav__item">
  372. <a href="%E6%95%B0%E6%8D%AE%E9%9B%86%E7%94%9F%E6%88%90%E4%B8%8E%E4%BC%98%E5%8C%96%28%E6%89%A9%E5%B1%95%E8%B5%84%E6%96%99%29.html" class="md-nav__link">
  373. <span class="md-ellipsis">
  374. 数据集生成与优化(扩展资料)
  375. </span>
  376. </a>
  377. </li>
  378. </ul>
  379. </nav>
  380. </li>
  381. <li class="md-nav__item md-nav__item--nested">
  382. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_6">
  383. <label class="md-nav__link" for="__nav_6" id="__nav_6_label" tabindex="0">
  384. <span class="md-ellipsis">
  385. ⑤ RAG系统评估
  386. </span>
  387. <span class="md-nav__icon md-icon"></span>
  388. </label>
  389. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_6_label" aria-expanded="false">
  390. <label class="md-nav__title" for="__nav_6">
  391. <span class="md-nav__icon md-icon"></span>
  392. ⑤ RAG系统评估
  393. </label>
  394. <ul class="md-nav__list" data-md-scrollfix>
  395. <li class="md-nav__item">
  396. <a href="../%E2%91%A4%20RAG%E7%B3%BB%E7%BB%9F%E8%AF%84%E4%BC%B0/5.1%20RAG%E7%B3%BB%E7%BB%9F%E8%AF%84%E4%BC%B0%E5%B7%A5%E5%85%B7.html" class="md-nav__link">
  397. <span class="md-ellipsis">
  398. 5.1 RAG系统评估工具
  399. </span>
  400. </a>
  401. </li>
  402. <li class="md-nav__item">
  403. <a href="../%E2%91%A4%20RAG%E7%B3%BB%E7%BB%9F%E8%AF%84%E4%BC%B0/5.2%20%E5%AE%9E%E7%8E%B0EnterpriseKB-RAG%E7%B3%BB%E7%BB%9F%E7%9A%84%E8%AF%84%E4%BC%B0.html" class="md-nav__link">
  404. <span class="md-ellipsis">
  405. 5.2 实现企业知识库RAG系统的评估
  406. </span>
  407. </a>
  408. </li>
  409. <li class="md-nav__item">
  410. <a href="../%E2%91%A4%20RAG%E7%B3%BB%E7%BB%9F%E8%AF%84%E4%BC%B0/5.3%20RAG%E8%AF%84%E4%BC%B0%E6%95%B0%E6%8D%AE%E9%9B%86%E7%9A%84%E6%9E%84%E9%80%A0%28%E6%89%A9%E5%B1%95%29.html" class="md-nav__link">
  411. <span class="md-ellipsis">
  412. 5.3 RAG评估数据集的构造(扩展)
  413. </span>
  414. </a>
  415. </li>
  416. </ul>
  417. </nav>
  418. </li>
  419. <li class="md-nav__item md-nav__item--nested">
  420. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_7">
  421. <label class="md-nav__link" for="__nav_7" id="__nav_7_label" tabindex="0">
  422. <span class="md-ellipsis">
  423. ⑥ 融合Mysql的RAG系统
  424. </span>
  425. <span class="md-nav__icon md-icon"></span>
  426. </label>
  427. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_7_label" aria-expanded="false">
  428. <label class="md-nav__title" for="__nav_7">
  429. <span class="md-nav__icon md-icon"></span>
  430. ⑥ 融合Mysql的RAG系统
  431. </label>
  432. <ul class="md-nav__list" data-md-scrollfix>
  433. <li class="md-nav__item">
  434. <a href="../%E2%91%A5%20%E8%9E%8D%E5%90%88Mysql%E7%9A%84RAG%E7%B3%BB%E7%BB%9F/6.1%20%E8%9E%8D%E5%90%88FQA%E5%92%8C%E7%9F%A5%E8%AF%86%E5%BA%93%E6%9F%A5%E8%AF%A2.html" class="md-nav__link">
  435. <span class="md-ellipsis">
  436. 6.1 融合FQA和知识库查询
  437. </span>
  438. </a>
  439. </li>
  440. <li class="md-nav__item">
  441. <a href="../%E2%91%A5%20%E8%9E%8D%E5%90%88Mysql%E7%9A%84RAG%E7%B3%BB%E7%BB%9F/6.2%20%E8%9E%8D%E5%90%88FAQ%20%E5%92%8C%E7%9F%A5%E8%AF%86%E5%BA%93%E6%9F%A5%E8%AF%A2%EF%BC%88%E4%BC%98%E5%8C%96%E7%89%88%EF%BC%89.html" class="md-nav__link">
  442. <span class="md-ellipsis">
  443. 6.2 融合FAQ 和知识库查询(优化版)
  444. </span>
  445. </a>
  446. </li>
  447. <li class="md-nav__item">
  448. <a href="../%E2%91%A5%20%E8%9E%8D%E5%90%88Mysql%E7%9A%84RAG%E7%B3%BB%E7%BB%9F/6.3%20%E5%9F%BA%E4%BA%8E%20FastAPI%20%E7%9A%84%20FAQ%20%E4%B8%8E%E7%9F%A5%E8%AF%86%E5%BA%93%E6%9F%A5%E8%AF%A2%E6%8E%A5%E5%8F%A3.html" class="md-nav__link">
  449. <span class="md-ellipsis">
  450. 6.3 基于 FastAPI 的 FAQ 与知识库查询接口
  451. </span>
  452. </a>
  453. </li>
  454. <li class="md-nav__item">
  455. <a href="../%E2%91%A5%20%E8%9E%8D%E5%90%88Mysql%E7%9A%84RAG%E7%B3%BB%E7%BB%9F/6.4%20FastAPI%E5%92%8CFlask%E7%9A%84%E5%8C%BA%E5%88%AB%E5%92%8C%E8%81%94%E7%B3%BB%28%E6%89%A9%E5%B1%95%E8%B5%84%E6%96%99%29.html" class="md-nav__link">
