diff options
Diffstat (limited to 'node_modules/sanitizer/lib/uri.js')
| -rw-r--r-- | node_modules/sanitizer/lib/uri.js | 752 |
1 files changed, 752 insertions, 0 deletions
diff --git a/node_modules/sanitizer/lib/uri.js b/node_modules/sanitizer/lib/uri.js new file mode 100644 index 0000000..8b64dc6 --- /dev/null +++ b/node_modules/sanitizer/lib/uri.js @@ -0,0 +1,752 @@ +// Copyright (C) 2010 Google Inc. +// +// 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 + * Implements RFC 3986 for parsing/formatting URIs. + * + * @author mikesamuel@gmail.com + * \@provides URI + * \@overrides window + */ + +var URI = (function () { + + /** + * creates a uri from the string form. The parser is relaxed, so special + * characters that aren't escaped but don't cause ambiguities will not cause + * parse failures. + * + * @return {URI|null} + */ + function parse(uriStr) { + var m = ('' + uriStr).match(URI_RE_); + if (!m) { return null; } + return new URI( + nullIfAbsent(m[1]), + nullIfAbsent(m[2]), + nullIfAbsent(m[3]), + nullIfAbsent(m[4]), + nullIfAbsent(m[5]), + nullIfAbsent(m[6]), + nullIfAbsent(m[7])); + } + + + /** + * creates a uri from the given parts. + * + * @param scheme {string} an unencoded scheme such as "http" or null + * @param credentials {string} unencoded user credentials or null + * @param domain {string} an unencoded domain name or null + * @param port {number} a port number in [1, 32768]. + * -1 indicates no port, as does null. + * @param path {string} an unencoded path + * @param query {Array.<string>|string|null} a list of unencoded cgi + * parameters where even values are keys and odds the corresponding values + * or an unencoded query. + * @param fragment {string} an unencoded fragment without the "#" or null. + * @return {URI} + */ + function create(scheme, credentials, domain, port, path, query, fragment) { + var uri = new URI( + encodeIfExists2(scheme, URI_DISALLOWED_IN_SCHEME_OR_CREDENTIALS_), + encodeIfExists2( + credentials, URI_DISALLOWED_IN_SCHEME_OR_CREDENTIALS_), + encodeIfExists(domain), + port > 0 ? port.toString() : null, + encodeIfExists2(path, URI_DISALLOWED_IN_PATH_), + null, + encodeIfExists(fragment)); + if (query) { + if ('string' === typeof query) { + uri.setRawQuery(query.replace(/[^?&=0-9A-Za-z_\-~.%]/g, encodeOne)); + } else { + uri.setAllParameters(query); + } + } + return uri; + } + function encodeIfExists(unescapedPart) { + if ('string' == typeof unescapedPart) { + return encodeURIComponent(unescapedPart); + } + return null; + }; + /** + * if unescapedPart is non null, then escapes any characters in it that aren't + * valid characters in a url and also escapes any special characters that + * appear in extra. + * + * @param unescapedPart {string} + * @param extra {RegExp} a character set of characters in [\01-\177]. + * @return {string|null} null iff unescapedPart == null. + */ + function encodeIfExists2(unescapedPart, extra) { + if ('string' == typeof unescapedPart) { + return encodeURI(unescapedPart).replace(extra, encodeOne); + } + return null; + }; + /** converts a character in [\01-\177] to its url encoded equivalent. */ + function encodeOne(ch) { + var n = ch.charCodeAt(0); + return '%' + '0123456789ABCDEF'.charAt((n >> 4) & 0xf) + + '0123456789ABCDEF'.charAt(n & 0xf); + } + + /** + * {@updoc + * $ normPath('foo/./bar') + * # 'foo/bar' + * $ normPath('./foo') + * # 'foo' + * $ normPath('foo/.') + * # 'foo' + * $ normPath('foo//bar') + * # 'foo/bar' + * } + */ + function normPath(path) { + return path.replace(/(^|\/)\.(?:\/|$)/g, '$1').replace(/\/{2,}/g, '/'); + } + + var PARENT_DIRECTORY_HANDLER = new RegExp( + '' + // A path break + + '(/|^)' + // followed by a non .. path element + // (cannot be . because normPath is used prior to this RegExp) + + '(?:[^./][^/]*|\\.{2,}(?:[^./][^/]*)|\\.{3,}[^/]*)' + // followed by .. followed by a path break. + + '/\\.\\.(?:/|$)'); + + var PARENT_DIRECTORY_HANDLER_RE = new RegExp(PARENT_DIRECTORY_HANDLER); + + var EXTRA_PARENT_PATHS_RE = /^(?:\.\.\/)*(?:\.\.$)?