2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* Copyright Joyent, Inc. and other Node contributors.
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
|
|
* copy of this software and associated documentation files (the
|
|
|
|
* "Software"), to deal in the Software without restriction, including
|
|
|
|
* without limitation the rights to use, copy, modify, merge, publish,
|
|
|
|
* distribute, sublicense, and/or sell copies of the Software, and to permit
|
|
|
|
* persons to whom the Software is furnished to do so, subject to the
|
|
|
|
* following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice shall be included
|
|
|
|
* in all copies or substantial portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
|
|
|
|
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
|
|
|
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
|
|
|
|
* NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
|
|
|
|
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
|
|
|
|
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
|
|
|
|
* USE OR OTHER DEALINGS IN THE SOFTWARE.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
'use strict';
|
|
|
|
|
|
|
|
var punycode = require('punycode');
|
|
|
|
|
|
|
|
function Url() {
|
|
|
|
this.protocol = null;
|
|
|
|
this.slashes = null;
|
|
|
|
this.auth = null;
|
|
|
|
this.host = null;
|
|
|
|
this.port = null;
|
|
|
|
this.hostname = null;
|
|
|
|
this.hash = null;
|
|
|
|
this.search = null;
|
|
|
|
this.query = null;
|
|
|
|
this.pathname = null;
|
|
|
|
this.path = null;
|
|
|
|
this.href = null;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Reference: RFC 3986, RFC 1808, RFC 2396
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* define these here so at least they only have to be
|
|
|
|
* compiled once on the first module load.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
var protocolPattern = /^([a-z0-9.+-]+:)/i,
|
2024-01-16 21:26:16 +08:00
|
|
|
portPattern = /:[0-9]*$/,
|
|
|
|
|
|
|
|
// Special case for a simple path URL
|
|
|
|
simplePathPattern = /^(\/\/?(?!\/)[^?\s]*)(\?[^\s]*)?$/,
|
|
|
|
|
|
|
|
/*
|
|
|
|
* RFC 2396: characters reserved for delimiting URLs.
|
|
|
|
* We actually just auto-escape these.
|
|
|
|
*/
|
|
|
|
delims = [
|
|
|
|
'<', '>', '"', '`', ' ', '\r', '\n', '\t'
|
|
|
|
],
|
|
|
|
|
|
|
|
// RFC 2396: characters not allowed for various reasons.
|
|
|
|
unwise = [
|
|
|
|
'{', '}', '|', '\\', '^', '`'
|
|
|
|
].concat(delims),
|
|
|
|
|
|
|
|
// Allowed by RFCs, but cause of XSS attacks. Always escape these.
|
|
|
|
autoEscape = ['\''].concat(unwise),
|
|
|
|
/*
|
|
|
|
* Characters that are never ever allowed in a hostname.
|
|
|
|
* Note that any invalid chars are also handled, but these
|
|
|
|
* are the ones that are *expected* to be seen, so we fast-path
|
|
|
|
* them.
|
|
|
|
*/
|
|
|
|
nonHostChars = [
|
|
|
|
'%', '/', '?', ';', '#'
|
|
|
|
].concat(autoEscape),
|
|
|
|
hostEndingChars = [
|
|
|
|
'/', '?', '#'
|
|
|
|
],
|
|
|
|
hostnameMaxLen = 255,
|
|
|
|
hostnamePartPattern = /^[+a-z0-9A-Z_-]{0,63}$/,
|
|
|
|
hostnamePartStart = /^([+a-z0-9A-Z_-]{0,63})(.*)$/,
|
|
|
|
// protocols that can allow "unsafe" and "unwise" chars.
|
|
|
|
unsafeProtocol = {
|
|
|
|
javascript: true,
|
|
|
|
'javascript:': true
|
|
|
|
},
|
|
|
|
// protocols that never have a hostname.
|
|
|
|
hostlessProtocol = {
|
|
|
|
javascript: true,
|
|
|
|
'javascript:': true
|
|
|
|
},
|
|
|
|
// protocols that always contain a // bit.