  456. <span class="md-ellipsis">
  457. 6.4 FastAPI和Flask的区别和联系(扩展资料)
  458. </span>
  459. </a>
  460. </li>
  461. </ul>
  462. </nav>
  463. </li>
  464. <li class="md-nav__item md-nav__item--nested">
  465. <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_8">
  466. <label class="md-nav__link" for="__nav_8" id="__nav_8_label" tabindex="0">
  467. <span class="md-ellipsis">
  468. ⑦ 企业级生产环境部署
  469. </span>
  470. <span class="md-nav__icon md-icon"></span>
  471. </label>
  472. <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_8_label" aria-expanded="false">
  473. <label class="md-nav__title" for="__nav_8">
  474. <span class="md-nav__icon md-icon"></span>
  475. ⑦ 企业级生产环境部署
  476. </label>
  477. <ul class="md-nav__list" data-md-scrollfix>
  478. <li class="md-nav__item">
  479. <a href="../%E2%91%A6%20%E4%BC%81%E4%B8%9A%E7%BA%A7%E7%94%9F%E4%BA%A7%E7%8E%AF%E5%A2%83%E9%83%A8%E7%BD%B2/7.1%20Docker%E7%9A%84%E5%8E%9F%E7%90%86%E5%92%8C%E5%9F%BA%E6%9C%AC%E4%BD%BF%E7%94%A8.html" class="md-nav__link">
  480. <span class="md-ellipsis">
  481. 7.1 Docker的原理和基本使用
  482. </span>
  483. </a>
  484. </li>
  485. <li class="md-nav__item">
  486. <a href="../%E2%91%A6%20%E4%BC%81%E4%B8%9A%E7%BA%A7%E7%94%9F%E4%BA%A7%E7%8E%AF%E5%A2%83%E9%83%A8%E7%BD%B2/7.2%20%E6%9C%8D%E5%8A%A1%E9%83%A8%E7%BD%B2.html" class="md-nav__link">
  487. <span class="md-ellipsis">
  488. 7.2 服务部署
  489. </span>
  490. </a>
  491. </li>
  492. </ul>
  493. </nav>
  494. </li>
  495. </ul>
  496. </nav>
  497. </div>
  498. </div>
  499. </div>
  500. <div class="md-sidebar md-sidebar--secondary" data-md-component="sidebar" data-md-type="toc">
  501. <div class="md-sidebar__scrollwrap">
  502. <div class="md-sidebar__inner">
  503. <nav class="md-nav md-nav--secondary" aria-label="目录">
  504. <label class="md-nav__title" for="__toc">
  505. <span class="md-nav__icon md-icon"></span>
  506. 目录
  507. </label>
  508. <ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
  509. <li class="md-nav__item">
  510. <a href="#_1" class="md-nav__link">
  511. <span class="md-ellipsis">
  512. 学习目标
  513. </span>
  514. </a>
  515. </li>
  516. <li class="md-nav__item">
  517. <a href="#_2" class="md-nav__link">
  518. <span class="md-ellipsis">
  519. 一、文档解析
  520. </span>
  521. </a>
  522. <nav class="md-nav" aria-label="一、文档解析">
  523. <ul class="md-nav__list">
  524. <li class="md-nav__item">
  525. <a href="#1" class="md-nav__link">
  526. <span class="md-ellipsis">
  527. 1 基本介绍
  528. </span>
  529. </a>
  530. </li>
  531. <li class="md-nav__item">
  532. <a href="#2" class="md-nav__link">
  533. <span class="md-ellipsis">
  534. 2 代码实现
  535. </span>
  536. </a>
  537. </li>
  538. </ul>
  539. </nav>
  540. </li>
  541. <li class="md-nav__item">
  542. <a href="#_3" class="md-nav__link">
  543. <span class="md-ellipsis">
  544. 二、说明
  545. </span>
  546. </a>
  547. </li>
  548. <li class="md-nav__item">
  549. <a href="#_4" class="md-nav__link">
  550. <span class="md-ellipsis">
  551. 本章小结
  552. </span>
  553. </a>
  554. </li>
  555. </ul>
  556. </nav>
  557. </div>
  558. </div>
  559. </div>
  560. <div class="md-content" data-md-component="content">
  561. <article class="md-content__inner md-typeset">
  562. <h1>4.3 文档处理模块</h1>
  563. <h2 id="_1">学习目标<a class="headerlink" href="#_1" title="Permanent link">¶</a></h2>
  564. <ul>
  565. <li>1.了解不通类型文档处理的基本逻辑。</li>
  566. <li>2.掌握文档加载和分块的基本原理。</li>
  567. </ul>
  568. <hr>
  569. <h2 id="_2">一、文档解析<a class="headerlink" href="#_2" title="Permanent link">¶</a></h2>
  570. <h3 id="1">1 基本介绍<a class="headerlink" href="#1" title="Permanent link">¶</a></h3>
  571. <p><code>document_processor.py</code>是企业知识库RAG系统的核心模块之一,用于文档解析。</p>
  572. <p>主要负责加载多种格式的文档(如<code>.txt</code>、<code>.pdf</code>等),并对其进行分层切分,生成父块和子块,为后续的向量存储和检索做好准备。</p>
  573. <h3 id="2">2 代码实现<a class="headerlink" href="#2" title="Permanent link">¶</a></h3>
