diff --git a/doc/api.txt b/doc/api.txt index 2fb568c12f64fe..4e125bf8f77536 100644 --- a/doc/api.txt +++ b/doc/api.txt @@ -1531,6 +1531,99 @@ require("path").exists("/etc/passwd", function (exists) { ------------------------------------ +=== URL Module + +This module has utilities for URL resolution and parsing. + +Parsed URL objects have some or all of the following fields, depending on whether or not +they exist in the URL string. Any parts that are not in the URL string will not be in the +parsed object. Examples are shown for the URL +"http://user:pass@host.com:8080/p/a/t/h?query=string#hash"+ + ++href+:: +The full URL that was originally parsed. Example: +"http://user:pass@host.com:8080/p/a/t/h?query=string#hash"+ + ++protocol+:: +The request protocol. Example: +"http:"+ + ++host+:: +The full host portion of the URL, including port and authentication information. Example: ++"user:pass@host.com:8080"+ + ++auth+:: +The authentication information portion of a URL. Example: +"user:pass"+ + ++hostname+:: +Just the hostname portion of the host. Example: +"host.com"+ + ++port+:: +The port number portion of the host. Example: +"8080"+ + ++pathname+:: +The path section of the URL, that comes after the host and before the query, including the +initial slash if present. Example: +"/p/a/t/h"+ + ++search+:: +The "query string" portion of the URL, including the leading question mark. Example: ++"?query=string"+ + ++query+:: +Either the "params" portion of the query string, or a querystring-parsed object. Example: ++"query=string"+ or +{"query":"string"}+ + ++hash+:: +The portion of the URL after the pound-sign. Example: +"#hash"+ + +The following methods are provided by the URL module: + ++url.parse(urlStr, parseQueryString=false)+:: +Take a URL string, and return an object. Pass +true+ as the second argument to also parse +the query string using the +querystring+ module. + ++url.format(urlObj)+:: +Take a parsed URL object, and return a formatted URL string. + ++url.resolve(from, to)+:: +Take a base URL, and a href URL, and resolve them as a browser would for an anchor tag. + + +=== Query String Module + +This module provides utilities for dealing with query strings. It provides the following methods: + ++querystring.stringify(obj, sep="&", eq="=")+:: +Serialize an object to a query string. Optionally override the default separator and assignment characters. +Example: ++ +------------------------------------ +node> require("querystring").stringify({foo:"bar", baz : {quux:"asdf", oof : "rab"}, boo:[1,2,3]}) +"foo=bar&baz%5Bquux%5D=asdf&baz%5Boof%5D=rab&boo%5B%5D=1&boo%5B%5D=2&boo%5B%5D=3" +------------------------------------ ++ + ++querystring.parse(str, sep="&", eq="=")+:: +Deserialize a query string to an object. Optionally override the default separator and assignment characters. ++ +------------------------------------ +node> require("querystring").parse("foo=bar&baz%5Bquux%5D=asdf&baz%5Boof%5D=rab&boo%5B%5D=1") +{ + "foo": "bar", + "baz": { + "quux": "asdf", + "oof": "rab" + }, + "boo": [ + 1 + ] +} +------------------------------------ ++ + ++querystring.escape+ +The escape function used by +querystring.stringify+, provided so that it could be overridden if necessary. + ++querystring.unescape+ +The unescape function used by +querystring.parse+, provided so that it could be overridden if necessary. + == REPL A Read-Eval-Print-Loop is available both as a standalone program and easily diff --git a/lib/querystring.js b/lib/querystring.js new file mode 100644 index 00000000000000..d258fc297d7d95 --- /dev/null +++ b/lib/querystring.js @@ -0,0 +1,177 @@ +// Query String Utilities + +var QueryString = exports; + +QueryString.unescape = function (str, decodeSpaces) { + return decodeURIComponent(decodeSpaces ? str.replace(/\+/g, " ") : str); +}; + +QueryString.escape = function (str) { + return encodeURIComponent(str); +}; + + +var stack = []; +/** + *