|
|
|
|
slashedProtocol = {
|
|
|
|
http: true,
|
|
|
|
https: true,
|
|
|
|
ftp: true,
|
|
|
|
gopher: true,
|
|
|
|
file: true,
|
|
|
|
'http:': true,
|
|
|
|
'https:': true,
|
|
|
|
'ftp:': true,
|
|
|
|
'gopher:': true,
|
|
|
|
'file:': true
|
|
|
|
},
|
|
|
|
querystring = require('qs');
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
function urlParse(url, parseQueryString, slashesDenoteHost) {
|
2024-01-16 21:26:16 +08:00
|
|
|
if (url && typeof url === 'object' && url instanceof Url) { return url; }
|
2023-12-18 13:12:25 +08:00
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
var u = new Url();
|
2023-12-18 13:12:25 +08:00
|
|
|
u.parse(url, parseQueryString, slashesDenoteHost);
|
|
|
|
return u;
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
Url.prototype.parse = function (url, parseQueryString, slashesDenoteHost) {
|
|
|
|
if (typeof url !== 'string') {
|
2023-12-18 13:12:25 +08:00
|
|
|
throw new TypeError("Parameter 'url' must be a string, not " + typeof url);
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* Copy chrome, IE, opera backslash-handling behavior.
|
|
|
|
* Back slashes before the query string get converted to forward slashes
|
|
|
|
* See: https://code.google.com/p/chromium/issues/detail?id=25916
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
var queryIndex = url.indexOf('?'),
|
2024-01-16 21:26:16 +08:00
|
|
|
splitter = queryIndex !== -1 && queryIndex < url.indexOf('#') ? '?' : '#',
|
|
|
|
uSplit = url.split(splitter),
|
|
|
|
slashRegex = /\\/g;
|
2023-12-18 13:12:25 +08:00
|
|
|
uSplit[0] = uSplit[0].replace(slashRegex, '/');
|
|
|
|
url = uSplit.join(splitter);
|
|
|
|
|
|
|
|
var rest = url;
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* trim before proceeding.
|
|
|
|
* This is to support parse stuff like " http://foo.com \n"
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
rest = rest.trim();
|
|
|
|
|
|
|
|
if (!slashesDenoteHost && url.split('#').length === 1) {
|
|
|
|
// Try fast path regexp
|
|
|
|
var simplePath = simplePathPattern.exec(rest);
|
|
|
|
if (simplePath) {
|
|
|
|
this.path = rest;
|
|
|
|
this.href = rest;
|
|
|
|
this.pathname = simplePath[1];
|
|
|
|
if (simplePath[2]) {
|
|
|
|
this.search = simplePath[2];
|
|
|
|
if (parseQueryString) {
|
|
|
|
this.query = querystring.parse(this.search.substr(1));
|
|
|
|
} else {
|
|
|
|
this.query = this.search.substr(1);
|
|
|
|
}
|
|
|
|
} else if (parseQueryString) {
|
|
|
|
this.search = '';
|
|
|
|
this.query = {};
|
|
|
|
}
|
|
|
|
return this;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
var proto = protocolPattern.exec(rest);
|
|
|
|
if (proto) {
|
|
|
|
proto = proto[0];
|
|
|
|
var lowerProto = proto.toLowerCase();
|
|
|
|
this.protocol = lowerProto;
|
|
|
|
rest = rest.substr(proto.length);
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* figure out if it's got a host
|
|
|
|
* user@server is *always* interpreted as a hostname, and url
|
|
|
|
* resolution will treat //foo/bar as host=foo,path=bar because that's
|
|
|
|
* how the browser resolves relative URLs.
|
|
|
|
*/
|
|
|
|
if (slashesDenoteHost || proto || rest.match(/^\/\/[^@/]+@[^@/]+/)) {
|
2023-12-18 13:12:25 +08:00
|
|
|
var slashes = rest.substr(0, 2) === '//';
|
|
|
|
if (slashes && !(proto && hostlessProtocol[proto])) {
|
|
|
|
rest = rest.substr(2);
|
|
|
|
this.slashes = true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (!hostlessProtocol[proto] && (slashes || (proto && !slashedProtocol[proto]))) {
|
|
|
|
|
|
|
|
/*
|
|
|
|
* there's a hostname.
|
|
|
|
* the first instance of /, ?, ;, or # ends the host.
|
|
|
|
*
|
|
|
|
* If there is an @ in the hostname, then non-host chars *are* allowed
|
|
|
|
* to the left of the last @ sign, unless some host-ending character
|
|
|
|
* comes *before* the @-sign.
|
|
|
|
* URLs are obnoxious.