  574. <div class="highlight"><pre><span></span><code><span class="c1"># core/document_processor.py</span>
  575. <span class="kn">import</span><span class="w"> </span><span class="nn">os</span>
  576. <span class="kn">from</span><span class="w"> </span><span class="nn">langchain_community.document_loaders</span><span class="w"> </span><span class="kn">import</span> <span class="n">TextLoader</span>
  577. <span class="kn">from</span><span class="w"> </span><span class="nn">langchain_community.document_loaders.markdown</span><span class="w"> </span><span class="kn">import</span> <span class="n">UnstructuredMarkdownLoader</span>
  578. <span class="kn">from</span><span class="w"> </span><span class="nn">langchain_text_splitters</span><span class="w"> </span><span class="kn">import</span> <span class="n">MarkdownTextSplitter</span>
  579. <span class="kn">from</span><span class="w"> </span><span class="nn">datetime</span><span class="w"> </span><span class="kn">import</span> <span class="n">datetime</span>
  580. <span class="kn">from</span><span class="w"> </span><span class="nn">rag_qa.kb_text_spliter</span><span class="w"> </span><span class="kn">import</span> <span class="n">ChineseRecursiveTextSplitter</span>
  581. <span class="kn">from</span><span class="w"> </span><span class="nn">rag_qa.kb_text_spliter</span><span class="w"> </span><span class="kn">import</span> <span class="n">AliTextSplitter</span>
  582. <span class="kn">from</span><span class="w"> </span><span class="nn">rag_qa.kb_document_loaders</span><span class="w"> </span><span class="kn">import</span> <span class="n">OCRPDFLoader</span><span class="p">,</span> <span class="n">OCRDOCLoader</span><span class="p">,</span> <span class="n">OCRPPTLoader</span><span class="p">,</span> <span class="n">OCRIMGLoader</span>
  583. <span class="kn">from</span><span class="w"> </span><span class="nn">base.config</span><span class="w"> </span><span class="kn">import</span> <span class="n">config</span>
  584. <span class="kn">from</span><span class="w"> </span><span class="nn">base.logger</span><span class="w"> </span><span class="kn">import</span> <span class="n">logger</span>
  585. <span class="n">conf</span> <span class="o">=</span> <span class="n">Config</span><span class="p">()</span>
  586. <span class="c1"># 定义支持的文件类型及其对应的加载器字典</span>
  587. <span class="n">document_loaders</span> <span class="o">=</span> <span class="p">{</span>
  588. <span class="c1"># 文本文件使用 TextLoader</span>
  589. <span class="s2">".txt"</span><span class="p">:</span> <span class="n">TextLoader</span><span class="p">,</span>
  590. <span class="c1"># PDF 文件使用 OCRPDFLoader</span>
  591. <span class="s2">".pdf"</span><span class="p">:</span> <span class="n">OCRPDFLoader</span><span class="p">,</span>
  592. <span class="c1"># Word 文件使用 OCRDOCLoader</span>
  593. <span class="s2">".docx"</span><span class="p">:</span> <span class="n">OCRDOCLoader</span><span class="p">,</span>
  594. <span class="c1"># PPT 文件使用 OCRPPTLoader</span>
  595. <span class="s2">".ppt"</span><span class="p">:</span> <span class="n">OCRPPTLoader</span><span class="p">,</span>
  596. <span class="c1"># PPTX 文件使用 OCRPPTLoader</span>
  597. <span class="s2">".pptx"</span><span class="p">:</span> <span class="n">OCRPPTLoader</span><span class="p">,</span>
  598. <span class="c1"># JPG 文件使用 OCRIMGLoader</span>
  599. <span class="s2">".jpg"</span><span class="p">:</span> <span class="n">OCRIMGLoader</span><span class="p">,</span>
  600. <span class="c1"># PNG 文件使用 OCRIMGLoader</span>
  601. <span class="s2">".png"</span><span class="p">:</span> <span class="n">OCRIMGLoader</span><span class="p">,</span>
  602. <span class="c1"># Markdown 文件使用 UnstructuredMarkdownLoader</span>
  603. <span class="s2">".md"</span><span class="p">:</span> <span class="n">UnstructuredMarkdownLoader</span>
  604. <span class="p">}</span>
