Url-knife 
Overview
Extract and decompose (fuzzy) URLs (including emails, which are conceptually a part of URLs) in texts with robust patterns.
URL knife (recommended)
Fuzzy URL knife (false positives, but detect human errors)
Installation
For ES5 users,
<html>
<body>
<p id="content"></p>
<script src="../dist/url-knife.bundle.js"></script>
<script type="text/javascript">
</script>
</body>
</html>For ES6 npm users, run 'npm install --save url-knife' on console.
import Pattern from 'url-knife';Syntax & Usage
Chapter 1. Normalize or parse one URL
Chapter 2. Extract all URLs or emails
Chapter 3. Extract URIs with certain names
Chapter 4. Extract all fuzzy URLs (False positives detected)
Chapter 5. Extract all URLs in raw HTML or XML
Chapter 1. Normalize or parse one URL
The following two methods should be used for only one url, not for multiple ones in texts. (for multiple ones, refer to Chapter 2 & 4)
normalizeUrl vs parseUrl
If you want to parse a normal url with no typos, it is safe to use parseUrl. However, normalizeUrl is for parsing urls with possible human errors.
-
normalizeUrl
/**
* @brief
* Normalize an url with potential human errors (Intranet urls are not allowed.)
*/
var sample1 = Pattern.UrlArea.normalizeUrl("htp/:/abcgermany.,def;:9094 #park//noon??abc=retry")
var sample2 = Pattern.UrlArea.normalizeUrl("'://abc.jppp:9091 /park/noon'")
var sample3 = Pattern.UrlArea.normalizeUrl("ss hd : /university,.acd. ;jpkp: 9091/adc??abc=.com")console.log()
{
"url": "htp/:/abcgermany.,def;:9094 #park//noon??abc=retry",
"normalizedUrl": "http://abcgermany.de:9094#park/noon?abc=retry",
"removedTailOnUrl": "",
"protocol": "http",
"onlyDomain": "abcgermany.de",
"onlyParams": "?abc=retry",
"onlyUri": "#park/noon",
"onlyUriWithParams": "#park/noon?abc=retry",
"onlyParamsJsn": {
"abc": "retry"
},
"type": "domain",
"port": "9094"
}
{
"url": "'://abc.jppp:9091 /park/noon'",
"normalizedUrl": "abc.jp:9091/park/noon",
"removedTailOnUrl": "'",
"protocol": null,
"onlyDomain": "abc.jp",
"onlyParams": null,
"onlyUri": "/park/noon'",
"onlyUriWithParams": "/park/noon'",
"onlyParamsJsn": null,
"type": "domain",
"port": "9091"
}
{
"url": "ss hd : /university,.acd. ;jpkp로 접속",
"normalizedUrl": "ssh://university.ac.jp",
"removedTailOnUrl": "",
"protocol": "ssh",
"onlyDomain": "university.ac.jp",
"onlyParams": null,
"onlyUri": null,
"onlyUriWithParams": null,
"onlyParamsJsn": null,
"type": "domain",
"port": null
}-
parseUrl
/**
* @brief
* Parse an url with no potential human errors
*/
var url = Pattern.UrlArea.parseUrl("xtp://gooppalgo.com/park/tree/?abc=1")console.log()
{
"url": "xtp://gooppalgo.com/park/tree/?abc=1",
"removedTailOnUrl": "",
"protocol": "xtp (unknown protocol)",
"onlyDomain": "gooppalgo.com",
"onlyParams": "?abc=1",
"onlyUri": "/park/tree/",
"onlyUriWithParams": "/park/tree/?abc=1",
"onlyParamsJsn": {
"abc": "1"
},
"type": "domain",
"port": null
}Chapter 2. Extract all URLs or emails
The following methods are recommended to use in most cases.