|
|
|
|
*
|
|
|
|
* ex:
|
|
|
|
* http://a@b@c/ => user:a@b host:c
|
|
|
|
* http://a@b?@c => user:a host:c path:/?@c
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* v0.12 TODO(isaacs): This is not quite how Chrome does things.
|
|
|
|
* Review our test case against browsers more comprehensively.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
// find the first instance of any hostEndingChars
|
|
|
|
var hostEnd = -1;
|
|
|
|
for (var i = 0; i < hostEndingChars.length; i++) {
|
|
|
|
var hec = rest.indexOf(hostEndingChars[i]);
|
2024-01-16 21:26:16 +08:00
|
|
|
if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) { hostEnd = hec; }
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* at this point, either we have an explicit point where the
|
|
|
|
* auth portion cannot go past, or the last @ char is the decider.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
var auth, atSign;
|
|
|
|
if (hostEnd === -1) {
|
|
|
|
// atSign can be anywhere.
|
|
|
|
atSign = rest.lastIndexOf('@');
|
|
|
|
} else {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* atSign must be in auth portion.
|
|
|
|
* http://a@b/c@d => host:b auth:a path:/c@d
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
atSign = rest.lastIndexOf('@', hostEnd);
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* Now we have a portion which is definitely the auth.
|
|
|
|
* Pull that off.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (atSign !== -1) {
|
|
|
|
auth = rest.slice(0, atSign);
|
|
|
|
rest = rest.slice(atSign + 1);
|
|
|
|
this.auth = decodeURIComponent(auth);
|
|
|
|
}
|
|
|
|
|
|
|
|
// the host is the remaining to the left of the first non-host char
|
|
|
|
hostEnd = -1;
|
|
|
|
for (var i = 0; i < nonHostChars.length; i++) {
|
|
|
|
var hec = rest.indexOf(nonHostChars[i]);
|
2024-01-16 21:26:16 +08:00
|
|
|
if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) { hostEnd = hec; }
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
// if we still have not hit it, then the entire thing is a host.
|
2024-01-16 21:26:16 +08:00
|
|
|
if (hostEnd === -1) { hostEnd = rest.length; }
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
this.host = rest.slice(0, hostEnd);
|
|
|
|
rest = rest.slice(hostEnd);
|
|
|
|
|
|
|
|
// pull out port.
|
|
|
|
this.parseHost();
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* we've indicated that there is a hostname,
|
|
|
|
* so even if it's empty, it has to be present.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
this.hostname = this.hostname || '';
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* if hostname begins with [ and ends with ]
|
|
|
|
* assume that it's an IPv6 address.
|
|
|
|
*/
|
|
|
|
var ipv6Hostname = this.hostname[0] === '[' && this.hostname[this.hostname.length - 1] === ']';
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
// validate a little.
|
|
|
|
if (!ipv6Hostname) {
|
|
|
|
var hostparts = this.hostname.split(/\./);
|
|
|
|
for (var i = 0, l = hostparts.length; i < l; i++) {
|
|
|
|
var part = hostparts[i];
|
2024-01-16 21:26:16 +08:00
|
|
|
if (!part) { continue; }
|
2023-12-18 13:12:25 +08:00
|
|
|
if (!part.match(hostnamePartPattern)) {
|
|
|
|
var newpart = '';
|
|
|
|
for (var j = 0, k = part.length; j < k; j++) {
|
|
|
|
if (part.charCodeAt(j) > 127) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* we replace non-ASCII char with a temporary placeholder
|
|
|
|
* we need this to make sure size of hostname is not
|
|
|
|
* broken by replacing non-ASCII by nothing
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
newpart += 'x';
|
|
|
|
} else {
|
|
|
|
newpart += part[j];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
// we test again with ASCII char only
|
|
|
|
if (!newpart.match(hostnamePartPattern)) {
|
|
|
|
var validParts = hostparts.slice(0, i);
|
|
|
|
var notHost = hostparts.slice(i + 1);
|
|
|
|
var bit = part.match(hostnamePartStart);
|
|
|
|
if (bit) {
|
|
|
|
validParts.push(bit[1]);
|
|
|
|
notHost.unshift(bit[2]);
|
|
|
|
}
|
|
|
|
if (notHost.length) {
|
|
|
|
rest = '/' + notHost.join('.') + rest;
|
|
|
|
}
|
|
|
|
this.hostname = validParts.join('.');
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (this.hostname.length > hostnameMaxLen) {
|
|
|
|
this.hostname = '';
|
|
|
|
} else {
|
|
|
|
// hostnames are always lower case.