/; + + /** + * Normalizes its input path and collapses all . and .. sequences except for + * .. sequences that would take it above the root of the current parent + * directory. + * {@updoc + * $ collapse_dots('foo/../bar') + * # 'bar' + * $ collapse_dots('foo/./bar') + * # 'foo/bar' + * $ collapse_dots('foo/../bar/./../../baz') + * # 'baz' + * $ collapse_dots('../foo') + * # '../foo' + * $ collapse_dots('../foo').replace(EXTRA_PARENT_PATHS_RE, '') + * # 'foo' + * } + */ + function collapse_dots(path) { + if (path === null) { return null; } + var p = normPath(path); + // Only /../ left to flatten + var r = PARENT_DIRECTORY_HANDLER_RE; + // We replace with $1 which matches a / before the .. because this + // guarantees that: + // (1) we have at most 1 / between the adjacent place, + // (2) always have a slash if there is a preceding path section, and + // (3) we never turn a relative path into an absolute path. + for (var q; (q = p.replace(r, '$1')) != p; p = q) {}; + return p; + } + + /** + * resolves a relative url string to a base uri. + * @return {URI} + */ + function resolve(baseUri, relativeUri) { + // there are several kinds of relative urls: + // 1. //foo - replaces everything from the domain on. foo is a domain name + // 2. foo - replaces the last part of the path, the whole query and fragment + // 3. /foo - replaces the the path, the query and fragment + // 4. ?foo - replace the query and fragment + // 5. #foo - replace the fragment only + + var absoluteUri = baseUri.clone(); + // we satisfy these conditions by looking for the first part of relativeUri + // that is not blank and applying defaults to the rest + + var overridden = relativeUri.hasScheme(); + + if (overridden) { + absoluteUri.setRawScheme(relativeUri.getRawScheme()); + } else { + overridden = relativeUri.hasCredentials(); + } + + if (overridden) { + absoluteUri.setRawCredentials(relativeUri.getRawCredentials()); + } else { + overridden = relativeUri.hasDomain(); + } + + if (overridden) { + absoluteUri.setRawDomain(relativeUri.getRawDomain()); + } else { + overridden = relativeUri.hasPort(); + } + + var rawPath = relativeUri.getRawPath(); + var simplifiedPath = collapse_dots(rawPath); + if (overridden) { + absoluteUri.setPort(relativeUri.getPort()); + simplifiedPath = simplifiedPath + && simplifiedPath.replace(EXTRA_PARENT_PATHS_RE, ''); + } else { + overridden = !!rawPath; + if (overridden) { + // resolve path properly + if (simplifiedPath.charCodeAt(0) !== 0x2f /* / */) { // path is relative + var absRawPath = collapse_dots(absoluteUri.getRawPath() || '') + .replace(EXTRA_PARENT_PATHS_RE, ''); + var slash = absRawPath.lastIndexOf('/') + 1; + simplifiedPath = collapse_dots( + (slash ? absRawPath.substring(0, slash) : '') + + collapse_dots(rawPath)) + .replace(EXTRA_PARENT_PATHS_RE, ''); + } + } else { + simplifiedPath = simplifiedPath + && simplifiedPath.replace(EXTRA_PARENT_PATHS_RE, ''); + if (simplifiedPath !== rawPath) { + absoluteUri.setRawPath(simplifiedPath); + } + } + } + + if (overridden) { + absoluteUri.setRawPath(simplifiedPath); + } else { + overridden = relativeUri.hasQuery(); + } + + if (overridden) { + absoluteUri.setRawQuery(relativeUri.getRawQuery()); + } else { + overridden = relativeUri.hasFragment(); + } + + if (overridden) { + absoluteUri.setRawFragment(relativeUri.getRawFragment()); + } + + return absoluteUri; + } + + /** + * a mutable URI. + * + * This class contains setters and getters for the parts of the URI. + * The <tt>getXYZ</tt>/<tt>setXYZ</tt> methods return the decoded part -- so + * <code>uri.parse('/foo%20bar').getPath()</code> will return the decoded path, + * <tt>/foo bar</tt>. + * + * <p>The raw versions of fields are available too. + * <code>uri.parse('/foo%20bar').getRawPath()</code> will return the raw path, + * <tt>/foo%20bar</tt>. Use the raw setters with care, since + * <code>URI::toString</code> is not guaranteed to return a valid url if a + * raw setter was used. + * + * <p>All setters return <tt>this</tt> and so may be chained, a la + * <code>uri.parse('/foo').setFragment('part').toString()</code>. + * + * <p>You should not use this constructor directly -- please prefer the factory + * functions {@link uri.parse}, {@link uri.create}, {@link uri.resolve} + * instead.