From 6fd19069deb06988b3d34482a151a79a53b589d0 Mon Sep 17 00:00:00 2001
From: Jacky Zhao <j.zhao2k19@gmail.com>
Date: Sun, 24 Oct 2021 23:17:13 -0700
Subject: [PATCH] search improvements

---
 layouts/partials/search.html | 373 ++++++++++++++++++-----------------
 1 file changed, 192 insertions(+), 181 deletions(-)

diff --git a/layouts/partials/search.html b/layouts/partials/search.html
index e2285ea..7ee139d 100644
--- a/layouts/partials/search.html
+++ b/layouts/partials/search.html
@@ -1,208 +1,219 @@
 <div id="search-container">
-    <div>
-        <input autocomplete="off" id="search-bar" name="search" type="text" aria-label="Search" placeholder="Search for something...">
+    <div id="search-space">
+        <input autoComplete="off" id="search-bar" name="search" type="text" aria-label="Search"
+               placeholder="Search for something...">
         <div id="results-container">
         </div>
     </div>
 </div>
 <script src="https://cdn.jsdelivr.net/gh/nextapps-de/flexsearch@0.7.2/dist/flexsearch.bundle.js"></script>
 <script>
-  // code from https://github.com/danestves/markdown-to-text
-  const removeMarkdown = (
-    markdown,
-    options = {
-      listUnicodeChar: false,
-      stripListLeaders: true,
-      gfm: true,
-      useImgAltText: false,
-      preserveLinks: false,
-    }
-  ) => {
-    let output = markdown || "";
-    output = output.replace(/^(-\s*?|\*\s*?|_\s*?){3,}\s*$/gm, "");
-
-    try {
-      if (options.stripListLeaders) {
-        if (options.listUnicodeChar)
-          output = output.replace(
-            /^([\s\t]*)([\*\-\+]|\d+\.)\s+/gm,
-            options.listUnicodeChar + " $1"
-          );
-        else output = output.replace(/^([\s\t]*)([\*\-\+]|\d+\.)\s+/gm, "$1");
-      }
-      if (options.gfm) {
-        output = output
-          .replace(/\n={2,}/g, "\n")
-          .replace(/~{3}.*\n/g, "")
-          .replace(/~~/g, "")
-          .replace(/`{3}.*\n/g, "");
-      }
-      if(options.preserveLinks) {
-        output = output.replace(/\[(.*?)\][\[\(](.*?)[\]\)]/g, "$1 ($2)")
-      }
-      output = output
-        .replace(/<[^>]*>/g, "")
-        .replace(/^[=\-]{2,}\s*$/g, "")
-        .replace(/\[\^.+?\](\: .*?$)?/g, "")
-        .replace(/\s{0,2}\[.*?\]: .*?$/g, "")
-        .replace(/\!\[(.*?)\][\[\(].*?[\]\)]/g, options.useImgAltText ? "$1" : "")
-        .replace(/\[(.*?)\][\[\(].*?[\]\)]/g, "$1")
-        .replace(/^\s{0,3}>\s?/g, "")
-        .replace(/(^|\n)\s{0,3}>\s?/g, "\n\n")
-        .replace(/^\s{1,2}\[(.*?)\]: (\S+)( ".*?")?\s*$/g, "")
-        .replace(
-          /^(\n)?\s{0,}#{1,6}\s+| {0,}(\n)?\s{0,}#{0,} {0,}(\n)?\s{0,}$/gm,
-          "$1$2$3"
-        )
-        .replace(/([\*_]{1,3})(\S.*?\S{0,1})\1/g, "$2")
-        .replace(/([\*_]{1,3})(\S.*?\S{0,1})\1/g, "$2")
-        .replace(/(`{3,})(.*?)\1/gm, "$2")
-        .replace(/`(.+?)`/g, "$1")
-        .replace(/\n{2,}/g, "\n\n");
-    } catch (e) {
-      console.error(e);
-      return markdown;
-    }
-    return output;
-  };
+    // code from https://github.com/danestves/markdown-to-text
+    const removeMarkdown = (
+        markdown,
+        options = {
+            listUnicodeChar: false,
+            stripListLeaders: true,
+            gfm: true,
+            useImgAltText: false,
+            preserveLinks: false,
+        }
+    ) => {
+        let output = markdown || "";
+        output = output.replace(/^(-\s*?|\*\s*?|_\s*?){3,}\s*$/gm, "");
+
+        try {
+            if (options.stripListLeaders) {
+                if (options.listUnicodeChar)
+                    output = output.replace(
+                        /^([\s\t]*)([\*\-\+]|\d+\.)\s+/gm,
+                        options.listUnicodeChar + " $1"
+                    );
+                else output = output.replace(/^([\s\t]*)([\*\-\+]|\d+\.)\s+/gm, "$1");
+            }
+            if (options.gfm) {