|
|
|
|
this.hostname = this.hostname.toLowerCase();
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!ipv6Hostname) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* IDNA Support: Returns a punycoded representation of "domain".
|
|
|
|
* It only converts parts of the domain name that
|
|
|
|
* have non-ASCII characters, i.e. it doesn't matter if
|
|
|
|
* you call it with a domain that already is ASCII-only.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
this.hostname = punycode.toASCII(this.hostname);
|
|
|
|
}
|
|
|
|
|
|
|
|
var p = this.port ? ':' + this.port : '';
|
|
|
|
var h = this.hostname || '';
|
|
|
|
this.host = h + p;
|
|
|
|
this.href += this.host;
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* strip [ and ] from the hostname
|
|
|
|
* the host field still retains them, though
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (ipv6Hostname) {
|
|
|
|
this.hostname = this.hostname.substr(1, this.hostname.length - 2);
|
|
|
|
if (rest[0] !== '/') {
|
|
|
|
rest = '/' + rest;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* now rest is set to the post-host stuff.
|
|
|
|
* chop off any delim chars.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (!unsafeProtocol[lowerProto]) {
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* First, make 100% sure that any "autoEscape" chars get
|
|
|
|
* escaped, even if encodeURIComponent doesn't think they
|
|
|
|
* need to be.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
for (var i = 0, l = autoEscape.length; i < l; i++) {
|
|
|
|
var ae = autoEscape[i];
|
2024-01-16 21:26:16 +08:00
|
|
|
if (rest.indexOf(ae) === -1) { continue; }
|
2023-12-18 13:12:25 +08:00
|
|
|
var esc = encodeURIComponent(ae);
|
|
|
|
if (esc === ae) {
|
|
|
|
esc = escape(ae);
|
|
|
|
}
|
|
|
|
rest = rest.split(ae).join(esc);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// chop off from the tail first.
|
|
|
|
var hash = rest.indexOf('#');
|
|
|
|
if (hash !== -1) {
|
|
|
|
// got a fragment string.
|
|
|
|
this.hash = rest.substr(hash);
|
|
|
|
rest = rest.slice(0, hash);
|
|
|
|
}
|
|
|
|
var qm = rest.indexOf('?');
|
|
|
|
if (qm !== -1) {
|
|
|
|
this.search = rest.substr(qm);
|
|
|
|
this.query = rest.substr(qm + 1);
|
|
|
|
if (parseQueryString) {
|
|
|
|
this.query = querystring.parse(this.query);
|
|
|
|
}
|
|
|
|
rest = rest.slice(0, qm);
|
|
|
|
} else if (parseQueryString) {
|
|
|
|
// no query string, but parseQueryString still requested
|
|
|
|
this.search = '';
|
|
|
|
this.query = {};
|
|
|
|
}
|
2024-01-16 21:26:16 +08:00
|
|
|
if (rest) { this.pathname = rest; }
|
|
|
|
if (slashedProtocol[lowerProto] && this.hostname && !this.pathname) {
|
2023-12-18 13:12:25 +08:00
|
|
|
this.pathname = '/';
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
// to support http.request
|
2023-12-18 13:12:25 +08:00
|
|
|
if (this.pathname || this.search) {
|
|
|
|
var p = this.pathname || '';
|
|
|
|
var s = this.search || '';
|
|
|
|
this.path = p + s;
|
|
|
|
}
|
|
|
|
|
|
|
|
// finally, reconstruct the href based on what has been validated.
|
|
|
|
this.href = this.format();
|
|
|
|
return this;
|
|
|
|
};
|
|
|
|
|
|
|
|
// format a parsed object into a url string
|
|
|
|
function urlFormat(obj) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* ensure it's an object, and not a string url.
|
|
|
|
* If it's an obj, this is a no-op.
|
|
|
|
* this way, you can call url_format() on strings
|
|
|
|
* to clean up potentially wonky urls.