</p> + * + * <p>The parameters are all raw (assumed to be properly escaped) parts, and + * any (but not all) may be null. Undefined is not allowed.</p> + * + * @constructor + */ + function URI( + rawScheme, + rawCredentials, rawDomain, port, + rawPath, rawQuery, rawFragment) { + this.scheme_ = rawScheme; + this.credentials_ = rawCredentials; + this.domain_ = rawDomain; + this.port_ = port; + this.path_ = rawPath; + this.query_ = rawQuery; + this.fragment_ = rawFragment; + /** + * @type {Array|null} + */ + this.paramCache_ = null; + } + + /** returns the string form of the url. */ + URI.prototype.toString = function () { + var out = []; + if (null !== this.scheme_) { out.push(this.scheme_, ':'); } + if (null !== this.domain_) { + out.push('//'); + if (null !== this.credentials_) { out.push(this.credentials_, '@'); } + out.push(this.domain_); + if (null !== this.port_) { out.push(':', this.port_.toString()); } + } + if (null !== this.path_) { out.push(this.path_); } + if (null !== this.query_) { out.push('?', this.query_); } + if (null !== this.fragment_) { out.push('#', this.fragment_); } + return out.join(''); + }; + + URI.prototype.clone = function () { + return new URI(this.scheme_, this.credentials_, this.domain_, this.port_, + this.path_, this.query_, this.fragment_); + }; + + URI.prototype.getScheme = function () { + // HTML5 spec does not require the scheme to be lowercased but + // all common browsers except Safari lowercase the scheme. + return this.scheme_ && decodeURIComponent(this.scheme_).toLowerCase(); + }; + URI.prototype.getRawScheme = function () { + return this.scheme_; + }; + URI.prototype.setScheme = function (newScheme) { + this.scheme_ = encodeIfExists2( + newScheme, URI_DISALLOWED_IN_SCHEME_OR_CREDENTIALS_); + return this; + }; + URI.prototype.setRawScheme = function (newScheme) { + this.scheme_ = newScheme ? newScheme : null; + return this; + }; + URI.prototype.hasScheme = function () { + return null !== this.scheme_; + }; + + + URI.prototype.getCredentials = function () { + return this.credentials_ && decodeURIComponent(this.credentials_); + }; + URI.prototype.getRawCredentials = function () { + return this.credentials_; + }; + URI.prototype.setCredentials = function (newCredentials) { + this.credentials_ = encodeIfExists2( + newCredentials, URI_DISALLOWED_IN_SCHEME_OR_CREDENTIALS_); + + return this; + }; + URI.prototype.setRawCredentials = function (newCredentials) { + this.credentials_ = newCredentials ? newCredentials : null; + return this; + }; + URI.prototype.hasCredentials = function () { + return null !== this.credentials_; + }; + + + URI.prototype.getDomain = function () { + return this.domain_ && decodeURIComponent(this.domain_); + }; + URI.prototype.getRawDomain = function () { + return this.domain_; + }; + URI.prototype.setDomain = function (newDomain) { + return this.setRawDomain(newDomain && encodeURIComponent(newDomain)); + }; + URI.prototype.setRawDomain = function (newDomain) { + this.domain_ = newDomain ? newDomain : null; + // Maintain the invariant that paths must start with a slash when the URI + // is not path-relative. + return this.setRawPath(this.path_); + }; + URI.prototype.hasDomain = function () { + return null !== this.domain_; + }; + + + URI.prototype.getPort = function () { + return this.port_ && decodeURIComponent(this.port_); + }; + URI.prototype.setPort = function (newPort) { + if (newPort) { + newPort = Number(newPort); + if (newPort !== (newPort & 0xffff)) { + throw new Error('Bad port number ' + newPort); + } + this.port_ = '' + newPort; + } else { + this.port_ = null; + } + return this; + }; + URI.prototype.hasPort = function () { + return null !== this.port_; + }; + + + URI.prototype.getPath = function () { + return this.path_ && decodeURIComponent(this.path_); + }; + URI.prototype.getRawPath = function () { + return this.path_; + }; + URI.prototype.setPath = function (newPath) { + return this.setRawPath(encodeIfExists2(newPath, URI_DISALLOWED_IN_PATH_)); + }; + URI.prototype.setRawPath = function (newPath) { + if (newPath) { + newPath = String(newPath); + this.path_ = + // Paths must start with '/' unless this is a path-relative URL. + (!this.domain_ || /^\//.test(newPath)) ? newPath : '/' + newPath; + } else { + this.path_ = null; + } + return this; + }; + URI.prototype.hasPath = function () { + return null !