chromium/third_party/google_input_tools/third_party/closure_library/closure/goog/uri/utils.js

// Copyright 2008 The Closure Library Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//      http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS-IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

/**
 * @fileoverview Simple utilities for dealing with URI strings.
 *
 * This is intended to be a lightweight alternative to constructing goog.Uri
 * objects.  Whereas goog.Uri adds several kilobytes to the binary regardless
 * of how much of its functionality you use, this is designed to be a set of
 * mostly-independent utilities so that the compiler includes only what is
 * necessary for the task.  Estimated savings of porting is 5k pre-gzip and
 * 1.5k post-gzip.  To ensure the savings remain, future developers should
 * avoid adding new functionality to existing functions, but instead create
 * new ones and factor out shared code.
 *
 * Many of these utilities have limited functionality, tailored to common
 * cases.  The query parameter utilities assume that the parameter keys are
 * already encoded, since most keys are compile-time alphanumeric strings.  The
 * query parameter mutation utilities also do not tolerate fragment identifiers.
 *
 * By design, these functions can be slower than goog.Uri equivalents.
 * Repeated calls to some of functions may be quadratic in behavior for IE,
 * although the effect is somewhat limited given the 2kb limit.
 *
 * One advantage of the limited functionality here is that this approach is
 * less sensitive to differences in URI encodings than goog.Uri, since these
 * functions operate on strings directly, rather than decoding them and
 * then re-encoding.
 *
 * Uses features of RFC 3986 for parsing/formatting URIs:
 *   http://www.ietf.org/rfc/rfc3986.txt
 *
 * @author [email protected] (Garrett Boyer) - The "lightened" design.
 */

goog.provide('goog.uri.utils');
goog.provide('goog.uri.utils.ComponentIndex');
goog.provide('goog.uri.utils.QueryArray');
goog.provide('goog.uri.utils.QueryValue');
goog.provide('goog.uri.utils.StandardQueryParam');

goog.require('goog.asserts');
goog.require('goog.string');
goog.require('goog.userAgent');


/**
 * Character codes inlined to avoid object allocations due to charCode.
 * @enum {number}
 * @private
 */
goog.uri.utils.CharCode_ = {
  AMPERSAND: 38,
  EQUAL: 61,
  HASH: 35,
  QUESTION: 63
};


/**
 * Builds a URI string from already-encoded parts.
 *
 * No encoding is performed.  Any component may be omitted as either null or
 * undefined.
 *
 * @param {?string=} opt_scheme The scheme such as 'http'.
 * @param {?string=} opt_userInfo The user name before the '@'.
 * @param {?string=} opt_domain The domain such as 'www.google.com', already
 *     URI-encoded.
 * @param {(string|number|null)=} opt_port The port number.
 * @param {?string=} opt_path The path, already URI-encoded.  If it is not
 *     empty, it must begin with a slash.
 * @param {?string=} opt_queryData The URI-encoded query data.
 * @param {?string=} opt_fragment The URI-encoded fragment identifier.
 * @return {string} The fully combined URI.
 */
goog.uri.utils.buildFromEncodedParts = function(opt_scheme, opt_userInfo,
    opt_domain, opt_port, opt_path, opt_queryData, opt_fragment) {
  var out = '';

  if (opt_scheme) {
    out += opt_scheme + ':';
  }

  if (opt_domain) {
    out += '//';

    if (opt_userInfo) {
      out += opt_userInfo + '@';
    }

    out += opt_domain;

    if (opt_port) {
      out += ':' + opt_port;
    }
  }

  if (opt_path) {
    out += opt_path;
  }

  if (opt_queryData) {
    out += '?' + opt_queryData;
  }

  if (opt_fragment) {
    out += '#' + opt_fragment;
  }

  return out;
};