|
|
|
|
*/
|
|
|
|
if (typeof obj === 'string') { obj = urlParse(obj); }
|
|
|
|
if (!(obj instanceof Url)) { return Url.prototype.format.call(obj); }
|
2023-12-18 13:12:25 +08:00
|
|
|
return obj.format();
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
Url.prototype.format = function () {
|
2023-12-18 13:12:25 +08:00
|
|
|
var auth = this.auth || '';
|
|
|
|
if (auth) {
|
|
|
|
auth = encodeURIComponent(auth);
|
|
|
|
auth = auth.replace(/%3A/i, ':');
|
|
|
|
auth += '@';
|
|
|
|
}
|
|
|
|
|
|
|
|
var protocol = this.protocol || '',
|
2024-01-16 21:26:16 +08:00
|
|
|
pathname = this.pathname || '',
|
|
|
|
hash = this.hash || '',
|
|
|
|
host = false,
|
|
|
|
query = '';
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
if (this.host) {
|
|
|
|
host = auth + this.host;
|
|
|
|
} else if (this.hostname) {
|
2024-01-16 21:26:16 +08:00
|
|
|
host = auth + (this.hostname.indexOf(':') === -1 ? this.hostname : '[' + this.hostname + ']');
|
2023-12-18 13:12:25 +08:00
|
|
|
if (this.port) {
|
|
|
|
host += ':' + this.port;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (this.query && typeof this.query === 'object' && Object.keys(this.query).length) {
|
|
|
|
query = querystring.stringify(this.query, {
|
|
|
|
arrayFormat: 'repeat',
|
|
|
|
addQueryPrefix: false
|
|
|
|
});
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
var search = this.search || (query && ('?' + query)) || '';
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (protocol && protocol.substr(-1) !== ':') { protocol += ':'; }
|
2023-12-18 13:12:25 +08:00
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* only the slashedProtocols get the //. Not mailto:, xmpp:, etc.
|
|
|
|
* unless they had them to begin with.
|
|
|
|
*/
|
|
|
|
if (this.slashes || (!protocol || slashedProtocol[protocol]) && host !== false) {
|
2023-12-18 13:12:25 +08:00
|
|
|
host = '//' + (host || '');
|
2024-01-16 21:26:16 +08:00
|
|
|
if (pathname && pathname.charAt(0) !== '/') { pathname = '/' + pathname; }
|
2023-12-18 13:12:25 +08:00
|
|
|
} else if (!host) {
|
|
|
|
host = '';
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (hash && hash.charAt(0) !== '#') { hash = '#' + hash; }
|
|
|
|
if (search && search.charAt(0) !== '?') { search = '?' + search; }
|
2023-12-18 13:12:25 +08:00
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
pathname = pathname.replace(/[?#]/g, function (match) {
|
2023-12-18 13:12:25 +08:00
|
|
|
return encodeURIComponent(match);
|
|
|
|
});
|
|
|
|
search = search.replace('#', '%23');
|
|
|
|
|
|
|
|
return protocol + host + pathname + search + hash;
|
|
|
|
};
|
|
|
|
|
|
|
|
function urlResolve(source, relative) {
|
|
|
|
return urlParse(source, false, true).resolve(relative);
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
Url.prototype.resolve = function (relative) {
|
2023-12-18 13:12:25 +08:00
|
|
|
return this.resolveObject(urlParse(relative, false, true)).format();
|
|
|
|
};
|
|
|
|
|
|
|
|
function urlResolveObject(source, relative) {
|
2024-01-16 21:26:16 +08:00
|
|
|
if (!source) { return relative; }
|
2023-12-18 13:12:25 +08:00
|
|
|
return urlParse(source, false, true).resolveObject(relative);
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
Url.prototype.resolveObject = function (relative) {
|
|
|
|
if (typeof relative === 'string') {
|
2023-12-18 13:12:25 +08:00
|
|
|
var rel = new Url();
|
|
|
|
rel.parse(relative, false, true);
|
|
|
|
relative = rel;
|
|
|
|
}
|
|
|
|
|
|
|
|
var result = new Url();
|
|
|
|
var tkeys = Object.keys(this);
|
|
|
|
for (var tk = 0; tk < tkeys.length; tk++) {
|
|
|
|
var tkey = tkeys[tk];
|
|
|
|
result[tkey] = this[tkey];
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* hash is always overridden, no matter what.
|
|
|
|
* even href="" will remove it.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
result.hash = relative.hash;
|
|
|
|
|
|
|
|
// if the relative url is empty, then there's nothing left to do here.