== this.path_; + }; + + + URI.prototype.getQuery = function () { + // From http://www.w3.org/Addressing/URL/4_URI_Recommentations.html + // Within the query string, the plus sign is reserved as shorthand notation + // for a space. + return this.query_ && decodeURIComponent(this.query_).replace(/\+/g, ' '); + }; + URI.prototype.getRawQuery = function () { + return this.query_; + }; + URI.prototype.setQuery = function (newQuery) { + this.paramCache_ = null; + this.query_ = encodeIfExists(newQuery); + return this; + }; + URI.prototype.setRawQuery = function (newQuery) { + this.paramCache_ = null; + this.query_ = newQuery ? newQuery : null; + return this; + }; + URI.prototype.hasQuery = function () { + return null !== this.query_; + }; + + /** + * sets the query given a list of strings of the form + * [ key0, value0, key1, value1, ... ]. + * + * <p><code>uri.setAllParameters(['a', 'b', 'c', 'd']).getQuery()</code> + * will yield <code>'a=b&c=d'</code>. + */ + URI.prototype.setAllParameters = function (params) { + if (typeof params === 'object') { + if (!(params instanceof Array) + && (params instanceof Object + || Object.prototype.toString.call(params) !== '[object Array]')) { + var newParams = []; + var i = -1; + for (var k in params) { + var v = params[k]; + if ('string' === typeof v) { + newParams[++i] = k; + newParams[++i] = v; + } + } + params = newParams; + } + } + this.paramCache_ = null; + var queryBuf = []; + var separator = ''; + for (var j = 0; j < params.length;) { + var k = params[j++]; + var v = params[j++]; + queryBuf.push(separator, encodeURIComponent(k.toString())); + separator = '&'; + if (v) { + queryBuf.push('=', encodeURIComponent(v.toString())); + } + } + this.query_ = queryBuf.join(''); + return this; + }; + URI.prototype.checkParameterCache_ = function () { + if (!this.paramCache_) { + var q = this.query_; + if (!q) { + this.paramCache_ = []; + } else { + var cgiParams = q.split(/[&\?]/); + var out = []; + var k = -1; + for (var i = 0; i < cgiParams.length; ++i) { + var m = cgiParams[i].match(/^([^=]*)(?:=(.*))?$/); + // From http://www.w3.org/Addressing/URL/4_URI_Recommentations.html + // Within the query string, the plus sign is reserved as shorthand + // notation for a space. + out[++k] = decodeURIComponent(m[1]).replace(/\+/g, ' '); + out[++k] = decodeURIComponent(m[2] || '').replace(/\+/g, ' '); + } + this.paramCache_ = out; + } + } + }; + /** + * sets the values of the named cgi parameters. + * + * <p>So, <code>uri.parse('foo?a=b&c=d&e=f').setParameterValues('c', ['new']) + * </code> yields <tt>foo?a=b&c=new&e=f</tt>.</p> + * + * @param key {string} + * @param values {Array.<string>} the new values. If values is a single string + * then it will be treated as the sole value. + */ + URI.prototype.setParameterValues = function (key, values) { + // be nice and avoid subtle bugs where [] operator on string performs charAt + // on some browsers and crashes on IE + if (typeof values === 'string') { + values = [ values ]; + } + + this.checkParameterCache_(); + var newValueIndex = 0; + var pc = this.paramCache_; + var params = []; + for (var i = 0, k = 0; i < pc.length; i += 2) { + if (key === pc[i]) { + if (newValueIndex < values.length) { + params.push(key, values[newValueIndex++]); + } + } else { + params.push(pc[i], pc[i + 1]); + } + } + while (newValueIndex < values.length) { + params.push(key, values[newValueIndex++]); + } + this.setAllParameters(params); + return this; + }; + URI.prototype.removeParameter = function (key) { + return this.setParameterValues(key, []); + }; + /** + * returns the parameters specified in the query part of the uri as a list of + * keys and values like [ key0, value0, key1, value1, ... ]. + * + * @return {Array.<string>} + */ + URI.prototype.getAllParameters = function () { + this.checkParameterCache_(); + return this.paramCache_.slice(0, this.paramCache_.length); + }; + /** + * returns the value<b>s</b> for a given cgi parameter as a list of decoded + * query parameter values. + * @return {Array.