Converts an arbitrary value to a Query String representation.
+ * + *Objects with cyclical references will trigger an exception.
+ * + * @method stringify + * @param obj {Variant} any arbitrary value to convert to query string + * @param sep {String} (optional) Character that should join param k=v pairs together. Default: "&" + * @param eq {String} (optional) Character that should join keys to their values. Default: "=" + * @param name {String} (optional) Name of the current key, for handling children recursively. + * @static + */ +QueryString.stringify = function (obj, sep, eq, name) { + sep = sep || "&"; + eq = eq || "="; + if (isA(obj, null) || isA(obj, undefined) || typeof(obj) === 'function') { + return name ? encodeURIComponent(name) + eq : ''; + } + + if (isBool(obj)) obj = +obj; + if (isNumber(obj) || isString(obj)) { + return encodeURIComponent(name) + eq + encodeURIComponent(obj); + } + if (isA(obj, [])) { + var s = []; + name = name+'[]'; + for (var i = 0, l = obj.length; i < l; i ++) { + s.push( QueryString.stringify(obj[i], sep, eq, name) ); + } + return s.join(sep); + } + // now we know it's an object. + + // Check for cyclical references in nested objects + for (var i = stack.length - 1; i >= 0; --i) if (stack[i] === obj) { + throw new Error("querystring.stringify. Cyclical reference"); + } + + stack.push(obj); + + var s = []; + var begin = name ? name + '[' : ''; + var end = name ? ']' : ''; + for (var i in obj) if (obj.hasOwnProperty(i)) { + var n = begin + i + end; + s.push(QueryString.stringify(obj[i], sep, eq, n)); + } + + stack.pop(); + + s = s.join(sep); + if (!s && name) return name + "="; + return s; +}; + +QueryString.parseQuery = QueryString.parse = function (qs, sep, eq) { + return qs + .split(sep||"&") + .map(pieceParser(eq||"=")) + .reduce(mergeParams); +}; + +// Parse a key=val string. +// These can get pretty hairy +// example flow: +// parse(foo[bar][][bla]=baz) +// return parse(foo[bar][][bla],"baz") +// return parse(foo[bar][], {bla : "baz"}) +// return parse(foo[bar], [{bla:"baz"}]) +// return parse(foo, {bar:[{bla:"baz"}]}) +// return {foo:{bar:[{bla:"baz"}]}} +var trimmerPattern = /^\s+|\s+$/g, + slicerPattern = /(.*)\[([^\]]*)\]$/; +var pieceParser = function (eq) { + return function parsePiece (key, val) { + if (arguments.length !== 2) { + // key=val, called from the map/reduce + key = key.split(eq); + return parsePiece( + QueryString.unescape(key.shift(), true), + QueryString.unescape(key.join(eq), true) + ); + } + key = key.replace(trimmerPattern, ''); + if (isString(val)) { + val = val.replace(trimmerPattern, ''); + // convert numerals to numbers + if (!isNaN(val)) { + var numVal = +val; + if (val === numVal.toString(10)) val = numVal; + } + } + var sliced = slicerPattern.exec(key); + if (!sliced) { + var ret = {}; + if (key) ret[key] = val; + return ret; + } + // ["foo[][bar][][baz]", "foo[][bar][]", "baz"] + var tail = sliced[2], head = sliced[1]; + + // array: key[]=val + if (!tail) return parsePiece(head, [val]); + + // obj: key[subkey]=val + var ret = {}; + ret[tail] = val; + return parsePiece(head, ret); + }; +}; + +// the reducer function that merges each query piece together into one set of params +function mergeParams (params, addition) { + return ( + // if it's uncontested, then just return the addition. + (!params) ? addition + // if the existing value is an array, then concat it. + : (isA(params, [])) ? params.concat(addition) + // if the existing value is not an array, and either are not objects, arrayify it. + : (!isA(params, {}) || !isA(addition, {})) ? [params].concat(addition) + // else