+                output = output
+                    .replace(/\n={2,}/g, "\n")
+                    .replace(/~{3}.*\n/g, "")
+                    .replace(/~~/g, "")
+                    .replace(/`{3}.*\n/g, "");
+            }
+            if (options.preserveLinks) {
+                output = output.replace(/\[(.*?)\][\[\(](.*?)[\]\)]/g, "$1 ($2)")
+            }
+            output = output
+                .replace(/<[^>]*>/g, "")
+                .replace(/^[=\-]{2,}\s*$/g, "")
+                .replace(/\[\^.+?\](\: .*?$)?/g, "")
+                .replace(/\s{0,2}\[.*?\]: .*?$/g, "")
+                .replace(/\!\[(.*?)\][\[\(].*?[\]\)]/g, options.useImgAltText ? "$1" : "")
+                .replace(/\[(.*?)\][\[\(].*?[\]\)]/g, "$1")
+                .replace(/^\s{0,3}>\s?/g, "")
+                .replace(/(^|\n)\s{0,3}>\s?/g, "\n\n")
+                .replace(/^\s{1,2}\[(.*?)\]: (\S+)( ".*?")?\s*$/g, "")
+                .replace(
+                    /^(\n)?\s{0,}#{1,6}\s+| {0,}(\n)?\s{0,}#{0,} {0,}(\n)?\s{0,}$/gm,
+                    "$1$2$3"
+                )
+                .replace(/([\*_]{1,3})(\S.*?\S{0,1})\1/g, "$2")
+                .replace(/([\*_]{1,3})(\S.*?\S{0,1})\1/g, "$2")
+                .replace(/(`{3,})(.*?)\1/gm, "$2")
+                .replace(/`(.+?)`/g, "$1")
+                .replace(/\n{2,}/g, "\n\n");
+        } catch (e) {
+            console.error(e);
+            return markdown;
+        }
+        return output;
+    };
 </script>
 <script>
-  const contentIndex = new FlexSearch.Worker({
-    tokenize: "strict",
-    charset: "latin:advanced",
-    context: true,
-    depth: 3,
-    cache: 10,
-    suggest: true,
-  })
-
-  const scrapedContent = {{$.Site.Data.contentIndex}}
-  for (const [key, value] of Object.entries(scrapedContent)) {
-    contentIndex.add(key, value.content)
-  }
-
-  const stopwords = ['i','me','my','myself','we','our','ours','ourselves','you','your','yours','yourself','yourselves','he','him','his','himself','she','her','hers','herself','it','its','itself','they','them','their','theirs','themselves','what','which','who','whom','this','that','these','those','am','is','are','was','were','be','been','being','have','has','had','having','do','does','did','doing','a','an','the','and','but','if','or','because','as','until','while','of','at','by','for','with','about','against','between','into','through','during','before','after','above','below','to','from','up','down','in','out','on','off','over','under','again','further','then','once','here','there','when','where','why','how','all','any','both','each','few','more','most','other','some','such','no','nor','not','only','own','same','so','than','too','very','s','t','can','will','just','don','should','now']
-  const highlight = (content, term) => {
-    const highlightWindow = 15
-    const tokenizedTerm = term.split(/\s+/).filter(t => t !== "")
-    const splitText = content.split(/\s+/).filter(t => t !== "")
-    const includesCheck = (token) => tokenizedTerm.some(term => token.toLowerCase().includes(term.toLowerCase()))
-
-    const occurrencesIndices = splitText
-      .map(includesCheck)
-
-    // calculate best index
-    let bestSum = 0
-    let bestIndex = 0
-    for (let i = 0; i < Math.max(occurrencesIndices.length - highlightWindow, 0); i++) {
-      const window = occurrencesIndices.slice(i, i + highlightWindow)
-      const windowSum = window.reduce((total, cur) => total + cur, 0)
-      if (windowSum > bestSum) {
-        bestSum = windowSum
-        bestIndex = i
-      }
+    const contentIndex = new FlexSearch.Worker({
+        tokenize: "reverse",
+        charset: "latin:extra",
+        suggest: true,
+    })
+
+    const scrapedContent = {{$.Site.Data.contentIndex}}
+    for (const [key, value] of Object.entries(scrapedContent)) {
+        contentIndex.add(key, value.content)
     }
 