<string>} + */ + URI.prototype.getParameterValues = function (paramNameUnescaped) { + this.checkParameterCache_(); + var values = []; + for (var i = 0; i < this.paramCache_.length; i += 2) { + if (paramNameUnescaped === this.paramCache_[i]) { + values.push(this.paramCache_[i + 1]); + } + } + return values; + }; + /** + * returns a map of cgi parameter names to (non-empty) lists of values. + * @return {Object.<string,Array.<string>>} + */ + URI.prototype.getParameterMap = function (paramNameUnescaped) { + this.checkParameterCache_(); + var paramMap = {}; + for (var i = 0; i < this.paramCache_.length; i += 2) { + var key = this.paramCache_[i++], + value = this.paramCache_[i++]; + if (!(key in paramMap)) { + paramMap[key] = [value]; + } else { + paramMap[key].push(value); + } + } + return paramMap; + }; + /** + * returns the first value for a given cgi parameter or null if the given + * parameter name does not appear in the query string. + * If the given parameter name does appear, but has no '<tt>=</tt>' following + * it, then the empty string will be returned. + * @return {string|null} + */ + URI.prototype.getParameterValue = function (paramNameUnescaped) { + this.checkParameterCache_(); + for (var i = 0; i < this.paramCache_.length; i += 2) { + if (paramNameUnescaped === this.paramCache_[i]) { + return this.paramCache_[i + 1]; + } + } + return null; + }; + + URI.prototype.getFragment = function () { + return this.fragment_ && decodeURIComponent(this.fragment_); + }; + URI.prototype.getRawFragment = function () { + return this.fragment_; + }; + URI.prototype.setFragment = function (newFragment) { + this.fragment_ = newFragment ? encodeURIComponent(newFragment) : null; + return this; + }; + URI.prototype.setRawFragment = function (newFragment) { + this.fragment_ = newFragment ? newFragment : null; + return this; + }; + URI.prototype.hasFragment = function () { + return null !== this.fragment_; + }; + + function nullIfAbsent(matchPart) { + return ('string' == typeof matchPart) && (matchPart.length > 0) + ? matchPart + : null; + } + + + + + /** + * a regular expression for breaking a URI into its component parts. + * + * <p>http://www.gbiv.com/protocols/uri/rfc/rfc3986.html#RFC2234 says + * 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. + * + * <p>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> + * + * <p>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> + * + * <p>msamuel: I have modified the regular expression slightly to expose the + * credentials, domain, and port separately from the authority. + * The modified version yields + * <pre> + * $1 = http scheme + * $2 = <undefined> credentials -\ + * $3 = www.ics.uci.edu domain | authority + * $4 = <undefined> port -/ + * $5 = /pub/ietf/uri/ path + * $6 = <undefined> query without ? + * $7 = Related fragment without # + * </pre> + */ + var URI_RE_ = new RegExp( + "^" + + "(?:" + + "([^:/?#]+)" + // scheme + ":)?" + + "(?://" + + "(?:([^/?#]*)@)?" + // credentials + "([^/?#:@]*)" + // domain + "(?::([0-9]+))?" + // port + ")?" + + "([^?#]+)?" + // path + "(?:\\?([^#]*))?" + // query + "(?:#(.*))?" + // fragment + "$" + ); + + var URI_DISALLOWED_IN_SCHEME_OR_CREDENTIALS_ = /[#\/\?@]/g; + var URI_DISALLOWED_IN_PATH_ = /[\#\?]/g; + + URI.parse = parse; + URI.create = create; + URI.resolve = resolve; + URI.collapse_dots = collapse_dots; // Visible for testing. + +// lightweight string-based api for loadModuleMaker + URI.utils = { + mimeTypeOf: function (uri) { + var uriObj = parse(uri); + if (/\.html$/.test(uriObj.getPath())) { + return 'text/html'; + } else { + return 'application/javascript'; + } + }, + resolve: function (base, uri) { + if (base) { + return resolve(parse(base), parse(uri)).toString(); + } else { + return '' + uri; + } + } + }; + + + return URI; +})(); + +if (typeof exports !== 'undefined') { + if (typeof module !== 'undefined' && module.exports) { + exports = module.exports = URI; + } + exports.URI = URI; +} else { + + // Exports for closure compiler. + if (typeof window !== 'undefined') { + window['URI'] = URI; + } +} |