|
|
|
|
if (relative.href === '') {
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
// hrefs like //foo/bar always cut to the protocol.
|
|
|
|
if (relative.slashes && !relative.protocol) {
|
|
|
|
// take everything except the protocol from relative
|
|
|
|
var rkeys = Object.keys(relative);
|
|
|
|
for (var rk = 0; rk < rkeys.length; rk++) {
|
|
|
|
var rkey = rkeys[rk];
|
2024-01-16 21:26:16 +08:00
|
|
|
if (rkey !== 'protocol') { result[rkey] = relative[rkey]; }
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
// urlParse appends trailing / to urls like http://www.example.com
|
|
|
|
if (slashedProtocol[result.protocol] && result.hostname && !result.pathname) {
|
|
|
|
result.pathname = '/';
|
|
|
|
result.path = result.pathname;
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (relative.protocol && relative.protocol !== result.protocol) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* if it's a known url protocol, then changing
|
|
|
|
* the protocol does weird things
|
|
|
|
* first, if it's not file:, then we MUST have a host,
|
|
|
|
* and if there was a path
|
|
|
|
* to begin with, then we MUST have a path.
|
|
|
|
* if it is file:, then the host is dropped,
|
|
|
|
* because that's known to be hostless.
|
|
|
|
* anything else is assumed to be absolute.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (!slashedProtocol[relative.protocol]) {
|
|
|
|
var keys = Object.keys(relative);
|
|
|
|
for (var v = 0; v < keys.length; v++) {
|
|
|
|
var k = keys[v];
|
|
|
|
result[k] = relative[k];
|
|
|
|
}
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
result.protocol = relative.protocol;
|
|
|
|
if (!relative.host && !hostlessProtocol[relative.protocol]) {
|
|
|
|
var relPath = (relative.pathname || '').split('/');
|
2024-01-16 21:26:16 +08:00
|
|
|
while (relPath.length && !(relative.host = relPath.shift())) { }
|
|
|
|
if (!relative.host) { relative.host = ''; }
|
|
|
|
if (!relative.hostname) { relative.hostname = ''; }
|
|
|
|
if (relPath[0] !== '') { relPath.unshift(''); }
|
|
|
|
if (relPath.length < 2) { relPath.unshift(''); }
|
2023-12-18 13:12:25 +08:00
|
|
|
result.pathname = relPath.join('/');
|
|
|
|
} else {
|
|
|
|
result.pathname = relative.pathname;
|
|
|
|
}
|
|
|
|
result.search = relative.search;
|
|
|
|
result.query = relative.query;
|
|
|
|
result.host = relative.host || '';
|
|
|
|
result.auth = relative.auth;
|
|
|
|
result.hostname = relative.hostname || relative.host;
|
|
|
|
result.port = relative.port;
|
|
|
|
// to support http.request
|
|
|
|
if (result.pathname || result.search) {
|
|
|
|
var p = result.pathname || '';
|
|
|
|
var s = result.search || '';
|
|
|
|
result.path = p + s;
|
|
|
|
}
|
|
|
|
result.slashes = result.slashes || relative.slashes;
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
var isSourceAbs = result.pathname && result.pathname.charAt(0) === '/',
|
|
|
|
isRelAbs = relative.host || relative.pathname && relative.pathname.charAt(0) === '/',
|
|
|
|
mustEndAbs = isRelAbs || isSourceAbs || (result.host && relative.pathname),
|
|
|
|
removeAllDots = mustEndAbs,
|
|
|
|
srcPath = result.pathname && result.pathname.split('/') || [],
|
|
|
|
relPath = relative.pathname && relative.pathname.split('/') || [],
|
|
|
|
psychotic = result.protocol && !slashedProtocol[result.protocol];
|
|
|
|
|
|
|
|
/*
|
|
|
|
* if the url is a non-slashed url, then relative
|
|
|
|
* links like ../.. should be able
|
|
|
|
* to crawl up to the hostname, as well. This is strange.
|
|
|
|
* result.protocol has already been set by now.