  605. <span class="c1"># 定义函数,从指定文件夹加载多种类型文件并添加元数据</span>
  606. <span class="k">def</span><span class="w"> </span><span class="nf">load_documents_from_directory</span><span class="p">(</span><span class="n">directory_path</span><span class="p">):</span>
  607. <span class="c1"># 初始化空列表,用于存储加载的文档</span>
  608. <span class="n">documents</span> <span class="o">=</span> <span class="p">[]</span>
  609. <span class="c1"># 获取支持的文件扩展名集合</span>
  610. <span class="n">supported_extensions</span> <span class="o">=</span> <span class="n">document_loaders</span><span class="o">.</span><span class="n">keys</span><span class="p">()</span>
  611. <span class="c1"># 从目录名提取知识域(如 "enterprise_data" -&gt; "enterprise")</span>
  612. <span class="n">source</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">basename</span><span class="p">(</span><span class="n">directory_path</span><span class="p">)</span><span class="o">.</span><span class="n">replace</span><span class="p">(</span><span class="s2">"_data"</span><span class="p">,</span> <span class="s2">""</span><span class="p">)</span>
  613. <span class="c1"># 遍历指定目录及其子目录</span>
  614. <span class="k">for</span> <span class="n">root</span><span class="p">,</span> <span class="n">_</span><span class="p">,</span> <span class="n">files</span> <span class="ow">in</span> <span class="n">os</span><span class="o">.</span><span class="n">walk</span><span class="p">(</span><span class="n">directory_path</span><span class="p">):</span>
  615. <span class="c1"># 遍历当前目录下的所有文件</span>
  616. <span class="k">for</span> <span class="n">file</span> <span class="ow">in</span> <span class="n">files</span><span class="p">:</span>
  617. <span class="c1"># 构造文件的完整路径</span>
  618. <span class="n">file_path</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">root</span><span class="p">,</span> <span class="n">file</span><span class="p">)</span>
  619. <span class="c1"># 获取文件扩展名并转换为小写</span>
  620. <span class="n">file_extension</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">splitext</span><span class="p">(</span><span class="n">file_path</span><span class="p">)[</span><span class="mi">1</span><span class="p">]</span><span class="o">.</span><span class="n">lower</span><span class="p">()</span>
  621. <span class="c1"># 检查文件类型是否在支持的扩展名列表中</span>
  622. <span class="k">if</span> <span class="n">file_extension</span> <span class="ow">in</span> <span class="n">supported_extensions</span><span class="p">:</span>
  623. <span class="c1"># 使用 try-except 捕获加载过程中的异常</span>
  624. <span class="k">try</span><span class="p">:</span>
  625. <span class="c1"># 根据文件扩展名获取对应的加载器类</span>
  626. <span class="n">loader_class</span> <span class="o">=</span> <span class="n">document_loaders</span><span class="p">[</span><span class="n">file_extension</span><span class="p">]</span>
  627. <span class="c1"># 实例化加载器对象,传入文件路径</span>
  628. <span class="k">if</span> <span class="n">file_extension</span> <span class="o">==</span> <span class="s2">".txt"</span><span class="p">:</span>
  629. <span class="n">loader</span> <span class="o">=</span> <span class="n">loader_class</span><span class="p">(</span><span class="n">file_path</span><span class="p">,</span> <span class="n">encoding</span><span class="o">=</span><span class="s2">"utf-8"</span><span class="p">)</span>
  630. <span class="k">else</span><span class="p">:</span>
  631. <span class="n">loader</span> <span class="o">=</span> <span class="n">loader_class</span><span class="p">(</span><span class="n">file_path</span><span class="p">)</span>
  632. <span class="c1"># 调用加载器加载文档内容,返回文档列表</span>
  633. <span class="n">loaded_docs</span> <span class="o">=</span> <span class="n">loader</span><span class="o">.</span><span class="n">load</span><span class="p">()</span>
  634. <span class="c1"># 遍历加载的每个文档</span>
