- Notifications
You must be signed in to change notification settings - Fork0
AlexJeffcott/adorn
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
Want keyword matching for large amounts of text (with large numbers of keywords) in Javascript-friendly environments?
Want to leverage knowledge resources to enrich a website's content without the overhead of curating and maintaining the extra content?
Want to add interactive elements (whether simply styling, links, or tooltips or something more complex) based on a user's profile or settings?
Adorn your content nimbly, but with power and precision.
Thecomplexity of theFlashText algorithm is linear and can search or replace keywords in one pass over a document. The time complexity of this algorithm is not dependent on the number of terms being searched or replaced. So for a document of size N (characters) and a dictionary of M keywords, the time complexity will be O(n). It is much faster than Regex, because regex time complexity is O(MxN).
TheAho–Corasick algorithm, by contrast, is linear in the length of the strings plus the length of the searched text plus the number of output matches.
There are several methods available depending on what you wish to achieve including, a simple list of match ids, a list of 'dirty' matches, the full text returned with matches wrapped in a custom tag.
It is ready to be used in static html and plain text, in Node and in the Browser. There is also support for simple implementation in React.
It is possible to have case-insensitive AND case-sensitive matches in the same Match instance. In other words, the flower 'rose' is not matched when looking for the name 'Rose' but both 'flower' and 'Flower' can be matched in the same matcher instance unlike the original flash-text implementation.
You can optionally listen for DOM mutations and scrolling.
One possible disadvantage of this algorithm is that it does not match substrings and returns the longest match only (if you want to do that then aho-corasick is probably a better choice for you).
While the below examples showcase htmlWrapping via annotateDOM with acustom element, there is no reason not to take a simpler approach and add a class, or inline styles, or even an anchor link with an href calculated from the matched keyword.
Checkout the React Typescript codesandbox
<htmllang="en"><head><style>x-annotate {border-bottom-color: cadetblue;border-bottom-style: solid;border-bottom-width:2px; }</style></head><body><p> Lorem ipsum dolor sit amet, consectetur adipiscing elit. In cursus cursus enim eu scelerisque. Nam eleifend purus sed quam facilisis aliquet. Fusce feugiat neque elit, non egestas ipsum molestie quis. Suspendisse quis ipsum malesuada, scelerisque tellus quis, auctor tortor. Nam gravida dolor at molestie facilisis. Donec faucibus nisl vitae ante accumsan, id vulputate lorem convallis. Integer condimentum nunc turpis, eget pellentesque nunc gravida nec. Maecenas in tincidunt eros. Nullam ac feugiat turpis. Interdum et malesuada fames ac ante ipsum primis in faucibus. Nullam at posuere urna. Phasellus fermentum dolor nec sapien congue feugiat. Duis aliquam, ex finibus porttitor viverra, quam augue gravida dui, quis cursus purus justo a mi.</p><scripttype="module">import{TextNodesFromDOM,Match,annotateDOM}from'../annotate/build';constinsensitive=newMap([['123',['Ipsum']],['456',['neque']],['789',['Ut']]]);constsensitive=newMap([['321',['Nullam']]]);constopts={tag:'x-annotate',getAttrs:(id:string)=>[['data-match-id',id]]};constmatch=newMatch(insensitive,sensitive,opts);consttextNodesFromDOM=newTextNodesFromDOM(document.body,[opts.tag.toUpperCase()]);annotateDOM(textNodesFromDOM.walk(document.body),match);textNodesFromDOM.watchDOM((ns)=>annotateDOM(ns,match));textNodesFromDOM.watchScroll((ns)=>annotateDOM(ns,match));</script></body></html>
import{FC,useEffect}from'react';import{TextNodesFromDOM,Match,annotateDOM}from'annotate';constinsensitive=newMap([['123',['Ipsum']],['456',['neque']],['789',['Ut']]]);constsensitive=newMap([['321',['Nullam']]]);constopts={tag:'x-annotate',getAttrs:(id:string)=>[['data-match-id',id]]};constmatch=newMatch(ipsumCaseInsensitive,ipsumCaseSensitive,opts);consttextNodesFromDOM=newTextNodesFromDOM(document.body,[opts.tag.toUpperCase()]);constIpsum:FC=()=>{useEffect(()=>{annotateDOM(textNodesFromDOM.walk(document.body),match);constscrollCB=textNodesFromDOM.watchScroll((ns:Node[])=>annotateDOM(ns,match));return()=>textNodesFromDOM.endWatchScroll(scrollCB);},[]);return(<p> Lorem ipsum dolor sit amet, consectetur adipiscing elit. In cursus cursus enim eu scelerisque. Nam eleifend purus sed quam facilisis aliquet. Fusce feugiat neque elit, non egestas ipsum molestie quis. Suspendisse quis ipsum malesuada, scelerisque tellus quis, auctor tortor. Nam gravida dolor at molestie facilisis. Donec faucibus nisl vitae ante accumsan, id vulputate lorem convallis. Integer condimentum nunc turpis, eget pellentesque nunc gravida nec. Maecenas in tincidunt eros. Nullam ac feugiat turpis. Interdum et malesuada fames ac ante ipsum primis in faucibus. Nullam at posuere urna. Phasellus fermentum dolor nec sapien congue feugiat. Duis aliquam, ex finibus porttitor viverra, quam augue gravida dui, quis cursus purus justo a mi.</p>);};
About
Performant and powerful keyword matching, extraction and annotation for the DOM