Hatena::Groupmoz-addon

Ci.nsIZIGOROu

2008-05-07ふー

XPCShellを使って特定のinterfaceのスタブを作る

| 20:12 |  XPCShellを使って特定のinterfaceのスタブを作る - Ci.nsIZIGOROu を含むブックマーク はてなブックマーク -  XPCShellを使って特定のinterfaceのスタブを作る - Ci.nsIZIGOROu

ベースはd:id:brazilさんのoct inaoduですが、コマンドラインとかでやれたらウマーかなと思ってXPCShellでやってみる事にしました。

nsIScriptableInterfaceInfoを使ってるんで、Gecko 1.9未満じゃないと動かないと思います。

ソース

no titleのソースを流用してるんで、コードが無駄に長いです。

const Cc = Components.classes;
const Ci = Components.interfaces;
const EXCLUDE_METHODS = { AddRef: 1, Release: 1, QueryInterface: 1 }; // nsISupports

/*
 JS Beautifier
---------------
  $Date: 2008-04-21 16:13:36 +0300 (Mon, 21 Apr 2008) $
  $Revision: 53 $


  Written by Einars "elfz" Lielmanis, <elfz@laacz.lv>
      http://elfz.laacz.lv/beautify/

  Originally converted to javascript by Vital, <vital76@gmail.com>
      http://my.opera.com/Vital/blog/2007/11/21/javascript-beautify-on-javascript-translated


  You are free to use this in any way you want, in case you find this useful or working for you.

  Usage:
    js_beautify(js_source_text);
*/


function js_beautify(js_source_text, indent_size, indent_character)
{

    var input, output, token_text, last_type, last_text, last_word, current_mode, modes, indent_level, indent_string;
    var whitespace, wordchar, punct, parser_pos, line_starters, in_case;
    var prefix, token_type;

    function print_newline(ignore_repeated)
    {
        ignore_repeated = typeof ignore_repeated === 'undefined' ? true: ignore_repeated;

        // remove trailing whitespace and indent
        while (output.length && (output[output.length - 1] === ' ' || output[output.length - 1] === indent_string)) {
            output.pop();
        }

        if (!output.length) {
            return; // no newline on start of file
        }

        if (output[output.length - 1] !== "\n" || !ignore_repeated) {
            output.push("\n");
        }
        for (var i = 0; i < indent_level; i++) {
            output.push(indent_string);
        }
    }



    function print_space()
    {
        var last_output = output.length ? output[output.length - 1] : ' ';
        if (last_output !== ' ' && last_output !== '\n' && last_output !== indent_string) { // prevent occassional duplicate space
            output.push(' ');
        }
    }


    function print_token()
    {
        output.push(token_text);
    }

    function indent()
    {
        indent_level++;
    }


    function unindent()
    {
        if (indent_level) {
            indent_level--;
        }
    }


    function remove_indent()
    {
        if (output.length && output[output.length - 1] === indent_string) {
            output.pop();
        }
    }


    function set_mode(mode)
    {
        modes.push(current_mode);
        current_mode = mode;
    }


    function restore_mode()
    {
        current_mode = modes.pop();
    }


    function in_array(what, arr)
    {
        for (var i = 0; i < arr.length; i++)
        {
            if (arr[i] === what) {
                return true;
            }
        }
        return false;
    }



    function get_next_token()
    {
        var n_newlines = 0;
        var c = '';

        do {
            if (parser_pos >= input.length) {
                return ['', 'TK_EOF'];
            }
            c = input.charAt(parser_pos);

            parser_pos += 1;
            if (c === "\n") {
                n_newlines += 1;
            }
        }
        while (in_array(c, whitespace));

        if (n_newlines > 1) {
            for (var i = 0; i < 2; i++) {
                print_newline(i === 0);
            }
        }
        var wanted_newline = (n_newlines === 1);


        if (in_array(c, wordchar)) {
            if (parser_pos < input.length) {
                while (in_array(input.charAt(parser_pos), wordchar)) {
                    c += input.charAt(parser_pos);
                    parser_pos += 1;
                    if (parser_pos === input.length) {
                        break;
                    }
                }
            }