/**
 * A regular expression for breaking a URI into its component parts.
 *
 * {@link http://www.ietf.org/rfc/rfc3986.txt} says in Appendix B
 * As the "first-match-wins" algorithm is identical to the "greedy"
 * disambiguation method used by POSIX regular expressions, it is natural and
 * commonplace to use a regular expression for parsing the potential five
 * components of a URI reference.
 *
 * The following line is the regular expression for breaking-down a
 * well-formed URI reference into its components.
 *
 * <pre>
 * ^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?
 *  12            3  4          5       6  7        8 9
 * </pre>
 *
 * The numbers in the second line above are only to assist readability; they
 * indicate the reference points for each subexpression (i.e., each paired
 * parenthesis). We refer to the value matched for subexpression <n> as $<n>.
 * For example, matching the above expression to
 * <pre>
 *     http://www.ics.uci.edu/pub/ietf/uri/#Related
 * </pre>
 * results in the following subexpression matches:
 * <pre>
 *    $1 = http:
 *    $2 = http
 *    $3 = //www.ics.uci.edu
 *    $4 = www.ics.uci.edu
 *    $5 = /pub/ietf/uri/
 *    $6 = <undefined>
 *    $7 = <undefined>
 *    $8 = #Related
 *    $9 = Related
 * </pre>
 * where <undefined> indicates that the component is not present, as is the
 * case for the query component in the above example. Therefore, we can
 * determine the value of the five components as
 * <pre>
 *    scheme    = $2
 *    authority = $4
 *    path      = $5
 *    query     = $7
 *    fragment  = $9
 * </pre>
 *
 * The regular expression has been modified slightly to expose the
 * userInfo, domain, and port separately from the authority.
 * The modified version yields
 * <pre>
 *    $1 = http              scheme
 *    $2 = <undefined>       userInfo -\
 *    $3 = www.ics.uci.edu   domain     | authority
 *    $4 = <undefined>       port     -/
 *    $5 = /pub/ietf/uri/    path
 *    $6 = <undefined>       query without ?
 *    $7 = Related           fragment without #
 * </pre>
 * @type {!RegExp}
 * @private
 */
goog.uri.utils.splitRe_ = new RegExp(
    '^' +
    '(?:' +
        '([^:/?#.]+)' +                  // scheme - ignore special characters
                                         // used by other URL parts such as :,
                                         // ?, /, #, and .
    ':)?' +
    '(?://' +
        '(?:([^/?#]*)@)?' +              // userInfo
        '([^/#?]*?)' +                   // domain
        '(?::([0-9]+))?' +               // port
        '(?=[/#?]|$)' +                  // authority-terminating character
    ')?' +
    '([^?#]+)?' +                        // path
    '(?:\\?([^#]*))?' +                  // query
    '(?:#(.*))?' +                       // fragment
    '$');


/**
 * The index of each URI component in the return value of goog.uri.utils.split.
 * @enum {number}
 */
goog.uri.utils.ComponentIndex = {
  SCHEME: 1,
  USER_INFO: 2,
  DOMAIN: 3,
  PORT: 4,
  PATH: 5,
  QUERY_DATA: 6,
  FRAGMENT: 7
};


/**
 * Splits a URI into its component parts.
 *
 * Each component can be accessed via the component indices; for example:
 * <pre>
 * goog.uri.utils.split(someStr)[goog.uri.utils.CompontentIndex.QUERY_DATA];
 * </pre>
 *
 * @param {string} uri The URI string to examine.
 * @return {!Array<string|undefined>} Each component still URI-encoded.
 *     Each component that is present will contain the encoded value, whereas
 *     components that are not present will be undefined or empty, depending
 *     on the browser's regular expression implementation.  Never null, since
 *     arbitrary strings may still look like path names.
 */
goog.uri.utils.split = function(uri) {
  goog.uri.utils.phishingProtection_();

  // See @return comment -- never null.
  return /** @type {!Array<string|undefined>} */ (
      uri.match(goog.uri.utils.splitRe_));
};


/**
 * Safari has a nasty bug where if you have an http URL with a username, e.g.,
 * http://evil.com%[email protected]/
 * Safari will report that window.location.href is
 * http://evil.com/google.com/
 * so that anyone who tries to parse the domain of that URL will get
 * the wrong domain. We've seen exploits where people use this to trick
 * Safari into loading resources from evil domains.
 *
 * To work around this, we run a little "Safari phishing check", and throw
 * an exception if we see this happening.
 *
 * There is no convenient place to put this check. We apply it to
 * anyone doing URI parsing on Webkit. We're not happy about this, but
 * it fixes the problem.
 *
 * This should be removed once Safari fixes their bug.
 *
 * Exploit reported by Masato Kinugawa.
 *
 * @type {boolean}
 * @private
 */
goog.uri.utils.needsPhishingProtection_ = goog.userAgent.WEBKIT;


/**
 * Check to see if the user is being phished.
 * @private
 */
goog.uri.utils.phishingProtection_ = function() {
  if (goog.uri.utils.needsPhishingProtection_) {
    // Turn protection off, so that we don't recurse.
    goog.uri.utils.needsPhishingProtection_ = false;

    // Use quoted access, just in case the user isn't using location externs.
    var location = goog.global['location'];
    if (location) {
      var href = location['href'];
      if (href) {
        var domain = goog.uri.utils.getDomain(href);
        if (domain && domain != location['hostname']) {
          // Phishing attack
          goog.uri.utils.needsPhishingProtection_ = true;
          throw Error();
        }
      }
    }
  }
};


/**
 * @param {?string} uri A possibly null string.
 * @param {boolean=} opt_preserveReserved If true, percent-encoding of RFC-3986
 *     reserved characters will not be removed.
 * @return {?string} The string URI-decoded, or null if uri is null.
 * @private
 */
goog.uri.utils.decodeIfPossible_ = function(uri, opt_preserveReserved) {
  if (!uri) {
    return uri;
  }

  return opt_preserveReserved ? decodeURI(uri) : decodeURIComponent(uri);
};