  635. <span class="k">for</span> <span class="n">doc</span> <span class="ow">in</span> <span class="n">loaded_docs</span><span class="p">:</span>
  636. <span class="c1"># 为文档添加知识域元数据</span>
  637. <span class="n">doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"source"</span><span class="p">]</span> <span class="o">=</span> <span class="n">source</span>
  638. <span class="c1"># 为文档添加文件路径元数据</span>
  639. <span class="n">doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"file_path"</span><span class="p">]</span> <span class="o">=</span> <span class="n">file_path</span>
  640. <span class="c1"># 为文档添加当前时间戳元数据</span>
  641. <span class="n">doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"timestamp"</span><span class="p">]</span> <span class="o">=</span> <span class="n">datetime</span><span class="o">.</span><span class="n">now</span><span class="p">()</span><span class="o">.</span><span class="n">isoformat</span><span class="p">()</span>
  642. <span class="c1"># 将加载的文档添加到总列表中</span>
  643. <span class="n">documents</span><span class="o">.</span><span class="n">extend</span><span class="p">(</span><span class="n">loaded_docs</span><span class="p">)</span>
  644. <span class="c1"># 记录成功加载文件的日志</span>
  645. <span class="n">logger</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">"成功加载文件: </span><span class="si">{</span><span class="n">file_path</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  646. <span class="c1"># 捕获加载过程中可能出现的异常</span>
  647. <span class="k">except</span> <span class="ne">Exception</span> <span class="k">as</span> <span class="n">e</span><span class="p">:</span>
  648. <span class="c1"># 记录加载失败的日志,包含错误信息</span>
  649. <span class="n">logger</span><span class="o">.</span><span class="n">error</span><span class="p">(</span><span class="sa">f</span><span class="s2">"加载文件 </span><span class="si">{</span><span class="n">file_path</span><span class="si">}</span><span class="s2"> 失败: </span><span class="si">{</span><span class="nb">str</span><span class="p">(</span><span class="n">e</span><span class="p">)</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  650. <span class="c1"># 如果文件类型不在支持列表中</span>
  651. <span class="k">else</span><span class="p">:</span>
  652. <span class="c1"># 记录警告日志,提示不支持的文件类型</span>
  653. <span class="n">logger</span><span class="o">.</span><span class="n">warning</span><span class="p">(</span><span class="sa">f</span><span class="s2">"不支持的文件类型: </span><span class="si">{</span><span class="n">file_path</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  654. <span class="c1"># 返回加载的所有文档列表</span>
  655. <span class="k">return</span> <span class="n">documents</span>
  656. <span class="c1"># 定义函数,处理文档并进行分层切分,返回子块结果</span>
  657. <span class="k">def</span><span class="w"> </span><span class="nf">process_documents</span><span class="p">(</span><span class="n">directory_path</span><span class="p">,</span> <span class="n">parent_chunk_size</span><span class="o">=</span><span class="n">conf</span><span class="o">.</span><span class="n">PARENT_CHUNK_SIZE</span><span class="p">,</span>
  658. <span class="n">child_chunk_size</span><span class="o">=</span><span class="n">conf</span><span class="o">.</span><span class="n">CHILD_CHUNK_SIZE</span><span class="p">,</span>
  659. <span class="n">chunk_overlap</span><span class="o">=</span><span class="n">conf</span><span class="o">.</span><span class="n">CHUNK_OVERLAP</span><span class="p">):</span>
  660. <span class="c1"># 从指定目录加载所有文档</span>
  661. <span class="n">documents</span> <span class="o">=</span> <span class="n">load_documents_from_directory</span><span class="p">(</span><span class="n">directory_path</span><span class="p">)</span>
  662. <span class="c1"># 记录加载的文档总数日志</span>
  663. <span class="n">logger</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">"加载的文档数量: </span><span class="si">{</span><span class="nb">len</span><span class="p">(</span><span class="n">documents</span><span class="p">)</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  664. <span class="c1"># 初始化父块和子块分词器(通用)</span>
