fix: fix indexing content between h1 and h2

Closes #154, #109
easyops-cn · May 19, 2022 · f38b307 · f38b307
1 parent f3c55d4
commit f38b307
Show file tree

Hide file tree

Showing 5 changed files with 71 additions and 37 deletions.
diff --git a/package.json b/package.json
@@ -27,6 +27,7 @@
   "license": "MIT",
   "dependencies": {
     "@docusaurus/utils": "^2.0.0-beta.4",
+    "@docusaurus/utils-common": "^2.0.0-beta.20",
     "@docusaurus/utils-validation": "^2.0.0-beta.4",
     "@easyops-cn/autocomplete.js": "^0.38.1",
     "@node-rs/jieba": "^1.6.0",
@@ -38,7 +39,7 @@
     "lunr": "^2.3.9",
     "lunr-languages": "^1.4.0",
     "mark.js": "^8.11.1",
-    "tslib": "^2.2.0"
+    "tslib": "^2.4.0"
   },
   "devDependencies": {
     "@babel/core": "^7.12.3",
@@ -79,6 +80,9 @@
     "prettier": "^2.1.2",
     "rimraf": "^3.0.2",
     "standard-version": "^9.0.0",
-    "typescript": "^4.0.3"
+    "typescript": "^4.6.4"
+  },
+  "resolutions": {
+    "joi": "17.6.0"
   }
 }
diff --git a/src/client/theme/SearchBar/SuggestionTemplate.ts b/src/client/theme/SearchBar/SuggestionTemplate.ts
@@ -49,7 +49,7 @@ export function SuggestionTemplate({
     !isInterOfTree && !isLastOfTree && explicitSearchResultPath;
   if (needsExplicitHitPath) {
     const pathItems = page
-      ? (page.b ?? []).concat(page.t).concat(document.s ?? [])
+      ? (page.b ?? []).concat(page.t).concat(!document.s || document.s === page.t ? [] : document.s)
       : document.b;
     wrapped.push(
       `<span class="${styles.hitPath}">${concatDocumentPath(

diff --git a/src/server/utils/parseDocument.spec.ts b/src/server/utils/parseDocument.spec.ts
@@ -51,6 +51,11 @@ describe("parseDocument", () => {
       {
         pageTitle: "Hello World",
         sections: [
+          {
+            title: "Hello World",
+            hash: "",
+            content: "",
+          },
           {
             title: "Hello fruits.",
             hash: "#hello-fruits",

diff --git a/src/server/utils/parseDocument.ts b/src/server/utils/parseDocument.ts
@@ -1,7 +1,9 @@
+import {blogPostContainerID} from '@docusaurus/utils-common';
 import { ParsedDocument, ParsedDocumentSection } from "../../shared/interfaces";
 import { getCondensedText } from "./getCondensedText";
 
 const HEADINGS = "h1, h2, h3";
+// const SUB_HEADINGS = "h2, h3";
 
 export function parseDocument($: cheerio.Root): ParsedDocument {
   const $pageTitle = $("article h1").first();
@@ -37,18 +39,39 @@ export function parseDocument($: cheerio.Root): ParsedDocument {
       const title = $h.contents().not("a.hash-link").text().trim();
       const hash = $h.find("a.hash-link").attr("href") || "";
 
-      let $sectionElements;
-      let $firstElement = $("article").children().first();
-      if ($h.is($pageTitle) && $firstElement.filter("header").length > 0) {
-        $firstElement = $("article")
-          .children() // div.markdown, header
-          .not("header") // div.markdown
-          .children() // h1, p, p, h2, ...
-          .first(); // h1 || p
-        if ($firstElement.filter(HEADINGS).length > 0) {
-          return;
+      // Find all content between h1 and h2/h3,
+      // which is considered as the content section of page title.
+      let $sectionElements = $([]);
+      if ($h.is($pageTitle)) {
+        const $header = $h.parent();
+        let $firstElement;
+        if ($header.is("header")) {
+          $firstElement = $header;
+        } else {
+          $firstElement = $h;
+        }
+        const blogPost = $(`#${blogPostContainerID}`);
+        if (blogPost.length) {
+          // Simplify blog post.
+          $firstElement = blogPost.children().first();
+          $sectionElements = $firstElement.nextUntil(HEADINGS).addBack();
+        } else {
+          const $nextElements = $firstElement.nextAll();
+          const $headings = $nextElements.filter(HEADINGS);
+          if ($headings.length) {
+            $sectionElements = $firstElement.nextUntil(HEADINGS);
+          } else {
+            for (const next of $nextElements.get()) {
+              const $heading = $(next).find(HEADINGS);
+              if ($heading.length) {
+                $sectionElements = $sectionElements.add($heading.first().prevAll());
+                break;
+              } else {
+                $sectionElements = $sectionElements.add(next);
+              }
+            }
+          }
         }
-        $sectionElements = $firstElement.nextUntil(HEADINGS).addBack();
       } else {
         $sectionElements = $h.nextUntil(HEADINGS);
       }