-
extractAllUrls
var textStr = 'http://[::1]:8000에서 http ://www.example.com/wpstyle/?p=364 is ok \n' +
'HTTP://foo.com/blah_blah_(wikipedia) https://www.google.com/maps/place/USA/@36.2218457,... tnae1ver.com:8000on the internet Asterisk\n ' +
'the packed1book.net. fakeshouldnotbedetected.url?abc=fake s5houl７十七日dbedetected.jp?japan=go&html=<span>가나다@pacbook.net</span>; abc.com/ad/fg/?kk=5 [email protected]' +
'Have you visited http://goasidaio.ac.kr?abd=5안녕하세요?5...,.&kkk=5rk.,, ' +
'http://✪df.ws/123\n' +
'http://142.42.1.1:8080/\n' +
'http://-.~_!$&\'()*+,;=:%40:80%2f::::::@example.com ' +
'Have <b>you</b> visited goasidaio.ac.kr?abd=5hell0?5...&kkk=5rk.,. ';
/**
* @brief
* Distill all urls from normal text
* @author Andrew Kang
* @param textStr string required
* @param noProtocolJsn object
* default : {
'ip_v4' : false,
'ip_v6' : false,
'localhost' : false,
'intranet' : false
}
var urls = Pattern.TextArea.extractAllUrls(textStr, {
'ip_v4' : true,
'ip_v6' : false,
'localhost' : false,
'intranet' : true
})-
extractAllEmails
/**
* @brief
* Distill all emails from normal text
* @author Andrew Kang
* @param textStr string required
* @param prefixSanitizer boolean (default : false)
* @return array
*/
var emails = Pattern.TextArea.extractAllEmails(textStr, true)
console.log()
You may be wondering what the 'pass' property below means. If 'pass' is true, that is the email pattern is strictly true following RFC rules.
[{
"value": {
"email": "가나다@apacbook.ac.kr",
"removedTailOnEmail": null,
"type": "domain"
},
"area": "text",
"index": {
"start": 222,
"end": 240
},
"pass": false
},
{
"value": {
"email": "[email protected]",
"removedTailOnEmail": null,
"type": "domain",
"removedTailOnUrl": "..."
},
"area": "text",
"index": {
"start": 242,
"end": 263
},
"pass": true
}]Chapter 3. Extract URIs with certain names
var sampleText = 'https://google.com/abc/777?a=5&b=7 abc/def 333/kak abc/55에서 abc/53 abc/53a/ka /123a/abc/556/dd /abc/123?a=5&b=tkt /xyj/asff' +
'a333/kak nice/guy/ bad/or/nice/guy ssh://nice.guy.com/?a=dkdfl';
/**
* @brief
* Distill uris with certain names from normal text
* @author Andrew Kang
* @param textStr string required
* @param uris array required
* for example, [['a','b'], ['c','d']]
* If you use {number}, this means 'only number' ex) [['a','{number}'], ['c','d']]
* @param endBoundary boolean (default : false)
* @return array
*/
var uris = Pattern.TextArea.extractCertainUris(sampleText,
[['{number}', 'kak'], ['nice','guy'],['abc', '{number}']], true)
// 'If endBoundary is set to false, more uris are detected.'
// This detects all URIs containing '{number}/kak' or nice/guy' or 'abc/{number}'console.log()
[
{
"uri_detected": {
"value": {
"url": "/abc/777?a=5&b=7",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": "",
"onlyParams": "?a=5&b=7",
"onlyUri": "/abc/777",
"onlyUriWithParams": "/abc/777?a=5&b=7",
"onlyParamsJsn": {
"a": "5",
"b": "7"
},
"type": "domain",
"port": null
},
"area": "text",
"index": {
"start": 18,
"end": 34
}
},
"in_what_url": {
"value": {
"url": "https://google.com/abc/777?a=5&b=7",
"removedTailOnUrl": "",
"protocol": "https",
"onlyDomain": "google.com",
"onlyParams": "?a=5&b=7",
"onlyUri": "/abc/777",
"onlyUriWithParams": "/abc/777?a=5&b=7",
"onlyParamsJsn": {
"a": "5",
"b": "7"
},
"type": "domain",
"port": null
},
"area": "text",
"index": {
"start": 0,
"end": 34
}
}
},
{
"uri_detected": {
"value": {
"url": "333/kak",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "333/kak",
"onlyUriWithParams": "333/kak",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 43,
"end": 51
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "abc/53",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "abc/53",
"onlyUriWithParams": "abc/53",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 60,
"end": 67
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "abc/533/ka",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "abc/533/ka",
"onlyUriWithParams": "abc/533/ka",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 67,
"end": 77
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "/123a/abc/556/dd",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "/123a/abc/556/dd",
"onlyUriWithParams": "/123a/abc/556/dd",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 89,
"end": 105
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "/abc/123?a=5&b=tkt",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": "?a=5&b=tkt",
"onlyUri": "/abc/123",
"onlyUriWithParams": "/abc/123?a=5&b=tkt",
"onlyParamsJsn": {
"a": "5",
"b": "tkt"
},
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 106,
"end": 124
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "nice/guy",
"removedTailOnUrl": "/",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "nice/guy",
"onlyUriWithParams": "nice/guy",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 144,
"end": 153
}
},
"in_what_url": null
},
{
"uri_detected": {
"value": {
"url": "/or/nice/guy",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": null,
"onlyParams": null,
"onlyUri": "/or/nice/guy",
"onlyUriWithParams": "/or/nice/guy",
"onlyParamsJsn": null,
"type": "uri",
"port": null
},
"area": "text",
"index": {
"start": 157,
"end": 170
}
},
"in_what_url": null
}
]Chapter 4. Extract all fuzzy URLs
The strongest url extracting method of URL-knife in natural language texts. However, this does not detect intranets due to false positives. If you need to extract intranets, go back to the Chapter 2 above.