-    const startIndex = Math.max(bestIndex - highlightWindow, 0)
-    const endIndex = Math.min(startIndex + 2 * highlightWindow, splitText.length)
-    const mappedText = splitText
-      .slice(startIndex, endIndex)
-      .map(token => {
-        if (includesCheck(token)) {
-          return `<span class="search-highlight">${token}</span>`
+    const highlight = (content, term) => {
+        const highlightWindow = 15
+        const tokenizedTerm = term.split(/\s+/).filter(t => t !== "")
+        const splitText = content.split(/\s+/).filter(t => t !== "")
+        const includesCheck = (token) => tokenizedTerm.some(term => token.toLowerCase().includes(term.toLowerCase()))
+
+        const occurrencesIndices = splitText
+            .map(includesCheck)
+
+        // calculate best index
+        let bestSum = 0
+        let bestIndex = 0
+        for (let i = 0; i < Math.max(occurrencesIndices.length - highlightWindow, 0); i++) {
+            const window = occurrencesIndices.slice(i, i + highlightWindow)
+            const windowSum = window.reduce((total, cur) => total + cur, 0)
+            if (windowSum >= bestSum) {
+                bestSum = windowSum
+                bestIndex = i
+            }
         }
-        return token
-      })
-      .join(" ")
-      .replaceAll('</span> <span class="search-highlight">', " ")
-    return `${startIndex === 0 ? "" : "..."}${mappedText}${endIndex === splitText.length ? "" : "..."}`
-  }
-
-  const resultToHTML = ({url, title, content, term}) => {
-    const md = content.split("---")[2]
-    const text = removeMarkdown(md)
-    const resultTitle = highlight(title, term)
-    const resultText = highlight(text, term)
-    return `<div class="result-card" id="${url}">
+
+        const startIndex = Math.max(bestIndex - highlightWindow, 0)
+        const endIndex = Math.min(startIndex + 2 * highlightWindow, splitText.length)
+        const mappedText = splitText
+            .slice(startIndex, endIndex)
+            .map(token => {
+                if (includesCheck(token)) {
+                    return `<span class="search-highlight">${token}</span>`
+                }
+                return token
+            })
+            .join(" ")
+            .replaceAll('</span> <span class="search-highlight">', " ")
+        return `${startIndex === 0 ? "" : "..."}${mappedText}${endIndex === splitText.length ? "" : "..."}`
+    }
+
+    const resultToHTML = ({url, title, content, term}) => {
+        const md = content.split("---")[2]
+        const text = removeMarkdown(md)
+        const resultTitle = highlight(title, term)
+        const resultText = highlight(text, term)
+        return `<button class="result-card" id="${url}">
         <h3>${resultTitle}</h3>
         <p>${resultText}</p>
-    </div>`
-  }
-
-  const source = document.getElementById('search-bar')
-  const results = document.getElementById("results-container")
-  source.addEventListener('input', (e) => {
-    const term = e.target.value
-    contentIndex.search(term, {
-      limit: 5,
-      depth: 3,
-      suggest: true,
-    }).then(searchResults => {
-      const resultIds = [...new Set(searchResults)]
-      const finalResults = resultIds.map(id => ({
-        url: id,
-        title: scrapedContent[id].title,
-        content: scrapedContent[id].content
-      }))
-
-      // display
-      if (finalResults.length === 0) {
-        results.innerHTML = `<div class="result-card">
+    </button>`
+    }
+
+    const source = document.getElementById('search-bar')
+    const results = document.getElementById("results-container")
+    let term
+    source.addEventListener("keyup", (e) => {
+        if (e.key === "Enter") {
+            const anchor = document.getElementsByClassName("result-card")[0]
+            window.location.href = `${anchor.id}#:~:text=${encodeURIComponent(term)}`
+        }
+    })
+    source.addEventListener('input', (e) => {
+        term = e.target.value
+        contentIndex.search(term, {
+            limit: 20,
+            depth: 3,
+            suggest: true,
+        }).then(searchResults => {
+            const resultIds = [...new Set(searchResults)]
+            const finalResults = resultIds.map(id => ({
+                url: id,
+                title: scrapedContent[id].title,
+                content: scrapedContent[id].content
+            }))
+
+            // display
+            if (finalResults.length === 0) {
+                results.innerHTML = `<div class="result-card">
             <p>No results.</p>
         </div>`
-      } else {
-        results.innerHTML = finalResults
-          .map(result => resultToHTML({
-            ...result,
-            term,
-          }))
-          .join("\n")
-        const anchors = document.getElementsByClassName("result-card");
-        [...anchors].forEach(anchor => {
-          anchor.onclick = () => {
-            window.location.href = `${anchor.id}#:~:text=${encodeURIComponent(term)}`
-          }
+            } else {
+                results.innerHTML = finalResults
+                    .map(result => resultToHTML({
+                        ...result,
+                        term,
+                    }))
+                    .join("\n")
+                const anchors = document.getElementsByClassName("result-card");
+                [...anchors].forEach(anchor => {
+                    anchor.onclick = () => {
+                        window.location.href = `${anchor.id}#:~:text=${encodeURIComponent(term)}`
+                    }
+                })
+            }
         })
-      }
     })
-  })
-
-
-  const searchContainer = document.getElementById("search-container")
-  function openSearch() {
-    if (searchContainer.style.display === "none" || searchContainer.style.display === "") {
-      source.value = ""
-      results.innerHTML = ""
-      searchContainer.style.display = "block"
-      source.focus()
-    } else {
-      searchContainer.style.display = "none"
-    }
-  }
 