/**
 * Gets a URI component by index.
 *
 * It is preferred to use the getPathEncoded() variety of functions ahead,
 * since they are more readable.
 *
 * @param {goog.uri.utils.ComponentIndex} componentIndex The component index.
 * @param {string} uri The URI to examine.
 * @return {?string} The still-encoded component, or null if the component
 *     is not present.
 * @private
 */
goog.uri.utils.getComponentByIndex_ = function(componentIndex, uri) {
  // Convert undefined, null, and empty string into null.
  return goog.uri.utils.split(uri)[componentIndex] || null;
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The protocol or scheme, or null if none.  Does not
 *     include trailing colons or slashes.
 */
goog.uri.utils.getScheme = function(uri) {
  return goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.SCHEME, uri);
};


/**
 * Gets the effective scheme for the URL.  If the URL is relative then the
 * scheme is derived from the page's location.
 * @param {string} uri The URI to examine.
 * @return {string} The protocol or scheme, always lower case.
 */
goog.uri.utils.getEffectiveScheme = function(uri) {
  var scheme = goog.uri.utils.getScheme(uri);
  if (!scheme && goog.global.self && goog.global.self.location) {
    var protocol = goog.global.self.location.protocol;
    scheme = protocol.substr(0, protocol.length - 1);
  }
  // NOTE: When called from a web worker in Firefox 3.5, location maybe null.
  // All other browsers with web workers support self.location from the worker.
  return scheme ? scheme.toLowerCase() : '';
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The user name still encoded, or null if none.
 */
goog.uri.utils.getUserInfoEncoded = function(uri) {
  return goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.USER_INFO, uri);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The decoded user info, or null if none.
 */
goog.uri.utils.getUserInfo = function(uri) {
  return goog.uri.utils.decodeIfPossible_(
      goog.uri.utils.getUserInfoEncoded(uri));
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The domain name still encoded, or null if none.
 */
goog.uri.utils.getDomainEncoded = function(uri) {
  return goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.DOMAIN, uri);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The decoded domain, or null if none.
 */
goog.uri.utils.getDomain = function(uri) {
  return goog.uri.utils.decodeIfPossible_(
      goog.uri.utils.getDomainEncoded(uri), true /* opt_preserveReserved */);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?number} The port number, or null if none.
 */
goog.uri.utils.getPort = function(uri) {
  // Coerce to a number.  If the result of getComponentByIndex_ is null or
  // non-numeric, the number coersion yields NaN.  This will then return
  // null for all non-numeric cases (though also zero, which isn't a relevant
  // port number).
  return Number(goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.PORT, uri)) || null;
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The path still encoded, or null if none. Includes the
 *     leading slash, if any.
 */
goog.uri.utils.getPathEncoded = function(uri) {
  return goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.PATH, uri);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The decoded path, or null if none.  Includes the leading
 *     slash, if any.
 */
goog.uri.utils.getPath = function(uri) {
  return goog.uri.utils.decodeIfPossible_(
      goog.uri.utils.getPathEncoded(uri), true /* opt_preserveReserved */);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The query data still encoded, or null if none.  Does not
 *     include the question mark itself.
 */
goog.uri.utils.getQueryData = function(uri) {
  return goog.uri.utils.getComponentByIndex_(
      goog.uri.utils.ComponentIndex.QUERY_DATA, uri);
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The fragment identifier, or null if none.  Does not
 *     include the hash mark itself.
 */
goog.uri.utils.getFragmentEncoded = function(uri) {
  // The hash mark may not appear in any other part of the URL.
  var hashIndex = uri.indexOf('#');
  return hashIndex < 0 ? null : uri.substr(hashIndex + 1);
};


/**
 * @param {string} uri The URI to examine.
 * @param {?string} fragment The encoded fragment identifier, or null if none.
 *     Does not include the hash mark itself.
 * @return {string} The URI with the fragment set.
 */
goog.uri.utils.setFragmentEncoded = function(uri, fragment) {
  return goog.uri.utils.removeFragment(uri) + (fragment ? '#' + fragment : '');
};


/**
 * @param {string} uri The URI to examine.
 * @return {?string} The decoded fragment identifier, or null if none.  Does
 *     not include the hash mark.
 */
goog.uri.utils.getFragment = function(uri) {
  return goog.uri.utils.decodeIfPossible_(
      goog.uri.utils.getFragmentEncoded(uri));
};