|
|
|
|
* Later on, put the first path part into the host field.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (psychotic) {
|
|
|
|
result.hostname = '';
|
|
|
|
result.port = null;
|
|
|
|
if (result.host) {
|
2024-01-16 21:26:16 +08:00
|
|
|
if (srcPath[0] === '') { srcPath[0] = result.host; } else { srcPath.unshift(result.host); }
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
result.host = '';
|
|
|
|
if (relative.protocol) {
|
|
|
|
relative.hostname = null;
|
|
|
|
relative.port = null;
|
|
|
|
if (relative.host) {
|
2024-01-16 21:26:16 +08:00
|
|
|
if (relPath[0] === '') { relPath[0] = relative.host; } else { relPath.unshift(relative.host); }
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
relative.host = null;
|
|
|
|
}
|
|
|
|
mustEndAbs = mustEndAbs && (relPath[0] === '' || srcPath[0] === '');
|
|
|
|
}
|
|
|
|
|
|
|
|
if (isRelAbs) {
|
|
|
|
// it's absolute.
|
2024-01-16 21:26:16 +08:00
|
|
|
result.host = relative.host || relative.host === '' ? relative.host : result.host;
|
|
|
|
result.hostname = relative.hostname || relative.hostname === '' ? relative.hostname : result.hostname;
|
2023-12-18 13:12:25 +08:00
|
|
|
result.search = relative.search;
|
|
|
|
result.query = relative.query;
|
|
|
|
srcPath = relPath;
|
|
|
|
// fall through to the dot-handling below.
|
|
|
|
} else if (relPath.length) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* it's relative
|
|
|
|
* throw away the existing file, and take the new path instead.
|
|
|
|
*/
|
|
|
|
if (!srcPath) { srcPath = []; }
|
2023-12-18 13:12:25 +08:00
|
|
|
srcPath.pop();
|
|
|
|
srcPath = srcPath.concat(relPath);
|
|
|
|
result.search = relative.search;
|
|
|
|
result.query = relative.query;
|
2024-01-16 21:26:16 +08:00
|
|
|
} else if (relative.search != null) {
|
|
|
|
/*
|
|
|
|
* just pull out the search.
|
|
|
|
* like href='?foo'.
|
|
|
|
* Put this after the other two cases because it simplifies the booleans
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
if (psychotic) {
|
2024-01-16 21:26:16 +08:00
|
|
|
result.host = srcPath.shift();
|
|
|
|
result.hostname = result.host;
|
|
|
|
/*
|
|
|
|
* occationaly the auth can get stuck only in host
|
|
|
|
* this especially happens in cases like
|
|
|
|
* url.resolveObject('mailto:local1@domain1', 'local2@domain2')
|
|
|
|
*/
|
|
|
|
var authInHost = result.host && result.host.indexOf('@') > 0 ? result.host.split('@') : false;
|
2023-12-18 13:12:25 +08:00
|
|
|
if (authInHost) {
|
|
|
|
result.auth = authInHost.shift();
|
2024-01-16 21:26:16 +08:00
|
|
|
result.hostname = authInHost.shift();
|
|
|
|
result.host = result.hostname;
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
result.search = relative.search;
|
|
|
|
result.query = relative.query;
|
2024-01-16 21:26:16 +08:00
|
|
|
// to support http.request
|
|
|
|
if (result.pathname !== null || result.search !== null) {
|
|
|
|
result.path = (result.pathname ? result.pathname : '') + (result.search ? result.search : '');
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!srcPath.length) {
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* no path at all. easy.
|
|
|
|
* we've already handled the other stuff above.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
result.pathname = null;
|
2024-01-16 21:26:16 +08:00
|
|
|
// to support http.request
|
2023-12-18 13:12:25 +08:00
|
|
|
if (result.search) {
|
|
|
|
result.path = '/' + result.search;
|
|
|
|
} else {
|
|
|
|
result.path = null;
|
|
|
|
}
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* if a url ENDs in . or .., then it must get a trailing slash.
|
|
|
|
* however, if it ends in anything else non-slashy,
|
|
|
|
* then it must NOT get a trailing slash.