  665. <span class="n">parent_splitter</span> <span class="o">=</span> <span class="n">ChineseRecursiveTextSplitter</span><span class="p">(</span><span class="n">chunk_size</span><span class="o">=</span><span class="n">parent_chunk_size</span><span class="p">,</span> <span class="n">chunk_overlap</span><span class="o">=</span><span class="n">chunk_overlap</span><span class="p">)</span>
  666. <span class="n">child_splitter</span> <span class="o">=</span> <span class="n">ChineseRecursiveTextSplitter</span><span class="p">(</span><span class="n">chunk_size</span><span class="o">=</span><span class="n">child_chunk_size</span><span class="p">,</span> <span class="n">chunk_overlap</span><span class="o">=</span><span class="n">chunk_overlap</span><span class="p">)</span>
  667. <span class="c1"># 初始化 Markdown 专用分词器</span>
  668. <span class="n">markdown_parent_splitter</span> <span class="o">=</span> <span class="n">MarkdownTextSplitter</span><span class="p">(</span><span class="n">chunk_size</span><span class="o">=</span><span class="n">parent_chunk_size</span><span class="p">,</span> <span class="n">chunk_overlap</span><span class="o">=</span><span class="n">chunk_overlap</span><span class="p">)</span>
  669. <span class="n">markdown_child_splitter</span> <span class="o">=</span> <span class="n">MarkdownTextSplitter</span><span class="p">(</span><span class="n">chunk_size</span><span class="o">=</span><span class="n">child_chunk_size</span><span class="p">,</span> <span class="n">chunk_overlap</span><span class="o">=</span><span class="n">chunk_overlap</span><span class="p">)</span>
  670. <span class="c1"># 初始化空列表,用于存储所有子块</span>
  671. <span class="n">child_chunks</span> <span class="o">=</span> <span class="p">[]</span>
  672. <span class="c1"># 遍历每个原始文档,带上索引 i</span>
  673. <span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="n">doc</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">documents</span><span class="p">):</span>
  674. <span class="c1"># print(doc)</span>
  675. <span class="c1"># 获取文件扩展名</span>
  676. <span class="n">file_extension</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">splitext</span><span class="p">(</span><span class="n">doc</span><span class="o">.</span><span class="n">metadata</span><span class="o">.</span><span class="n">get</span><span class="p">(</span><span class="s2">"file_path"</span><span class="p">,</span> <span class="s2">""</span><span class="p">))[</span><span class="mi">1</span><span class="p">]</span><span class="o">.</span><span class="n">lower</span><span class="p">()</span>
  677. <span class="c1"># 选择切分器</span>
  678. <span class="n">is_markdown</span> <span class="o">=</span> <span class="p">(</span><span class="n">file_extension</span> <span class="o">==</span> <span class="s2">".md"</span><span class="p">)</span>
  679. <span class="n">parent_splitter_to_use</span> <span class="o">=</span> <span class="n">markdown_parent_splitter</span> <span class="k">if</span> <span class="n">is_markdown</span> <span class="k">else</span> <span class="n">parent_splitter</span>
  680. <span class="c1"># print(f'parent_splitter_to_use--&gt;{parent_splitter_to_use}')</span>
  681. <span class="n">child_splitter_to_use</span> <span class="o">=</span> <span class="n">markdown_child_splitter</span> <span class="k">if</span> <span class="n">is_markdown</span> <span class="k">else</span> <span class="n">child_splitter</span>