/**
 * Extracts everything up to the port of the URI.
 * @param {string} uri The URI string.
 * @return {string} Everything up to and including the port.
 */
goog.uri.utils.getHost = function(uri) {
  var pieces = goog.uri.utils.split(uri);
  return goog.uri.utils.buildFromEncodedParts(
      pieces[goog.uri.utils.ComponentIndex.SCHEME],
      pieces[goog.uri.utils.ComponentIndex.USER_INFO],
      pieces[goog.uri.utils.ComponentIndex.DOMAIN],
      pieces[goog.uri.utils.ComponentIndex.PORT]);
};


/**
 * Extracts the path of the URL and everything after.
 * @param {string} uri The URI string.
 * @return {string} The URI, starting at the path and including the query
 *     parameters and fragment identifier.
 */
goog.uri.utils.getPathAndAfter = function(uri) {
  var pieces = goog.uri.utils.split(uri);
  return goog.uri.utils.buildFromEncodedParts(null, null, null, null,
      pieces[goog.uri.utils.ComponentIndex.PATH],
      pieces[goog.uri.utils.ComponentIndex.QUERY_DATA],
      pieces[goog.uri.utils.ComponentIndex.FRAGMENT]);
};


/**
 * Gets the URI with the fragment identifier removed.
 * @param {string} uri The URI to examine.
 * @return {string} Everything preceding the hash mark.
 */
goog.uri.utils.removeFragment = function(uri) {
  // The hash mark may not appear in any other part of the URL.
  var hashIndex = uri.indexOf('#');
  return hashIndex < 0 ? uri : uri.substr(0, hashIndex);
};


/**
 * Ensures that two URI's have the exact same domain, scheme, and port.
 *
 * Unlike the version in goog.Uri, this checks protocol, and therefore is
 * suitable for checking against the browser's same-origin policy.
 *
 * @param {string} uri1 The first URI.
 * @param {string} uri2 The second URI.
 * @return {boolean} Whether they have the same scheme, domain and port.
 */
goog.uri.utils.haveSameDomain = function(uri1, uri2) {
  var pieces1 = goog.uri.utils.split(uri1);
  var pieces2 = goog.uri.utils.split(uri2);
  return pieces1[goog.uri.utils.ComponentIndex.DOMAIN] ==
             pieces2[goog.uri.utils.ComponentIndex.DOMAIN] &&
         pieces1[goog.uri.utils.ComponentIndex.SCHEME] ==
             pieces2[goog.uri.utils.ComponentIndex.SCHEME] &&
         pieces1[goog.uri.utils.ComponentIndex.PORT] ==
             pieces2[goog.uri.utils.ComponentIndex.PORT];
};


/**
 * Asserts that there are no fragment or query identifiers, only in uncompiled
 * mode.
 * @param {string} uri The URI to examine.
 * @private
 */
goog.uri.utils.assertNoFragmentsOrQueries_ = function(uri) {
  // NOTE: would use goog.asserts here, but jscompiler doesn't know that
  // indexOf has no side effects.
  if (goog.DEBUG && (uri.indexOf('#') >= 0 || uri.indexOf('?') >= 0)) {
    throw Error('goog.uri.utils: Fragment or query identifiers are not ' +
        'supported: [' + uri + ']');
  }
};


/**
 * Supported query parameter values by the parameter serializing utilities.
 *
 * If a value is null or undefined, the key-value pair is skipped, as an easy
 * way to omit parameters conditionally.  Non-array parameters are converted
 * to a string and URI encoded.  Array values are expanded into multiple
 * &key=value pairs, with each element stringized and URI-encoded.
 *
 * @typedef {*}
 */
goog.uri.utils.QueryValue;


/**
 * An array representing a set of query parameters with alternating keys
 * and values.
 *
 * Keys are assumed to be URI encoded already and live at even indices.  See
 * goog.uri.utils.QueryValue for details on how parameter values are encoded.
 *
 * Example:
 * <pre>
 * var data = [
 *   // Simple param: ?name=BobBarker
 *   'name', 'BobBarker',
 *   // Conditional param -- may be omitted entirely.
 *   'specialDietaryNeeds', hasDietaryNeeds() ? getDietaryNeeds() : null,
 *   // Multi-valued param: &house=LosAngeles&house=NewYork&house=null
 *   'house', ['LosAngeles', 'NewYork', null]
 * ];
 * </pre>
 *
 * @typedef {!Array<string|goog.uri.utils.QueryValue>}
 */
goog.uri.utils.QueryArray;