var textStr = '142 .42.1.1:8080 123.45 xtp://--[::1]:8000에서 h ttpp ;//-www.ex ample;com -/wpstyle/??p=3?6/4&x=5/3 in the ssh h::/;/ww.example.com/wpstyle/?p=364 is ok ' +
'h ttp:/://132 .42.,1.1 HT TP:// foo, co,.kr/blah_blah_(wikipedia) https://www.google .org :8005/maps/place/USA/@36.2218457,... tnae1ver.co. jp;8000on the internet Asterisk\n ' +
'the packed1book.net. 가나다@apacbook.ac.kr fakeshouldnotbedetected.url?abc=fake s5houl７十七日dbedetected.jp?japan=go&html=<span>가나다@pacbook.travelersinsurance</span>;' +
' abc,com//ad/fg/?kk=5 [email protected] Have you visited http://agoasidaio.ac.kr?abd=55...,.&kkk=5rk.,, ' +
'Have <b>you</b> visited goasidaio.ac.kr?abd=5hell0?5...&kkk=5rk.,. ';
/**
* @brief
* Distill all urls including fuzzy matched ones from normal text
* @author Andrew Kang
* @param textStr string required
var urls = Pattern.TextArea.extractAllFuzzyUrls(textStr)console.log()
Chapter 5. Extract all URLs in raw HTML or XML
// The sample of 'XML (HTML)'
var xmlStr =
'en.wikipedia.org/wiki/Wikipedia:About\n' +
'<body><p>packed1book.net?user[name][first]=tj&user[name][last]=holowaychuk</p>\n' +
'fakeshouldnotbedetected.url?abc=fake -s5houl７十七日dbedetected.jp?japan=go- ' +
'plus.google.co.kr0에서.., \n' +
'https://plus.google.com/+google\n' +
'https://www.google.com/maps/place/USA/@36.2218457,...' +
'<img style=\' = > float : none ; height: 200px;max-width: 50%;margin-top : 3%\' alt="undefined" src="http://www.aaa가가.com/image/showWorkOrderImg?fileName=12345.png"/>\n' +
'<!--how about adackedbooked.co.kr-the site? 请发邮件给我abc件给@navered.com ssh://www.aaa가.com" <p >--邮件给aa件给@daum.net</p> www.naver.com\n <p style="width: 100%"></p>--> "[email protected]"로 보내주세요. ' +
'-gigi.dau.ac.kr?mac=10 -dau.ac.kr?mac=10 <p id="abc" class="def xxx gh" style="<>">[email protected]에서 가나다@pacbook.net<span style="color: rgb(127,127,127);">Please align the paper to the left.</span> </p>\n' +
'<p> 구루.com <img style="float:none;height: 200px;margin-top : 3%" src="/image/showWorkOrderImg?fileName=123456.png" alt="undefined" abc/></p>\n' +
'http: //ne1ver.com:8000?abc=1&dd=5 localhost:80 estonia.ee/ estonia.ee? <p class="https://www.aadc给s.cn"> https://flaviocopes.com/how-to-inspect-javascript-object/ ※Please ask 203.35.33.555:8000 if you have any issues! ※ </p></body> Have you visited goasidaioaaa.ac.kr';
var urls = PatternExtractor.XmlArea.extractAllUrls(xmlStr); console.log()
[
// Not all listed
{
"value": {
"url": "packed1book.net?user[name][first]=tj&user[name][last]=holowaychuk",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": "packed1book.net",
"onlyParams": "?user[name][first]=tj&user[name][last]=holowaychuk",
"onlyUri": null,
"onlyUriWithParams": "?user[name][first]=tj&user[name][last]=holowaychuk",
"onlyParamsJsn": {
"user": {
"name": {
"first": "tj",
"last": "holowaychuk"
}
}
},
"type": "domain",
"port": null
},
"area": "text"
},
{
"value": {
"url": "adackedbooked.co.kr",
"removedTailOnUrl": "",
"protocol": null,
"onlyDomain": "adackedbooked.co.kr",
"onlyParams": null,
"onlyUri": null,
"onlyUriWithParams": null,
"onlyParamsJsn": null,
"type": "domain",
"port": null
},
"area": "comment"
}
.....
]Please inform me of more sophisticated patterns you need by leaving issues on Github or emailing me at [email protected].