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
var last = srcPath.slice(-1)[0];
|
2024-01-16 21:26:16 +08:00
|
|
|
var hasTrailingSlash = (result.host || relative.host || srcPath.length > 1) && (last === '.' || last === '..') || last === '';
|
2023-12-18 13:12:25 +08:00
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
/*
|
|
|
|
* strip single dots, resolve double dots to parent dir
|
|
|
|
* if the path tries to go above the root, `up` ends up > 0
|
|
|
|
*/
|
2023-12-18 13:12:25 +08:00
|
|
|
var up = 0;
|
|
|
|
for (var i = srcPath.length; i >= 0; i--) {
|
|
|
|
last = srcPath[i];
|
|
|
|
if (last === '.') {
|
|
|
|
srcPath.splice(i, 1);
|
|
|
|
} else if (last === '..') {
|
|
|
|
srcPath.splice(i, 1);
|
|
|
|
up++;
|
|
|
|
} else if (up) {
|
|
|
|
srcPath.splice(i, 1);
|
|
|
|
up--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// if the path is allowed to go above the root, restore leading ..s
|
|
|
|
if (!mustEndAbs && !removeAllDots) {
|
|
|
|
for (; up--; up) {
|
|
|
|
srcPath.unshift('..');
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (mustEndAbs && srcPath[0] !== '' && (!srcPath[0] || srcPath[0].charAt(0) !== '/')) {
|
2023-12-18 13:12:25 +08:00
|
|
|
srcPath.unshift('');
|
|
|
|
}
|
|
|
|
|
|
|
|
if (hasTrailingSlash && (srcPath.join('/').substr(-1) !== '/')) {
|
|
|
|
srcPath.push('');
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
var isAbsolute = srcPath[0] === '' || (srcPath[0] && srcPath[0].charAt(0) === '/');
|
2023-12-18 13:12:25 +08:00
|
|
|
|
|
|
|
// put the host back
|
|
|
|
if (psychotic) {
|
2024-01-16 21:26:16 +08:00
|
|
|
result.hostname = isAbsolute ? '' : srcPath.length ? srcPath.shift() : '';
|
|
|
|
result.host = result.hostname;
|
|
|
|
/*
|
|
|
|
* occationaly the auth can get stuck only in host
|
|
|
|
* this especially happens in cases like
|
|
|
|
* url.resolveObject('mailto:local1@domain1', 'local2@domain2')
|
|
|
|
*/
|
|
|
|
var authInHost = result.host && result.host.indexOf('@') > 0 ? result.host.split('@') : false;
|
2023-12-18 13:12:25 +08:00
|
|
|
if (authInHost) {
|
|
|
|
result.auth = authInHost.shift();
|
2024-01-16 21:26:16 +08:00
|
|
|
result.hostname = authInHost.shift();
|
|
|
|
result.host = result.hostname;
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
mustEndAbs = mustEndAbs || (result.host && srcPath.length);
|
|
|
|
|
|
|
|
if (mustEndAbs && !isAbsolute) {
|
|
|
|
srcPath.unshift('');
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
if (srcPath.length > 0) {
|
|
|
|
result.pathname = srcPath.join('/');
|
|
|
|
} else {
|
2023-12-18 13:12:25 +08:00
|
|
|
result.pathname = null;
|
|
|
|
result.path = null;
|
|
|
|
}
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
// to support request.http
|
|
|
|
if (result.pathname !== null || result.search !== null) {
|
|
|
|
result.path = (result.pathname ? result.pathname : '') + (result.search ? result.search : '');
|
2023-12-18 13:12:25 +08:00
|
|
|
}
|
|
|
|
result.auth = relative.auth || result.auth;
|
|
|
|
result.slashes = result.slashes || relative.slashes;
|
|
|
|
result.href = result.format();
|
|
|
|
return result;
|
|
|
|
};
|
|
|
|
|
2024-01-16 21:26:16 +08:00
|
|
|
Url.prototype.parseHost = function () {
|
2023-12-18 13:12:25 +08:00
|
|
|
var host = this.host;
|
|
|
|
var port = portPattern.exec(host);
|
|
|
|
if (port) {
|
|
|
|
port = port[0];
|
|
|
|
if (port !== ':') {
|
|
|
|
this.port = port.substr(1);
|
|
|
|
}
|
|
|
|
host = host.substr(0, host.length - port.length);
|
|
|
|
}
|
2024-01-16 21:26:16 +08:00
|
|
|
if (host) { this.hostname = host; }
|
2023-12-18 13:12:25 +08:00
|
|
|
};
|
2024-01-16 21:26:16 +08:00
|
|
|
|
|
|
|
exports.parse = urlParse;
|
|
|
|
exports.resolve = urlResolve;
|
|
|
|
exports.resolveObject = urlResolveObject;
|
|
|
|
exports.format = urlFormat;
|
|
|
|
|
|
|
|
exports.Url = Url;
|