/**
 * Parses encoded query parameters and calls callback function for every
 * parameter found in the string.
 *
 * Missing value of parameter (e.g. “…&key&…”) is treated as if the value was an
 * empty string.  Keys may be empty strings (e.g. “…&=value&…”) which also means
 * that “…&=&…” and “…&&…” will result in an empty key and value.
 *
 * @param {string} encodedQuery Encoded query string excluding question mark at
 *     the beginning.
 * @param {function(string, string)} callback Function called for every
 *     parameter found in query string.  The first argument (name) will not be
 *     urldecoded (so the function is consistent with buildQueryData), but the
 *     second will.  If the parameter has no value (i.e. “=” was not present)
 *     the second argument (value) will be an empty string.
 */
goog.uri.utils.parseQueryData = function(encodedQuery, callback) {
  if (!encodedQuery) {
    return;
  }
  var pairs = encodedQuery.split('&');
  for (var i = 0; i < pairs.length; i++) {
    var indexOfEquals = pairs[i].indexOf('=');
    var name = null;
    var value = null;
    if (indexOfEquals >= 0) {
      name = pairs[i].substring(0, indexOfEquals);
      value = pairs[i].substring(indexOfEquals + 1);
    } else {
      name = pairs[i];
    }
    callback(name, value ? goog.string.urlDecode(value) : '');
  }
};


/**
 * Appends a URI and query data in a string buffer with special preconditions.
 *
 * Internal implementation utility, performing very few object allocations.
 *
 * @param {!Array<string|undefined>} buffer A string buffer.  The first element
 *     must be the base URI, and may have a fragment identifier.  If the array
 *     contains more than one element, the second element must be an ampersand,
 *     and may be overwritten, depending on the base URI.  Undefined elements
 *     are treated as empty-string.
 * @return {string} The concatenated URI and query data.
 * @private
 */
goog.uri.utils.appendQueryData_ = function(buffer) {
  if (buffer[1]) {
    // At least one query parameter was added.  We need to check the
    // punctuation mark, which is currently an ampersand, and also make sure
    // there aren't any interfering fragment identifiers.
    var baseUri = /** @type {string} */ (buffer[0]);
    var hashIndex = baseUri.indexOf('#');
    if (hashIndex >= 0) {
      // Move the fragment off the base part of the URI into the end.
      buffer.push(baseUri.substr(hashIndex));
      buffer[0] = baseUri = baseUri.substr(0, hashIndex);
    }
    var questionIndex = baseUri.indexOf('?');
    if (questionIndex < 0) {
      // No question mark, so we need a question mark instead of an ampersand.
      buffer[1] = '?';
    } else if (questionIndex == baseUri.length - 1) {
      // Question mark is the very last character of the existing URI, so don't
      // append an additional delimiter.
      buffer[1] = undefined;
    }
  }

  return buffer.join('');
};


/**
 * Appends key=value pairs to an array, supporting multi-valued objects.
 * @param {string} key The key prefix.
 * @param {goog.uri.utils.QueryValue} value The value to serialize.
 * @param {!Array<string>} pairs The array to which the 'key=value' strings
 *     should be appended.
 * @private
 */
goog.uri.utils.appendKeyValuePairs_ = function(key, value, pairs) {
  if (goog.isArray(value)) {
    // Convince the compiler it's an array.
    goog.asserts.assertArray(value);
    for (var j = 0; j < value.length; j++) {
      // Convert to string explicitly, to short circuit the null and array
      // logic in this function -- this ensures that null and undefined get
      // written as literal 'null' and 'undefined', and arrays don't get
      // expanded out but instead encoded in the default way.
      goog.uri.utils.appendKeyValuePairs_(key, String(value[j]), pairs);
    }
  } else if (value != null) {
    // Skip a top-level null or undefined entirely.
    pairs.push('&', key,
        // Check for empty string. Zero gets encoded into the url as literal
        // strings.  For empty string, skip the equal sign, to be consistent
        // with UriBuilder.java.
        value === '' ? '' : '=',
        goog.string.urlEncode(value));
  }
};


/**
 * Builds a buffer of query data from a sequence of alternating keys and values.
 *
 * @param {!Array<string|undefined>} buffer A string buffer to append to.  The
 *     first element appended will be an '&', and may be replaced by the caller.
 * @param {!goog.uri.utils.QueryArray|!Arguments} keysAndValues An array with
 *     alternating keys and values -- see the typedef.
 * @param {number=} opt_startIndex A start offset into the arary, defaults to 0.
 * @return {!Array<string|undefined>} The buffer argument.
 * @private
 */
goog.uri.utils.buildQueryDataBuffer_ = function(
    buffer, keysAndValues, opt_startIndex) {
  goog.asserts.assert(Math.max(keysAndValues.length - (opt_startIndex || 0),
      0) % 2 == 0, 'goog.uri.utils: Key/value lists must be even in length.');

  for (var i = opt_startIndex || 0; i < keysAndValues.length; i += 2) {
    goog.uri.utils.appendKeyValuePairs_(
        keysAndValues[i], keysAndValues[i + 1], buffer);
  }

  return buffer;
};