  682. <span class="n">logger</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">"处理文档: </span><span class="si">{</span><span class="n">doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s1">'file_path'</span><span class="p">]</span><span class="si">}</span><span class="s2">, 使用切分器: </span><span class="si">{</span><span class="s1">'Markdown'</span><span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="n">is_markdown</span><span class="w"> </span><span class="k">else</span><span class="w"> </span><span class="s1">'ChineseRecursive'</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  683. <span class="c1"># 使用父块分词器将文档切分为父块</span>
  684. <span class="n">parent_docs</span> <span class="o">=</span> <span class="n">parent_splitter_to_use</span><span class="o">.</span><span class="n">split_documents</span><span class="p">([</span><span class="n">doc</span><span class="p">])</span>
  685. <span class="c1"># 遍历每个父块,带上索引 j</span>
  686. <span class="k">for</span> <span class="n">j</span><span class="p">,</span> <span class="n">parent_doc</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">parent_docs</span><span class="p">):</span>
  687. <span class="c1"># 为父块生成唯一 ID,格式为 "doc_i_parent_j"</span>
  688. <span class="n">parent_id</span> <span class="o">=</span> <span class="sa">f</span><span class="s2">"doc_</span><span class="si">{</span><span class="n">i</span><span class="si">}</span><span class="s2">_parent_</span><span class="si">{</span><span class="n">j</span><span class="si">}</span><span class="s2">"</span>
  689. <span class="c1"># 将父块 ID 添加到元数据</span>
  690. <span class="n">parent_doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"parent_id"</span><span class="p">]</span> <span class="o">=</span> <span class="n">parent_id</span>
  691. <span class="c1"># 将父块内容存储到元数据</span>
  692. <span class="n">parent_doc</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"parent_content"</span><span class="p">]</span> <span class="o">=</span> <span class="n">parent_doc</span><span class="o">.</span><span class="n">page_content</span>
  693. <span class="c1"># 使用子块分词器将父块切分为子块</span>
  694. <span class="n">sub_chunks</span> <span class="o">=</span> <span class="n">child_splitter_to_use</span><span class="o">.</span><span class="n">split_documents</span><span class="p">([</span><span class="n">parent_doc</span><span class="p">])</span>
  695. <span class="c1"># 遍历每个子块,带上索引 k</span>
  696. <span class="k">for</span> <span class="n">k</span><span class="p">,</span> <span class="n">sub_chunk</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">sub_chunks</span><span class="p">):</span>
  697. <span class="c1"># 为子块添加父块 ID 到元数据</span>
  698. <span class="n">sub_chunk</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"parent_id"</span><span class="p">]</span> <span class="o">=</span> <span class="n">parent_id</span>
  699. <span class="c1"># 为子块添加父块内容到元数据</span>
  700. <span class="n">sub_chunk</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"parent_content"</span><span class="p">]</span> <span class="o">=</span> <span class="n">parent_doc</span><span class="o">.</span><span class="n">page_content</span>
  701. <span class="c1"># 为子块生成唯一 ID,格式为 "parent_id_child_k"</span>
  702. <span class="n">sub_chunk</span><span class="o">.</span><span class="n">metadata</span><span class="p">[</span><span class="s2">"id"</span><span class="p">]</span> <span class="o">=</span> <span class="sa">f</span><span class="s2">"</span><span class="si">{</span><span class="n">parent_id</span><span class="si">}</span><span class="s2">_child_</span><span class="si">{</span><span class="n">k</span><span class="si">}</span><span class="s2">"</span>
  703. <span class="c1"># 将子块添加到子块列表中</span>
  704. <span class="n">child_chunks</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">sub_chunk</span><span class="p">)</span>
  705. <span class="c1"># 记录子块总数日志</span>
  706. <span class="n">logger</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">"子块数量: </span><span class="si">{</span><span class="nb">len</span><span class="p">(</span><span class="n">child_chunks</span><span class="p">)</span><span class="si">}</span><span class="s2">"</span><span class="p">)</span>
