xiantuo/node_modules/mdurl/parse.js

// 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.

'use strict';

//
// Changes from joyent/node:
//
// 1. No leading slash in paths,
//    e.g. in `url.parse('http://foo?bar')` pathname is ``, not `/`
//
// 2. Backslashes are not replaced with slashes,
//    so `http:\\example.org\` is treated like a relative path
//
// 3. Trailing colon is treated like a part of the path,
//    i.e. in `http://example.org:foo` pathname is `:foo`
//
// 4. Nothing is URL-encoded in the resulting object,
//    (in joyent/node some chars in auth and paths are encoded)
//
// 5. `url.parse()` does not have `parseQueryString` argument
//
// 6. Removed extraneous result properties: `host`, `path`, `query`, etc.,
//    which can be constructed using other parts of the url.
//


function Url() {
  this.protocol = null;
  this.slashes = null;
  this.auth = null;
  this.port = null;
  this.hostname = null;
  this.hash = null;
  this.search = null;
  this.pathname = null;
}

// Reference: RFC 3986, RFC 1808, RFC 2396

// define these here so at least they only have to be
// compiled once on the first module load.
var protocolPattern = /^([a-z0-9.+-]+:)/i,
    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.
    /* eslint-disable no-script-url */
    // 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
    };
    /* eslint-enable no-script-url */

function urlParse(url, slashesDenoteHost) {
  if (url && url instanceof Url) { return url; }

  var u = new Url();
  u.parse(url, slashesDenoteHost);
  return u;
}

Url.prototype.parse = function(url, slashesDenoteHost) {
  var i, l, lowerProto, hec, slashes,
      rest = url;

  // trim before proceeding.
  // This is to support parse stuff like "  http://foo.com  \n"
  rest = rest.trim();

  if (!slashesDenoteHost && url.split('#').length === 1) {
    // Try fast path regexp
    var simplePath = simplePathPattern.exec(rest);
    if (simplePath) {
      this.pathname = simplePath[1];
      if (simplePath[2]) {
        this.search = simplePath[2];
      }
      return this;
    }
  }

  var proto = protocolPattern.exec(rest);
  if (proto) {
    proto = proto[0];
    lowerProto = proto.toLowerCase();
    this.protocol = proto;
    rest = rest.substr(proto.length);
  }

  // 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(/^\/\/[^@\/]+@[^@\/]+/)) {
    slashes = rest.substr(0, 2) === '//';
    if (slashes && !(proto && hostlessProtocol[proto])) {
      rest = rest.substr(2);
      this.slashes = true;
    }
  }

  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.

    // find the first instance of any hostEndingChars
    var hostEnd = -1;
    for (i = 0; i < hostEndingChars.length; i++) {
      hec = rest.indexOf(hostEndingChars[i]);
      if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) {
        hostEnd = hec;
      }
    }

    // at this point, either we have an explicit point where the
    // auth portion cannot go past, or the last @ char is the decider.
    var auth, atSign;
    if (hostEnd === -1) {
      // atSign can be anywhere.
      atSign = rest.lastIndexOf('@');
    } else {
      // atSign must be in auth portion.
      // http://a@b/c@d => host:b auth:a path:/c@d
      atSign = rest.lastIndexOf('@', hostEnd);
    }

    // Now we have a portion which is definitely the auth.
    // Pull that off.
    if (atSign !== -1) {
      auth = rest.slice(0, atSign);
      rest = rest.slice(atSign + 1);
      this.auth = auth;
    }

    // the host is the remaining to the left of the first non-host char
    hostEnd = -1;
    for (i = 0; i < nonHostChars.length; i++) {
      hec = rest.indexOf(nonHostChars[i]);
      if (hec !== -1 && (hostEnd === -1 || hec < hostEnd)) {
        hostEnd = hec;
      }
    }
    // if we still have not hit it, then the entire thing is a host.
    if (hostEnd === -1) {
      hostEnd = rest.length;
    }

    if (rest[hostEnd - 1] === ':') { hostEnd--; }
    var host = rest.slice(0, hostEnd);
    rest = rest.slice(hostEnd);

    // pull out port.
    this.parseHost(host);

    // we've indicated that there is a hostname,
    // so even if it's empty, it has to be present.
    this.hostname = this.hostname || '';

    // 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] === ']';

    // validate a little.
    if (!ipv6Hostname) {
      var hostparts = this.hostname.split(/\./);
      for (i = 0, l = hostparts.length; i < l; i++) {
        var part = hostparts[i];
        if (!part) { continue; }
        if (!part.match(hostnamePartPattern)) {
          var newpart = '';
          for (var j = 0, k = part.length; j < k; j++) {
            if (part.charCodeAt(j) > 127) {
              // 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
              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 = '';
    }

    // strip [ and ] from the hostname
    // the host field still retains them, though
    if (ipv6Hostname) {
      this.hostname = this.hostname.substr(1, this.hostname.length - 2);
    }
  }

  // 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);
    rest = rest.slice(0, qm);
  }
  if (rest) { this.pathname = rest; }
  if (slashedProtocol[lowerProto] &&
      this.hostname && !this.pathname) {
    this.pathname = '';
  }

  return this;
};

Url.prototype.parseHost = function(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);
  }
  if (host) { this.hostname = host; }
};

module.exports = urlParse;