/**
 * Builds a query data string from a sequence of alternating keys and values.
 * Currently generates "&key&" for empty args.
 *
 * @param {goog.uri.utils.QueryArray} keysAndValues Alternating keys and
 *     values.  See the typedef.
 * @param {number=} opt_startIndex A start offset into the arary, defaults to 0.
 * @return {string} The encoded query string, in the form 'a=1&b=2'.
 */
goog.uri.utils.buildQueryData = function(keysAndValues, opt_startIndex) {
  var buffer = goog.uri.utils.buildQueryDataBuffer_(
      [], keysAndValues, opt_startIndex);
  buffer[0] = ''; // Remove the leading ampersand.
  return buffer.join('');
};


/**
 * Builds a buffer of query data from a map.
 *
 * @param {!Array<string|undefined>} buffer A string buffer to append to.  The
 *     first element appended will be an '&', and may be replaced by the caller.
 * @param {!Object<string, goog.uri.utils.QueryValue>} map An object where keys
 *     are URI-encoded parameter keys, and the values conform to the contract
 *     specified in the goog.uri.utils.QueryValue typedef.
 * @return {!Array<string|undefined>} The buffer argument.
 * @private
 */
goog.uri.utils.buildQueryDataBufferFromMap_ = function(buffer, map) {
  for (var key in map) {
    goog.uri.utils.appendKeyValuePairs_(key, map[key], buffer);
  }

  return buffer;
};


/**
 * Builds a query data string from a map.
 * Currently generates "&key&" for empty args.
 *
 * @param {!Object<string, goog.uri.utils.QueryValue>} map An object where keys
 *     are URI-encoded parameter keys, and the values are arbitrary types
 *     or arrays. Keys with a null value are dropped.
 * @return {string} The encoded query string, in the form 'a=1&b=2'.
 */
goog.uri.utils.buildQueryDataFromMap = function(map) {
  var buffer = goog.uri.utils.buildQueryDataBufferFromMap_([], map);
  buffer[0] = '';
  return buffer.join('');
};


/**
 * Appends URI parameters to an existing URI.
 *
 * The variable arguments may contain alternating keys and values.  Keys are
 * assumed to be already URI encoded.  The values should not be URI-encoded,
 * and will instead be encoded by this function.
 * <pre>
 * appendParams('http://www.foo.com?existing=true',
 *     'key1', 'value1',
 *     'key2', 'value?willBeEncoded',
 *     'key3', ['valueA', 'valueB', 'valueC'],
 *     'key4', null);
 * result: 'http://www.foo.com?existing=true&' +
 *     'key1=value1&' +
 *     'key2=value%3FwillBeEncoded&' +
 *     'key3=valueA&key3=valueB&key3=valueC'
 * </pre>
 *
 * A single call to this function will not exhibit quadratic behavior in IE,
 * whereas multiple repeated calls may, although the effect is limited by
 * fact that URL's generally can't exceed 2kb.
 *
 * @param {string} uri The original URI, which may already have query data.
 * @param {...(goog.uri.utils.QueryArray|string|goog.uri.utils.QueryValue)} var_args
 *     An array or argument list conforming to goog.uri.utils.QueryArray.
 * @return {string} The URI with all query parameters added.
 */
goog.uri.utils.appendParams = function(uri, var_args) {
  return goog.uri.utils.appendQueryData_(
      arguments.length == 2 ?
      goog.uri.utils.buildQueryDataBuffer_([uri], arguments[1], 0) :
      goog.uri.utils.buildQueryDataBuffer_([uri], arguments, 1));
};


/**
 * Appends query parameters from a map.
 *
 * @param {string} uri The original URI, which may already have query data.
 * @param {!Object<goog.uri.utils.QueryValue>} map An object where keys are
 *     URI-encoded parameter keys, and the values are arbitrary types or arrays.
 *     Keys with a null value are dropped.
 * @return {string} The new parameters.
 */
goog.uri.utils.appendParamsFromMap = function(uri, map) {
  return goog.uri.utils.appendQueryData_(
      goog.uri.utils.buildQueryDataBufferFromMap_([uri], map));
};


/**
 * Appends a single URI parameter.
 *
 * Repeated calls to this can exhibit quadratic behavior in IE6 due to the
 * way string append works, though it should be limited given the 2kb limit.
 *
 * @param {string} uri The original URI, which may already have query data.
 * @param {string} key The key, which must already be URI encoded.
 * @param {*=} opt_value The value, which will be stringized and encoded
 *     (assumed not already to be encoded).  If omitted, undefined, or null, the
 *     key will be added as a valueless parameter.
 * @return {string} The URI with the query parameter added.
 */
goog.uri.utils.appendParam = function(uri, key, opt_value) {
  var paramArr = [uri, '&', key];
  if (goog.isDefAndNotNull(opt_value)) {
    paramArr.push('=', goog.string.urlEncode(opt_value));
  }
  return goog.uri.utils.appendQueryData_(paramArr);
};