diff --git a/yarn.lock b/yarn.lock
@@ -973,6 +973,13 @@
     webpack "^5.40.0"
     webpack-merge "^5.8.0"
 
+"@docusaurus/utils-common@^2.0.0-beta.20":
+  version "2.0.0-beta.20"
+  resolved "https://registry.yarnpkg.com/@docusaurus/utils-common/-/utils-common-2.0.0-beta.20.tgz#adb914c331d711a3c0ef2ba3f64139acdf4cd781"
+  integrity sha512-HabHh23vOQn6ygs0PjuCSF/oZaNsYTFsxB2R6EwHNyw01nWgBC3QAcGVmyIWQhlb9p8V3byKgbzVS68hZX5t9A==
+  dependencies:
+    tslib "^2.4.0"
+
 "@docusaurus/utils-validation@^2.0.0-beta.4":
   version "2.0.0-beta.4"
   resolved "https://registry.yarnpkg.com/@docusaurus/utils-validation/-/utils-validation-2.0.0-beta.4.tgz#417ff389d61aab4c6544f169e31bb86573b518df"
@@ -1402,10 +1409,10 @@
     "@nodelib/fs.scandir" "2.1.3"
     fastq "^1.6.0"
 
-"@sideway/address@^4.1.0":
-  version "4.1.0"
-  resolved "https://registry.yarnpkg.com/@sideway/address/-/address-4.1.0.tgz#0b301ada10ac4e0e3fa525c90615e0b61a72b78d"
-  integrity sha512-wAH/JYRXeIFQRsxerIuLjgUu2Xszam+O5xKeatJ4oudShOOirfmsQ1D6LL54XOU2tizpCYku+s1wmU0SYdpoSA==
+"@sideway/address@^4.1.3":
+  version "4.1.4"
+  resolved "https://registry.yarnpkg.com/@sideway/address/-/address-4.1.4.tgz#03dccebc6ea47fdc226f7d3d1ad512955d4783f0"
+  integrity sha512-7vwq+rOHVWjyXxVlR76Agnvhy8I9rpzjosTESvmhNeXOXdZZB15Fl+TI9x1SiHZH5Jv2wTGduSxFDIaq0m3DUw==
   dependencies:
     "@hapi/hoek" "^9.0.0"
 
@@ -4473,14 +4480,14 @@ jest@^28.0.0:
     import-local "^3.0.2"
     jest-cli "^28.1.0"
 
-joi@^17.4.0:
-  version "17.4.0"
-  resolved "https://registry.yarnpkg.com/joi/-/joi-17.4.0.tgz#b5c2277c8519e016316e49ababd41a1908d9ef20"
-  integrity sha512-F4WiW2xaV6wc1jxete70Rw4V/VuMd6IN+a5ilZsxG4uYtUXWu2kq9W5P2dz30e7Gmw8RCbY/u/uk+dMPma9tAg==
+joi@17.6.0, joi@^17.4.0:
+  version "17.6.0"
+  resolved "https://registry.yarnpkg.com/joi/-/joi-17.6.0.tgz#0bb54f2f006c09a96e75ce687957bd04290054b2"
+  integrity sha512-OX5dG6DTbcr/kbMFj0KGYxuew69HPcAE3K/sZpEV2nP6e/j/C0HV+HNiBPCASxdx5T7DMoa0s8UeHWMnb6n2zw==
   dependencies:
     "@hapi/hoek" "^9.0.0"
     "@hapi/topo" "^5.0.0"
-    "@sideway/address" "^4.1.0"
+    "@sideway/address" "^4.1.3"
     "@sideway/formula" "^3.0.0"
     "@sideway/pinpoint" "^2.0.0"
 