merge them as objects, which is a little more complex + : mergeObjects(params, addition) + ); +}; + +// Merge two *objects* together. If this is called, we've already ruled +// out the simple cases, and need to do the for-in business. +function mergeObjects (params, addition) { + for (var i in addition) if (i && addition.hasOwnProperty(i)) { + params[i] = mergeParams(params[i], addition[i]); + } + return params; +}; + +// duck typing +function isA (thing, canon) { + return ( + // truthiness. you can feel it in your gut. + (!thing === !canon) + // typeof is usually "object" + && typeof(thing) === typeof(canon) + // check the constructor + && Object.prototype.toString.call(thing) === Object.prototype.toString.call(canon) + ); +}; +function isBool (thing) { + return ( + typeof(thing) === "boolean" + || isA(thing, new Boolean(thing)) + ); +}; +function isNumber (thing) { + return ( + typeof(thing) === "number" + || isA(thing, new Number(thing)) + ) && isFinite(thing); +}; +function isString (thing) { + return ( + typeof(thing) === "string" + || isA(thing, new String(thing)) + ); +}; diff --git a/lib/url.js b/lib/url.js new file mode 100644 index 00000000000000..dd82b831e5a7c4 --- /dev/null +++ b/lib/url.js @@ -0,0 +1,299 @@ + +exports.parse = url_parse; +exports.resolve = url_resolve; +exports.resolveObject = url_resolveObject; +exports.format = url_format; + +// define these here so at least they only have to be compiled once on the first module load. +var protocolPattern = /^([a-z0-9]+:)/, + portPattern = /:[0-9]+$/, + nonHostChars = ["/", "?", ";", "#"], + hostlessProtocol = { + "file":true, + "file:":true + }, + slashedProtocol = { + "http":true, "https":true, "ftp":true, "gopher":true, "file":true, + "http:":true, "https:":true, "ftp:":true, "gopher:":true, "file:":true + }, + path = require("path"), // internal module, guaranteed to be loaded already. + querystring; // don't load unless necessary. + +function url_parse (url, parseQueryString) { + if (url && typeof(url) === "object" && url.href) return url; + + var out = { href : url }, + rest = url; + + var proto = protocolPattern.exec(rest); + if (proto) { + proto = proto[0]; + out.protocol = proto; + rest = rest.substr(proto.length); + } + + // figure out if it's got a host + var slashes = rest.substr(0, 2) === "//"; + if (slashes && !(proto && hostlessProtocol[proto])) { + rest = rest.substr(2); + out.slashes = true; + } + if (!hostlessProtocol[proto] && (slashes || (proto && !slashedProtocol[proto]))) { + // there's a hostname. + // the first instance of /, ?, ;, or # ends the host. + // don't enforce full RFC correctness, just be unstupid about it. + var firstNonHost = -1; + for (var i = 0, l = nonHostChars.length; i < l; i ++) { + var index = rest.indexOf(nonHostChars[i]); + if (index !== -1 && (firstNonHost < 0 || index < firstNonHost)) firstNonHost = index; + } + if (firstNonHost !== -1) { + out.host = rest.substr(0, firstNonHost); + rest = rest.substr(firstNonHost); + } else { + out.host = rest; + rest = ""; + } + + // pull out the auth and port. + var p = parseHost(out.host); + for (var i in p) out[i] = p[i]; + // we've indicated that there is a hostname, so even if it's empty, it has to be present. + out.hostname = out.hostname || ""; + } + + // now rest is set to the post-host stuff. + // chop off from the tail first. + var hash = rest.indexOf("#"); + if (hash !== -1) { + // got a fragment string. + out.hash = rest.substr(hash); + rest = rest.slice(0, hash); + } + var qm = rest.indexOf("?"); + if (qm !