/**
 * Finds the next instance of a query parameter with the specified name.
 *
 * Does not instantiate any objects.
 *
 * @param {string} uri The URI to search.  May contain a fragment identifier
 *     if opt_hashIndex is specified.
 * @param {number} startIndex The index to begin searching for the key at.  A
 *     match may be found even if this is one character after the ampersand.
 * @param {string} keyEncoded The URI-encoded key.
 * @param {number} hashOrEndIndex Index to stop looking at.  If a hash
 *     mark is present, it should be its index, otherwise it should be the
 *     length of the string.
 * @return {number} The position of the first character in the key's name,
 *     immediately after either a question mark or a dot.
 * @private
 */
goog.uri.utils.findParam_ = function(
    uri, startIndex, keyEncoded, hashOrEndIndex) {
  var index = startIndex;
  var keyLength = keyEncoded.length;

  // Search for the key itself and post-filter for surronuding punctuation,
  // rather than expensively building a regexp.
  while ((index = uri.indexOf(keyEncoded, index)) >= 0 &&
      index < hashOrEndIndex) {
    var precedingChar = uri.charCodeAt(index - 1);
    // Ensure that the preceding character is '&' or '?'.
    if (precedingChar == goog.uri.utils.CharCode_.AMPERSAND ||
        precedingChar == goog.uri.utils.CharCode_.QUESTION) {
      // Ensure the following character is '&', '=', '#', or NaN
      // (end of string).
      var followingChar = uri.charCodeAt(index + keyLength);
      if (!followingChar ||
          followingChar == goog.uri.utils.CharCode_.EQUAL ||
          followingChar == goog.uri.utils.CharCode_.AMPERSAND ||
          followingChar == goog.uri.utils.CharCode_.HASH) {
        return index;
      }
    }
    index += keyLength + 1;
  }

  return -1;
};


/**
 * Regular expression for finding a hash mark or end of string.
 * @type {RegExp}
 * @private
 */
goog.uri.utils.hashOrEndRe_ = /#|$/;


/**
 * Determines if the URI contains a specific key.
 *
 * Performs no object instantiations.
 *
 * @param {string} uri The URI to process.  May contain a fragment
 *     identifier.
 * @param {string} keyEncoded The URI-encoded key.  Case-sensitive.
 * @return {boolean} Whether the key is present.
 */
goog.uri.utils.hasParam = function(uri, keyEncoded) {
  return goog.uri.utils.findParam_(uri, 0, keyEncoded,
      uri.search(goog.uri.utils.hashOrEndRe_)) >= 0;
};


/**
 * Gets the first value of a query parameter.
 * @param {string} uri The URI to process.  May contain a fragment.
 * @param {string} keyEncoded The URI-encoded key.  Case-sensitive.
 * @return {?string} The first value of the parameter (URI-decoded), or null
 *     if the parameter is not found.
 */
goog.uri.utils.getParamValue = function(uri, keyEncoded) {
  var hashOrEndIndex = uri.search(goog.uri.utils.hashOrEndRe_);
  var foundIndex = goog.uri.utils.findParam_(
      uri, 0, keyEncoded, hashOrEndIndex);

  if (foundIndex < 0) {
    return null;
  } else {
    var endPosition = uri.indexOf('&', foundIndex);
    if (endPosition < 0 || endPosition > hashOrEndIndex) {
      endPosition = hashOrEndIndex;
    }
    // Progress forth to the end of the "key=" or "key&" substring.
    foundIndex += keyEncoded.length + 1;
    // Use substr, because it (unlike substring) will return empty string
    // if foundIndex > endPosition.
    return goog.string.urlDecode(
        uri.substr(foundIndex, endPosition - foundIndex));
  }
};


/**
 * Gets all values of a query parameter.
 * @param {string} uri The URI to process.  May contain a fragment.
 * @param {string} keyEncoded The URI-encoded key.  Case-sensitive.
 * @return {!Array<string>} All URI-decoded values with the given key.
 *     If the key is not found, this will have length 0, but never be null.
 */
goog.uri.utils.getParamValues = function(uri, keyEncoded) {
  var hashOrEndIndex = uri.search(goog.uri.utils.hashOrEndRe_);
  var position = 0;
  var foundIndex;
  var result = [];

  while ((foundIndex = goog.uri.utils.findParam_(
      uri, position, keyEncoded, hashOrEndIndex)) >= 0) {
    // Find where this parameter ends, either the '&' or the end of the
    // query parameters.
    position = uri.indexOf('&', foundIndex);
    if (position < 0 || position > hashOrEndIndex) {
      position = hashOrEndIndex;
    }

