---
layout: post
title: Jekyll 的文章搜尋功能
date: 2017-07-12
---
# Jekyll 的文章搜尋功能
![Imgur](https://i.imgur.com/3y5O9tZ.gif)
前幾天在 HackerNews 上看到了這篇文章:[A search widget for static web sites](https://news.ycombinator.com/item?id=14717182),原理就是抓取 RSS 的 xml 檔案當做搜尋的資料庫。
既然都用 Jekyll 了,那產生一個更容易使用的格式,比如 JSON,就不用手寫 regex 去 match XML 啦。
我直接用了 [Jekyll search using lunr.js](https://learn.cloudcannon.com/jekyll/jekyll-search-using-lunr-js/) 這篇文章的做法,在 JavaScript 塞入文章內容:
```javascript
var store = {
{% for post in include.posts %}
"{{ post.url | slugify }}": {
"title": "{{ post.title | xml_escape }}",
"url": "{{ post.url | xml_escape }}",
"author": "{{ post.author | xml_escape }}",
"category": "{{ post.category | xml_escape }}",
"content": {{ post.content | strip_html | strip_newlines | jsonify }},
"url": "{{ post.url | xml_escape }}",
"date": "{{ post.date | date: "%b %-d, %Y" }}"
}
{% unless forloop.last %},{% endunless %}
{% endfor %}
}
```
產生出來的檔案長這樣:
```javascript
var store = {
"blog-2017-06-28-windows-unix-like-development-guide-nodejs-ruby": {
"title": "Windows Unix-Like 環境設定踩坑紀錄",
"url": "/blog/2017/06/28/windows-unix-like-development-guide-nodejs-ruby/",
"author": "",
"category": "",
"content": "...",
"url": "/blog/2017/06/28/windows-unix-like-development-guide-nodejs-ruby/",
"date": "Jun 28, 2017"
},
"blog-2017-05-04-deploy-redmine-with-puma-nginx-migration-from-mysql-to-postgresql": {
"title": "使用 puma 和 nginx 部屬 Redmine(加上從 MySQL 搬到 PostgreSQL)",
"url": "/blog/2017/05/04/deploy-redmine-with-puma-nginx-migration-from-mysql-to-postgresql/",
"author": "",
"category": "",
"content": "...",
"url": "/blog/2017/05/04/deploy-redmine-with-puma-nginx-migration-from-mysql-to-postgresql/",
"date": "May 4, 2017"
},
// ...
}
```
然後用 `lunr.js` 做搜尋:
```javascript
var idx = lunr(function () {
this.field('id');
this.field('title', { boost: 10 });
this.field('author');
this.field('category');
this.field('content');
this.field('url');
this.field('date');
for (var key in store) { // Add the data to lunr
this.add({
'id': key,
'title': store[key].title,
'author': store[key].author,
'category': store[key].category,
'content': store[key].content,
'url': store[key].url,
'date': store[key].date,
});
}
});
var results = idx.search(searchTerm); // Get lunr to perform a search
```
再把結果產生成列表就行了。花了比較多時間在處裡產生的 store 資料上,比如 `jekyll-mention` 插件會把 `@` 開頭的轉換成 GitHub User 連結,以及 `jemoji` 插件會把 `:smile:` 轉成 GitHub 的 emoji 圖片連結,但在:
```txt
post.content | strip_html | strip_newlines | jsonify
```
這個轉換 pipeline 裡面沒法被消除掉,於是轉出來的 JavaScript 就會出現 SyntaxError(多換行啦、double quote 沒有 escape 等)。
另一個問題是因為用到 HTML5 History API,跟 Turbolink 尬在一起就會炸。現在有時還會「回到上一頁但頁面不切換」之類的問題。就暫時放置 Play 了 XD
更多請參考 commit [`bb3a006`](https://github.com/Yukaii/Blog/commit/bb3a006f690f1ceed8793f9fa0b950f6c043bca9) 囉。
(完)