== -1) { + out.search = rest.substr(qm); + out.query = rest.substr(qm+1); + if (parseQueryString) out.query = (querystring || querystring = require("querystring")).parse(out.query); + rest = rest.slice(0, qm); + } + if (rest) out.pathname = rest; + + return out; +}; + +// format a parsed object into a url string +function url_format (obj) { + // 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 = url_parse(obj); + + var protocol = obj.protocol || "", + host = (obj.host !== undefined) ? obj.host + : obj.hostname !== undefined ? ( + (obj.auth ? obj.auth + "@" : "") + + obj.hostname + + (obj.port ? ":" + obj.port : "") + ) + : false, + pathname = obj.pathname || "", + search = obj.search || ( + obj.query && ( "?" + ( + typeof(obj.query) === "object" + ? require("querystring").stringify(obj.query) + : String(obj.query) + )) + ) || "", + hash = obj.hash || ""; + + if (protocol && protocol.substr(-1) !== ":") protocol += ":"; + + // only the slashedProtocols get the //. Not mailto:, xmpp:, etc. + // unless they had them to begin with. + if (obj.slashes || (!protocol || slashedProtocol[protocol]) && host !== false) { + host = "//" + (host || ""); + if (pathname && pathname.charAt(0) !== "/") pathname = "/" + pathname; + } else if (!host) host = ""; + + if (hash && hash.charAt(0) !== "#") hash = "#" + hash; + if (search && search.charAt(0) !== "?") search = "?" + search; + + return protocol + host + pathname + search + hash; +}; + +function url_resolve (source, relative) { + return url_format(url_resolveObject(source, relative)); +}; + +function url_resolveObject (source, relative) { + if (!source) return relative; + + source = url_parse(url_format(source)); + relative = url_parse(url_format(relative)); + + // hash is always overridden, no matter what. + source.hash = relative.hash; + + if (relative.href === "") return source; + + // hrefs like //foo/bar always cut to the protocol. + if (relative.slashes && !relative.protocol) { + relative.protocol = source.protocol; + return relative; + } + + if (relative.protocol && relative.protocol !== source.protocol) { + // 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. + + if (!slashedProtocol[relative.protocol]) return relative; + + source.protocol = relative.protocol; + if (!relative.host && !hostlessProtocol[relative.protocol]) { + var relPath = (relative.pathname || "").split("/"); + while (relPath.length && !(relative.host = relPath.shift())); + if (!relative.host) relative.host = ""; + if (relPath[0] !== "") relPath.unshift(""); + if (relPath.length < 2) relPath.unshift(""); + relative.pathname = relPath.join("/"); + } + source.pathname = relative.pathname; + source.search = relative.search; + source.query = relative.query; + source.host = relative.host || ""; + delete source.auth; + delete source.hostname; + source.port = relative.port; + return source; + } + + var isSourceAbs = (source.pathname && source.pathname.charAt(0) === "/"), + isRelAbs = ( + relative.host !== undefined + || relative.pathname && relative.pathname.charAt(0) === "/" + ), + mustEndAbs = (isRelAbs || isSourceAbs || (source.host && relative.pathname)), + removeAllDots = mustEndAbs, + srcPath = source.pathname && source.pathname.split("/") || [], + relPath = relative.pathname && relative.pathname.split("/") || [], + psychotic = source.protocol && !slashedProtocol[source.protocol] && source.host !== undefined; + + // 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. + // source.protocol has already been set by now. + // Later on, put the first path part into the host field. + if ( psychotic ) { + + delete source.hostname; + delete source.auth; + delete source.port; + if (source.host) { + if (srcPath[0] === "") srcPath[0] = source.host; + else srcPath.unshift(source.host); + } + delete source.host; + + if (relative.protocol) { + delete relative.hostname; + delete relative.auth; + delete relative.port; + if (relative.host) { + if (relPath[0] === "") relPath[0] = relative.host; + else relPath.unshift(relative.host); + } + delete relative.host; + } + mustEndAbs = mustEndAbs && (relPath[0] === "" || srcPath[0] === ""); + } + + if (isRelAbs) { + // it's absolute. + source.host = (relative.host || relative.host === "") ? relative.host : source.host; + source.search = relative.search; + source.query = relative.query; + srcPath = relPath; + // fall through to the dot-handling below. + } else if (relPath.length) { + // it's relative + // throw away the existing file, and take the new path instead. + if (!srcPath) srcPath = []; + srcPath.pop(); + srcPath = srcPath.concat(relPath); + source.search = relative.search; + source.query = relative.query; + } else if ("search" in relative) { + // just pull out the search. + // like href="?foo". + // Put this after the other two cases because it simplifies the booleans + if (psychotic) { + source.host = srcPath.shift(); + } + source.search = relative.search; + source.query = relative.query; + return source; + } + if (!srcPath.length) { + // no path at all. easy. + // we've already handled the other stuff above. + delete source.pathname; + return source; + } + + // resolve dots. + // 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. + var last = srcPath.slice(-1)[0]; + var hasTrailingSlash = ( + (source.host || relative.host) && (last === "." || last === "..") + || last === "" + ); + + // Figure out if this has to end up as an absolute url, or should continue to be relative. + srcPath = path.normalizeArray(srcPath, true); + if (srcPath.length === 1 && srcPath[0] === ".") srcPath = []; + if (mustEndAbs || removeAllDots) { + // all dots must go. + var dirs = []; + srcPath.forEach(function (dir, i) { + if (dir === "..") dirs.pop(); + else if (dir !== ".") dirs.push(dir); + }); + + if (mustEndAbs && dirs[0] !== "") { + dirs.unshift(""); + } + srcPath = dirs; + } + if (hasTrailingSlash && (srcPath.length < 2 || srcPath.slice(-1)[0] !== "")) srcPath.push(""); + + // put the host back + if ( psychotic ) source.host = srcPath[0] === "" ? "" : srcPath.shift(); + + mustEndAbs = mustEndAbs || (source.host && srcPath.length); + + if (mustEndAbs && srcPath[0] !== "") srcPath.unshift("") + + source.pathname = srcPath.join("/"); + + return source; +}; + +function parseHost (host) { + var out = {}; + var at = host.indexOf("@"); + if (at !== -1) { + out.auth = host.substr(0, at); + host = host.substr(at+1); // drop the @ + } + var port = portPattern.exec(host); + if (port) { + port = port[0]; + out.port = port.substr(1); + host = host.substr(0, host.length - port.length); + } + if (host) out.hostname = host; + return out; +} diff --git a/test/mjsunit/test-querystring.js b/test/mjsunit/test-querystring.js new file mode 100644 index 00000000000000..19f7ee3902c209 --- /dev/null +++ b/test/mjsunit/test-querystring.js @@ -0,0 +1,125 @@ +process.mixin(require("./common")); + +// test using assert + +var qs = require("querystring"); + +// folding block. +{ +// [ wonkyQS, canonicalQS, obj ] +var qsTestCases = [ + ["foo=bar", "foo=bar", {"foo" : "bar"}], + ["foo=bar&foo=quux", "foo%5B%5D=bar&foo%5B%5D=quux", {"foo" : ["bar", "quux"]}], + ["foo=1&bar=2", "foo=1&bar=2", {"foo" : 1, "bar" : 2}], + ["my+weird+field=q1%212%22%27w%245%267%2Fz8%29%3F", "my%20weird%20field=q1!2%22'w%245%267%2Fz8)%3F", {"my weird field" : "q1!2\"'w$5&7/z8)?" }], + ["foo%3Dbaz=bar", "foo%3Dbaz=bar", {"foo=baz" : "bar"}], + ["foo=baz=bar", "foo=baz%3Dbar", {"foo" : "baz=bar"}], + [ "str=foo&arr[]=1&arr[]=2&arr[]=3&obj[a]=bar&obj[b][]=4&obj[b][]=5&obj[b][]=6&obj[b][]=&obj[c][]=4&obj[c][]=5&obj[c][][somestr]=baz&obj[objobj][objobjstr]=blerg&somenull=&undef=", "str=foo&arr%5B%5D=1&arr%5B%5D=2&arr%5B%5D=3&obj%5Ba%5D=bar&obj%5Bb%5D%5B%5D=4&obj%5Bb%5D%5B%5D=5&obj%5Bb%5D%5B%5D=6&obj%5Bb%5D%5B%5D=&obj%5Bc%5D%5B%5D=4&obj%5Bc%5D%5B%5D=5&obj%5Bc%5D%5B%5D%5Bsomestr%5D=baz&obj%5Bobjobj%5D%5Bobjobjstr%5D=blerg&somenull=&undef=", { + "str":"foo", + "arr":[1,2,3], + "obj":{ + "a":"bar", + "b":[4,5,6,""], + "c":[4,5,{"somestr":"baz"}], + "objobj":{"objobjstr":"blerg"} + }, + "somenull":"", + "undef":"" + }], + ["foo[bar][bla]=baz&foo[bar][bla]=blo", "foo%5Bbar%5D%5Bbla%5D%5B%5D=baz&foo%5Bbar%5D%5Bbla%5D%5B%5D=blo", {"foo":{"bar":{"bla":["baz","blo"]}}}], + ["foo[bar][][bla]=baz&foo[bar][][bla]=blo", "foo%5Bbar%5D%5B%5D%5Bbla%5D=baz&foo%5Bbar%5D%5B%5D%5Bbla%5D=blo", {"foo":{"bar":[{"bla":"baz"},{"bla":"blo"}]}}], + ["foo[bar][bla][]=baz&foo[bar][bla][]=blo", "foo%5Bbar%5D%5Bbla%5D%5B%5D=baz&foo%5Bbar%5D%5Bbla%5D%5B%5D=blo", {"foo":{"bar":{"bla":["baz","blo"]}}}], + [" foo = bar ", "foo=bar", {"foo":"bar"}] +]; + +// [ wonkyQS, canonicalQS, obj ] +var qsColonTestCases = [ + ["foo:bar", "foo:bar", {"foo":"bar"}], + ["foo:bar;foo:quux", "foo%5B%5D:bar;foo%5B%5D:quux", {"foo" : ["bar", "quux"]}], + ["foo:1&bar:2;baz:quux", "foo:1%26bar%3A2;baz:quux", {"foo":"1&bar:2", "baz":"quux"}], + ["foo%3Abaz:bar", "foo%3Abaz:bar", {"foo:baz":"bar"}], + ["foo:baz:bar", "foo:baz%3Abar", {"foo":"baz:bar"}] +]; + +// [ wonkyObj, qs, canonicalObj ] +var extendedFunction = function () {}; +extendedFunction.prototype = {a:"b"}; +var qsWeirdObjects = [ + [ {regexp:/./g}, "regexp=", {"regexp":""} ], + [ {regexp: new RegExp(".", "g")}, "regexp=", {"regexp":""} ], + [ {fn:function () {}}, "fn=", {"fn":""}], + [ {fn:new Function("")}, "fn=", {"fn":""} ], + [ {math:Math}, "math=", {"math":""} ], + [ {e:extendedFunction}, "e=", {"e":""} ], + [ {d:new Date()}, "d=", {"d":""} ], + [ {d:Date}, "d=", {"d":""} ], + [ {f:new Boolean(false), t:new Boolean(true)}, "f=0&t=1", {"f":0, "t":1} ], + [ {f:false, t:true}, "f=0&t=1", {"f":0, "t":1} ], +]; +} + +// test that the canonical qs is parsed properly. +qsTestCases.forEach(function (testCase) { + assert.deepEqual(testCase[2], qs.parse(testCase[0])); +}); + +// test that the colon test cases can do the same +qsColonTestCases.forEach(function (testCase) { + assert.deepEqual(testCase[2], qs.parse(testCase[0], ";", ":")); +}); + +// test the weird objects, that they get parsed properly +qsWeirdObjects.forEach(function (testCase) { + assert.deepEqual(testCase[2], qs.parse(testCase[1])); +}); + +// test the nested qs-in-qs case +var f = qs.parse("a=b&q=x%3Dy%26y%3Dz"); +f.q = qs.parse(f.q); +assert.deepEqual(f, { a : "b", q : { x : "y", y : "z" } }); + +// nested in colon +var f = qs.parse("a:b;q:x%3Ay%3By%3Az", ";", ":"); +f.q = qs.parse(f.q, ";", ":"); +assert.deepEqual(f, { a : "b", q : { x : "y", y : "z" } }); + + +// now test stringifying +assert.throws(function () { + var f = {}; + f.f = f; + qs.stringify(f); +}); + +// basic +qsTestCases.forEach(function (testCase) { + assert.equal(testCase[1], qs.stringify(testCase[2])); +}); + +qsColonTestCases.forEach(function (testCase) { + assert.equal(testCase[1], qs.stringify(testCase[2], ";", ":")); +}); + +qsWeirdObjects.forEach(function (testCase) { + assert.equal(testCase[1], qs.stringify(testCase[0])); +}); + +// nested +var f = qs.stringify({ + a : "b", + q : qs.stringify({ + x : "y", + y : "z" + }) +}); +assert.equal(f, "a=b&q=x%3Dy%26y%3Dz"); + +// nested in colon +var f = qs.stringify({ + a : "b", + q : qs.stringify({ + x : "y", + y : "z" + }, ";", ":") +}, ";", ":"); +assert.equal(f, "a:b;q:x%3Ay%3By%3Az"); diff --git a/test/mjsunit/test-url.js b/test/mjsunit/test-url.js new file mode 100644 index 00000000000000..dfdd28777c5b00 --- /dev/null +++ b/test/mjsunit/test-url.js @@ -0,0 +1,495 @@ +process.mixin(require("./common")); + +var url = require("url"), + sys = require("sys"); + +// URLs to parse, and expected data +// { url : parsed } +var parseTests = { + "http://www.narwhaljs.org/blog/categories?id=news" : { + "href": "http://www.narwhaljs.org/blog/categories?id=news", + "protocol": "http:", + "host": "www.narwhaljs.org", + "hostname": "www.narwhaljs.org", + "search": "?id=news", + "query": "id=news", + "pathname": "/blog/categories" + }, + "http://mt0.google.com/vt/lyrs=m@114&hl=en&src=api&x=2&y=2&z=3&s=" : { + "href": "http://mt0.google.com/vt/lyrs=m@114&hl=en&src=api&x=2&y=2&z=3&s=", + "protocol": "http:", + "host": "mt0.google.com", + "hostname": "mt0.google.com", + "pathname": "/vt/lyrs=m@114&hl=en&src=api&x=2&y=2&z=3&s=" + }, + "http://mt0.google.com/vt/lyrs=m@114???&hl=en&src=api&x=2&y=2&z=3&s=" : { + "href": "http://mt0.google.com/vt/lyrs=m@114???&hl=en&src=api&x=2&y=2&z=3&s=", + "protocol": "http:", + "host": "mt0.google.com", + "hostname": "mt0.google.com", + "search": "???&hl=en&src=api&x=2&y=2&z=3&s=", + "query": "??&hl=en&src=api&x=2&y=2&z=3&s=", + "pathname": "/vt/lyrs=m@114" + }, + "http://user:pass@mt0.google.com/vt/lyrs=m@114???&hl=en&src=api&x=2&y=2&z=3&s=" : { + "href": "http://user:pass@mt0.google.com/vt/lyrs=m@114???&hl=en&src=api&x=2&y=2&z=3&s=", + "protocol": "http:", + "host": "user:pass@mt0.google.com", + "auth": "user:pass", + "hostname": "mt0.google.com", + "search": "???&hl=en&src=api&x=2&y=2&z=3&s=", + "query": "??&hl=en&src=api&x=2&y=2&z=3&s=", + "pathname": "/vt/lyrs=m@114" + }, + "file:///etc/passwd" : { + "href": "file:///etc/passwd", + "protocol": "file:", + "pathname": "///etc/passwd" + }, + "file:///etc/node/" : { + "href": "file:///etc/node/", + "protocol": "file:", + "pathname": "///etc/node/" + }, + "http:/baz/../foo/bar" : { + "href": "http:/baz/../foo/bar", + "protocol": "http:", + "pathname": "/baz/../foo/bar" + }, + "http://user:pass@example.com:8000/foo/bar?baz=quux#frag" : { + "href": "http://user:pass@example.com:8000/foo/bar?baz=quux#frag", + "protocol": "http:", + "host": "user:pass@example.com:8000", + "auth": "user:pass", + "port": "8000", + "hostname": "example.com", + "hash": "#frag", + "search": "?baz=quux", + "query": "baz=quux", + "pathname": "/foo/bar" + }, + "//user:pass@example.com:8000/foo/bar?baz=quux#frag" : { + "href": "//user:pass@example.com:8000/foo/bar?baz=quux#frag", + "host": "user:pass@example.com:8000", + "auth": "user:pass", + "port": "8000", + "hostname": "example.com", + "hash": "#frag", + "search": "?baz=quux", + "query": "baz=quux", + "pathname": "/foo/bar" + }, + "http://example.com?foo=bar#frag" : { + "href": "http://example.com?foo=bar#frag", + "protocol": "http:", + "host": "example.com", + "hostname": "example.com", + "hash": "#frag", + "search": "?foo=bar", + "query": "foo=bar" + }, + "http://example.com?foo=@bar#frag" : { + "href": "http://example.com?foo=@bar#frag", + "protocol": "http:", + "host": "example.com", + "hostname": "example.com", + "hash": "#frag", + "search": "?foo=@bar", + "query": "foo=@bar" + }, + "http://example.com?foo=/bar/#frag" : { + "href": "http://example.com?foo=/bar/#frag", + "protocol": "http:", + "host": "example.com", + "hostname": "example.com", + "hash": "#frag", + "search": "?foo=/bar/", + "query": "foo=/bar/" + }, + "http://example.com?foo=?bar/#frag" : { + "href": "http://example.com?foo=?bar/#frag", + "protocol": "http:", + "host": "example.com", + "hostname": "example.com", + "hash": "#frag", + "search": "?foo=?bar/", + "query": "foo=?bar/" + }, + "http://example.com#frag=?bar/#frag" : { + "href": "http://example.com#frag=?bar/#frag", + "protocol": "http:", + "host": "example.com", + "hostname": "example.com", + "hash": "#frag=?bar/#frag" + }, + "/foo/bar?baz=quux#frag" : { + "href": "/foo/bar?baz=quux#frag", + "hash": "#frag", + "search": "?baz=quux", + "query": "baz=quux", + "pathname": "/foo/bar" + }, + "http:/foo/bar?baz=quux#frag" : { + "href": "http:/foo/bar?baz=quux#frag", + "protocol": "http:", + "hash": "#frag", + "search": "?baz=quux", + "query": "baz=quux", + "pathname": "/foo/bar" + }, + "mailto:foo@bar.com?subject=hello" : { + "href": "mailto:foo@bar.com?subject=hello", + "protocol": "mailto:", + "host": "foo@bar.com", + "auth" : "foo", + "hostname" : "bar.com", + "search": "?subject=hello", + "query": "subject=hello" + }, + "javascript:alert('hello');" : { + "href": "javascript:alert('hello');", + "protocol": "javascript:", + "host": "alert('hello')", + "hostname": "alert('hello')", + "pathname" : ";" + }, + "xmpp:isaacschlueter@jabber.org" : { + "href": "xmpp:isaacschlueter@jabber.org", + "protocol": "xmpp:", + "host": "isaacschlueter@jabber.org", + "auth": "isaacschlueter", + "hostname": "jabber.org" + } +}; +for (var u in parseTests) { + var actual = url.parse(u), + expected = parseTests[u]; + for (var i in expected) { + var e = JSON.stringify(expected[i]), + a = JSON.stringify(actual[i]); + assert.equal(e, a, "parse(" + u + ")."+i+" == "+e+"\nactual: "+a); + } + + var expected = u, + actual = url.format(parseTests[u]); + + assert.equal(expected, actual, "format("+u+") == "+u+"\nactual:"+actual); +} + +// some extra formatting tests, just to verify that it'll format slightly wonky content to a valid url. +var formatTests = { + "http://a.com/a/b/c?s#h" : { + "protocol": "http", + "host": "a.com", + "pathname": "a/b/c", + "hash": "h", + "search": "s" + }, + "xmpp:isaacschlueter@jabber.org" : { + "href": "xmpp://isaacschlueter@jabber.org", + "protocol": "xmpp:", + "host": "isaacschlueter@jabber.org", + "auth": "isaacschlueter", + "hostname": "jabber.org" + } +}; +for (var u in formatTests) { + var actual = url.format(formatTests[u]); + assert.equal(actual, u, "wonky format("+u+") == "+u+"\nactual:"+actual); +} + +[ + // [from, path, expected] + ["/foo/bar/baz", "quux", "/foo/bar/quux"], + ["/foo/bar/baz", "quux/asdf", "/foo/bar/quux/asdf"], + ["/foo/bar/baz", "quux/baz", "/foo/bar/quux/baz"], + ["/foo/bar/baz", "../quux/baz", "/foo/quux/baz"], + ["/foo/bar/baz", "/bar", "/bar"], + ["/foo/bar/baz/", "quux", "/foo/bar/baz/quux"], + ["/foo/bar/baz/", "quux/baz", "/foo/bar/baz/quux/baz"], + ["/foo/bar/baz", "../../../../../../../../quux/baz", "/quux/baz"], + ["/foo/bar/baz", "../../../../../../../quux/baz", "/quux/baz"], + ["foo/bar", "../../../baz", "../../baz"], + ["foo/bar/", "../../../baz", "../baz"], + ["http://example.com/b//c//d;p?q#blarg","https:#hash2","https:///#hash2" ], + ["http://example.com/b//c//d;p?q#blarg","https:/p/a/t/h?s#hash2","https://p/a/t/h?s#hash2" ], + ["http://example.com/b//c//d;p?q#blarg","https://u:p@h.com/p/a/t/h?s#hash2","https://u:p@h.com/p/a/t/h?s#hash2"], + ["http://example.com/b//c//d;p?q#blarg","https:/a/b/c/d","https://a/b/c/d"], + ["http://example.com/b//c//d;p?q#blarg","http:#hash2","http://example.com/b//c//d;p?q#hash2" ], + ["http://example.com/b//c//d;p?q#blarg","http:/p/a/t/h?s#hash2","http://example.com/p/a/t/h?s#hash2" ], + ["http://example.com/b//c//d;p?q#blarg","http://u:p@h.com/p/a/t/h?s#hash2","http://u:p@h.com/p/a/t/h?s#hash2" ], + ["http://example.com/b//c//d;p?q#blarg","http:/a/b/c/d","http://example.com/a/b/c/d"], + ["/foo/bar/baz", "/../etc/passwd", "/etc/passwd"] +].forEach(function (relativeTest) { + var a = url.resolve(relativeTest[0], relativeTest[1]), + e = relativeTest[2]; + assert.equal(e, a, + "resolve("+[relativeTest[0], relativeTest[1]]+") == "+e+ + "\n actual="+a); +}); + + +// +// Tests below taken from Chiron +// http://code.google.com/p/chironjs/source/browse/trunk/src/test/http/url.js +// +// Copyright (c) 2002-2008 Kris Kowal