  707. <span class="c1"># 返回所有子块列表</span>
  708. <span class="k">return</span> <span class="n">child_chunks</span>
  709. <span class="k">if</span> <span class="vm">__name__</span> <span class="o">==</span> <span class="s1">'__main__'</span><span class="p">:</span>
  710. <span class="n">chunks</span> <span class="o">=</span> <span class="n">process_documents</span><span class="p">(</span>
  711. <span class="s1">'/app/data/enterprise_data'</span><span class="p">,</span>
  712. <span class="n">conf</span><span class="o">.</span><span class="n">PARENT_CHUNK_SIZE</span><span class="p">,</span>
  713. <span class="n">conf</span><span class="o">.</span><span class="n">CHILD_CHUNK_SIZE</span><span class="p">,</span>
  714. <span class="n">conf</span><span class="o">.</span><span class="n">CHUNK_OVERLAP</span><span class="p">,</span>
  715. <span class="p">)</span>
  716. <span class="nb">print</span><span class="p">(</span><span class="n">chunks</span><span class="p">)</span>
  717. </code></pre></div>
  718. <h2 id="_3">二、说明<a class="headerlink" href="#_3" title="Permanent link">¶</a></h2>
  719. <ul>
  720. <li><strong>文档加载</strong>:支持多种格式(如<code>.txt</code>、<code>.pdf</code>),使用专用加载器处理复杂文档。</li>
  721. <li><strong>分层切分</strong>:采用<code>ChineseRecursiveTextSplitter</code>生成父块和子块,优化中文文本处理。</li>
  722. <li><strong>元数据管理</strong>:为每个块添加唯一ID、来源和时间戳,便于检索和溯源。</li>
  723. </ul>
  724. <h2 id="_4">本章小结<a class="headerlink" href="#_4" title="Permanent link">¶</a></h2>
  725. <p><code>document_processor</code>模块为企业知识库RAG系统提供了以下核心支持:</p>
  726. <ul>
  727. <li><strong>文档处理</strong>:实现多格式文档的高效加载和分块。</li>
  728. <li><strong>图片或表格</strong>:采用paddleOCR实现高效识别。</li>
  729. </ul>
  730. </article>
  731. </div>
  732. <script>var target=document.getElementById(location.hash.slice(1));target&&target.name&&(target.checked=target.name.startsWith("__tabbed_"))</script>
  733. </div>
  734. </main>
  735. <footer class="md-footer">
  736. <div class="md-footer-meta md-typeset">
  737. <div class="md-footer-meta__inner md-grid">
  738. <div class="md-copyright">
  739. <div class="md-copyright__highlight">
  740. © EnterpriseKB-RAG 2026. Internal knowledge base demo site.
  741. </div>
  742. Made with
  743. <a href="https://squidfunk.github.io/mkdocs-material/" target="_blank" rel="noopener">
  744. Material for MkDocs
  745. </a>
  746. </div>
  747. </div>
  748. </div>
  749. </footer>
  750. </div>
  751. <div class="md-dialog" data-md-component="dialog">
  752. <div class="md-dialog__inner md-typeset"></div>
  753. </div>
  754. <script id="__config" type="application/json">{"base": "..", "features": [], "search": "../assets/javascripts/workers/search.973d3a69.min.js", "tags": null, "translations": {"clipboard.copied": "\u5df2\u590d\u5236", "clipboard.copy": "\u590d\u5236", "search.result.more.one": "\u5728\u8be5\u9875\u4e0a\u8fd8\u6709 1 \u4e2a\u7b26\u5408\u6761\u4ef6\u7684\u7ed3\u679c", "search.result.more.other": "\u5728\u8be5\u9875\u4e0a\u8fd8\u6709 # \u4e2a\u7b26\u5408\u6761\u4ef6\u7684\u7ed3\u679c", "search.result.none": "\u6ca1\u6709\u627e\u5230\u7b26\u5408\u6761\u4ef6\u7684\u7ed3\u679c", "search.result.one": "\u627e\u5230 1 \u4e2a\u7b26\u5408\u6761\u4ef6\u7684\u7ed3\u679c", "search.result.other": "# \u4e2a\u7b26\u5408\u6761\u4ef6\u7684\u7ed3\u679c", "search.result.placeholder": "\u952e\u5165\u4ee5\u5f00\u59cb\u641c\u7d22", "search.result.term.missing": "\u7f3a\u5c11", "select.version": "\u9009\u62e9\u5f53\u524d\u7248\u672c"}, "version": null}</script>
  755. <script src="../assets/javascripts/bundle.f55a23d4.min.js"></script>
  756. <script src="../javascripts/extra.js"></script>
  757. <script src="../javascripts/mathjax.js"></script>
  758. <!-- <script src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script>-->
  759. <script src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
  760. </body>
  761. </html>