@@ -4775,12 +4782,7 @@ lodash.templatesettings@^4.0.0:
   dependencies:
     lodash._reinterpolate "^3.0.0"
 
-lodash@^4.15.0, lodash@^4.17.15, lodash@^4.17.19, lodash@^4.17.20:
-  version "4.17.20"
-  resolved "https://registry.yarnpkg.com/lodash/-/lodash-4.17.20.tgz#b44a9b6297bcb698f1c51a3545a2b3b368d59c52"
-  integrity sha512-PlhdFcillOINfeV7Ni6oF1TAEayyZBoZ8bcshTHqOYJYlrqzRK5hagpagky5o4HfCzzd1TRkXPMFq6cKk9rGmA==
-
-lodash@^4.17.21:
+lodash@^4.15.0, lodash@^4.17.15, lodash@^4.17.19, lodash@^4.17.20, lodash@^4.17.21:
   version "4.17.21"
   resolved "https://registry.yarnpkg.com/lodash/-/lodash-4.17.21.tgz#679591c564c3bffaae8454cf0b3df370c3d6911c"
   integrity sha512-v2kDEe57lecTulaDIuNTPy3Ry4gLGJ6Z1O3vE1krgXZNrsQ+LFTGHVxVjcXPs17LhbZVGedAJv8XZ1tvj5FvSg==
@@ -6444,10 +6446,10 @@ tslib@^1.8.1, tslib@^1.9.0:
   resolved "https://registry.yarnpkg.com/tslib/-/tslib-1.14.1.tgz#cf2d38bdc34a134bcaf1091c41f6619e2f672d00"
   integrity sha512-Xni35NKzjgMrwevysHTCArtLDpPvye8zV/0E4EyYn43P7/7qvQwPh9BGkHewbMulVntbigmcT7rdX3BNo9wRJg==
 
-tslib@^2.1.0, tslib@^2.2.0:
-  version "2.2.0"
-  resolved "https://registry.yarnpkg.com/tslib/-/tslib-2.2.0.tgz#fb2c475977e35e241311ede2693cee1ec6698f5c"
-  integrity sha512-gS9GVHRU+RGn5KQM2rllAlR3dU6m7AcpJKdtH8gFvQiC4Otgk98XnmMU+nZenHt/+VhnBPWwgrJsyrdcw6i23w==
+tslib@^2.1.0, tslib@^2.2.0, tslib@^2.4.0:
+  version "2.4.0"
+  resolved "https://registry.yarnpkg.com/tslib/-/tslib-2.4.0.tgz#7cecaa7f073ce680a05847aa77be941098f36dc3"
+  integrity sha512-d6xOpEDfsi2CZVlPQzGeux8XMwLT9hssAsaPYExaQMuYskwb+x1x7J371tWlbBdWHroy99KnVB6qIkUbs5X3UQ==
 
 tsutils@^3.21.0:
   version "3.21.0"
@@ -6505,10 +6507,10 @@ typedarray@^0.0.6:
   resolved "https://registry.yarnpkg.com/typedarray/-/typedarray-0.0.6.tgz#867ac74e3864187b1d3d47d996a78ec5c8830777"
   integrity sha1-hnrHTjhkGHsdPUfZlqeOxciDB3c=
 
-typescript@^4.0.3:
-  version "4.0.3"
-  resolved "https://registry.yarnpkg.com/typescript/-/typescript-4.0.3.tgz#153bbd468ef07725c1df9c77e8b453f8d36abba5"
-  integrity sha512-tEu6DGxGgRJPb/mVPIZ48e69xCn2yRmCgYmDugAVwmJ6o+0u1RI18eO7E7WBTLYLaEVVOhwQmcdhQHweux/WPg==
+typescript@^4.6.4:
+  version "4.6.4"
+  resolved "https://registry.yarnpkg.com/typescript/-/typescript-4.6.4.tgz#caa78bbc3a59e6a5c510d35703f6a09877ce45e9"
+  integrity sha512-9ia/jWHIEbo49HfjrLGfKbZSuWo9iTMwXO+Ca3pRsSpbsMbc7/IU8NKdCZVRRBafVPGnoJeFL76ZOAA84I9fEg==
 
 uglify-js@^3.1.4:
   version "3.11.2"