    // Progress forth to the end of the "key=" or "key&" substring.
    foundIndex += keyEncoded.length + 1;
    // Use substr, because it (unlike substring) will return empty string
    // if foundIndex > position.
    result.push(goog.string.urlDecode(uri.substr(
        foundIndex, position - foundIndex)));
  }

  return result;
};


/**
 * Regexp to find trailing question marks and ampersands.
 * @type {RegExp}
 * @private
 */
goog.uri.utils.trailingQueryPunctuationRe_ = /[?&]($|#)/;


/**
 * Removes all instances of a query parameter.
 * @param {string} uri The URI to process.  Must not contain a fragment.
 * @param {string} keyEncoded The URI-encoded key.
 * @return {string} The URI with all instances of the parameter removed.
 */
goog.uri.utils.removeParam = function(uri, keyEncoded) {
  var hashOrEndIndex = uri.search(goog.uri.utils.hashOrEndRe_);
  var position = 0;
  var foundIndex;
  var buffer = [];

  // Look for a query parameter.
  while ((foundIndex = goog.uri.utils.findParam_(
      uri, position, keyEncoded, hashOrEndIndex)) >= 0) {
    // Get the portion of the query string up to, but not including, the ?
    // or & starting the parameter.
    buffer.push(uri.substring(position, foundIndex));
    // Progress to immediately after the '&'.  If not found, go to the end.
    // Avoid including the hash mark.
    position = Math.min((uri.indexOf('&', foundIndex) + 1) || hashOrEndIndex,
        hashOrEndIndex);
  }

  // Append everything that is remaining.
  buffer.push(uri.substr(position));

  // Join the buffer, and remove trailing punctuation that remains.
  return buffer.join('').replace(
      goog.uri.utils.trailingQueryPunctuationRe_, '$1');
};


/**
 * Replaces all existing definitions of a parameter with a single definition.
 *
 * Repeated calls to this can exhibit quadratic behavior due to the need to
 * find existing instances and reconstruct the string, though it should be
 * limited given the 2kb limit.  Consider using appendParams to append multiple
 * parameters in bulk.
 *
 * @param {string} uri The original URI, which may already have query data.
 * @param {string} keyEncoded The key, which must already be URI encoded.
 * @param {*} value The value, which will be stringized and encoded (assumed
 *     not already to be encoded).
 * @return {string} The URI with the query parameter added.
 */
goog.uri.utils.setParam = function(uri, keyEncoded, value) {
  return goog.uri.utils.appendParam(
      goog.uri.utils.removeParam(uri, keyEncoded), keyEncoded, value);
};


/**
 * Generates a URI path using a given URI and a path with checks to
 * prevent consecutive "//". The baseUri passed in must not contain
 * query or fragment identifiers. The path to append may not contain query or
 * fragment identifiers.
 *
 * @param {string} baseUri URI to use as the base.
 * @param {string} path Path to append.
 * @return {string} Updated URI.
 */
goog.uri.utils.appendPath = function(baseUri, path) {
  goog.uri.utils.assertNoFragmentsOrQueries_(baseUri);

  // Remove any trailing '/'
  if (goog.string.endsWith(baseUri, '/')) {
    baseUri = baseUri.substr(0, baseUri.length - 1);
  }
  // Remove any leading '/'
  if (goog.string.startsWith(path, '/')) {
    path = path.substr(1);
  }
  return goog.string.buildString(baseUri, '/', path);
};


/**
 * Replaces the path.
 * @param {string} uri URI to use as the base.
 * @param {string} path New path.
 * @return {string} Updated URI.
 */
goog.uri.utils.setPath = function(uri, path) {
  // Add any missing '/'.
  if (!goog.string.startsWith(path, '/')) {
    path = '/' + path;
  }
  var parts = goog.uri.utils.split(uri);
  return goog.uri.utils.buildFromEncodedParts(
      parts[goog.uri.utils.ComponentIndex.SCHEME],
      parts[goog.uri.utils.ComponentIndex.USER_INFO],
      parts[goog.uri.utils.ComponentIndex.DOMAIN],
      parts[goog.uri.utils.ComponentIndex.PORT],
      path,
      parts[goog.uri.utils.ComponentIndex.QUERY_DATA],
      parts[goog.uri.utils.ComponentIndex.FRAGMENT]);
};


/**
 * Standard supported query parameters.
 * @enum {string}
 */
goog.uri.utils.StandardQueryParam = {

  /** Unused parameter for unique-ifying. */
  RANDOM: 'zx'
};


/**
 * Sets the zx parameter of a URI to a random value.
 * @param {string} uri Any URI.
 * @return {string} That URI with the "zx" parameter added or replaced to
 *     contain a random string.
 */
goog.uri.utils.makeUnique = function(uri) {
  return goog.uri.utils.setParam(uri,
      goog.uri.utils.StandardQueryParam.RANDOM, goog.string.getRandomString());
};