-  function closeSearch() {
-    searchContainer.style.display = "none"
-  }
 
-  document.addEventListener('keydown', (event) => {
-    if (event.key === "/") {
-      event.preventDefault()
-      openSearch()
+    const searchContainer = document.getElementById("search-container")
+
+    function openSearch() {
+        if (searchContainer.style.display === "none" || searchContainer.style.display === "") {
+            source.value = ""
+            results.innerHTML = ""
+            searchContainer.style.display = "block"
+            source.focus()
+        } else {
+            searchContainer.style.display = "none"
+        }
     }
-    if (event.key === "Escape") {
-      event.preventDefault()
-      closeSearch()
+
+    function closeSearch() {
+        searchContainer.style.display = "none"
     }
-  })
 
-  window.addEventListener('DOMContentLoaded', () => {
-    const searchButton = document.getElementById("search-icon")
-    searchButton.addEventListener('click', (evt) => {
-      openSearch()
+    document.addEventListener('keydown', (event) => {
+        if (event.key === "/") {
+            event.preventDefault()
+            openSearch()
+        }
+        if (event.key === "Escape") {
+            event.preventDefault()
+            closeSearch()
+        }
     })
-    searchButton.addEventListener('keydown', (evt) => {
-      openSearch()
+
+    window.addEventListener('DOMContentLoaded', () => {
+        const searchButton = document.getElementById("search-icon")
+        searchButton.addEventListener('click', (evt) => {
+            openSearch()
+        })
+        searchButton.addEventListener('keydown', (evt) => {
+            openSearch()
+        })
+        searchContainer.addEventListener('click', (evt) => {
+            closeSearch()
+        })
+        document.getElementById("search-space").addEventListener('click', (evt) => {
+            evt.stopPropagation()
+        })
     })
-  })
 
 </script>
\ No newline at end of file