            // small and surprisingly unugly hack for 1E-10 representation
            if (parser_pos !== input.length && c.match(/^[0-9]+[Ee]$/) && input.charAt(parser_pos) === '-') {
                parser_pos += 1;

                var t = get_next_token(parser_pos);
                c += '-' + t[0];
                return [c, 'TK_WORD'];
            }

            if (c === 'in') { // hack for 'in' operator
                return [c, 'TK_OPERATOR'];
            }
            return [c, 'TK_WORD'];
        }

        if (c === '(' || c === '[') {
            return [c, 'TK_START_EXPR'];
        }

        if (c === ')' || c === ']') {
            return [c, 'TK_END_EXPR'];
        }

        if (c === '{') {
            return [c, 'TK_START_BLOCK'];
        }

        if (c === '}') {
            return [c, 'TK_END_BLOCK'];
        }

        if (c === ';') {
            return [c, 'TK_END_COMMAND'];
        }

        if (c === '/') {
            var comment = '';
            // peek for comment /* ... */
            if (input.charAt(parser_pos) === '*') {
                parser_pos += 1;
                if (parser_pos < input.length) {
                    while (! (input.charAt(parser_pos) === '*' && input.charAt(parser_pos + 1) && input.charAt(parser_pos + 1) === '/') && parser_pos < input.length) {
                        comment += input.charAt(parser_pos);
                        parser_pos += 1;
                        if (parser_pos >= input.length) {
                            break;
                        }
                    }
                }
                parser_pos += 2;
                return ['/*' + comment + '*/', 'TK_BLOCK_COMMENT'];
            }
            // peek for comment // ...
            if (input.charAt(parser_pos) === '/') {
                comment = c;
                while (input.charAt(parser_pos) !== "\x0d" && input.charAt(parser_pos) !== "\x0a") {
                    comment += input.charAt(parser_pos);
                    parser_pos += 1;
                    if (parser_pos >= input.length) {
                        break;
                    }
                }
                parser_pos += 1;
                if (wanted_newline) {
                    print_newline();
                }
                return [comment, 'TK_COMMENT'];
            }

        }

        if (c === "'" || // string
        c === '"' || // string
        (c === '/' &&
        ((last_type === 'TK_WORD' && last_text === 'return') || (last_type === 'TK_START_EXPR' || last_type === 'TK_END_BLOCK' || last_type === 'TK_OPERATOR' || last_type === 'TK_EOF' || last_type === 'TK_END_COMMAND')))) { // regexp
            var sep = c;
            var esc = false;
            c = '';

            if (parser_pos < input.length) {

                while (esc || input.charAt(parser_pos) !== sep) {
                    c += input.charAt(parser_pos);
                    if (!esc) {
                        esc = input.charAt(parser_pos) === '\\';
                    } else {
                        esc = false;
                    }
                    parser_pos += 1;
                    if (parser_pos >= input.length) {
                        break;
                    }
                }

            }

            parser_pos += 1;
            if (last_type === 'TK_END_COMMAND') {
                print_newline();
            }
            return [sep + c + sep, 'TK_STRING'];
        }

        if (in_array(c, punct)) {
            while (parser_pos < input.length && in_array(c + input.charAt(parser_pos), punct)) {
                c += input.charAt(parser_pos);
                parser_pos += 1;
                if (parser_pos >= input.length) {
                    break;
                }
            }
            return [c, 'TK_OPERATOR'];
        }

        return [c, 'TK_UNKNOWN'];
    }


    //----------------------------------

    indent_character = indent_character || ' ';
    indent_size = indent_size || 4;

    indent_string = '';
    while (indent_size--) {
        indent_string += indent_character;
    }

    input = js_source_text;

    last_word = ''; // last 'TK_WORD' passed
    last_type = 'TK_START_EXPR'; // last token type
    last_text = ''; // last token text
    output = [];

    whitespace = "\n\r\t ".split('');
    wordchar = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_$'.split('');
    punct = '+ - * / % & ++ -- = += -= *= /= %= == === != !== > < >= <= >> << >>> >>>= >>= <<= && &= | || ! !! , : ? ^ ^= |='.split(' ');

    // words which should always start on new line.
    line_starters = 'continue,try,throw,return,var,if,switch,case,default,for,while,break,function'.split(',');

    // states showing if we are currently in expression (i.e. "if" case) - 'EXPRESSION', or in usual block (like, procedure), 'BLOCK'.
    // some formatting depends on that.
    current_mode = 'BLOCK';
    modes = [current_mode];

    indent_level = 0;
    parser_pos = 0; // parser position
    in_case = false; // flag for parser that case/default has been processed, and next colon needs special attention
    while (true) {
        var t = get_next_token(parser_pos);
        token_text = t[0];
        token_type = t[1];
        if (token_type === 'TK_EOF') {
            break;
        }

        switch (token_type) {

        case 'TK_START_EXPR':

            set_mode('EXPRESSION');
            if (last_type === 'TK_END_EXPR' || last_type === 'TK_START_EXPR') {
                // do nothing on (( and )( and ][ and ]( ..
            } else if (last_type !== 'TK_WORD' && last_type !== 'TK_OPERATOR') {
                print_space();
            } else if (in_array(last_word, line_starters) && last_word !== 'function') {
                print_space();
            }
            print_token();
            break;

        case 'TK_END_EXPR':

            print_token();
            restore_mode();
            break;

        case 'TK_START_BLOCK':

            set_mode('BLOCK');
            if (last_type !== 'TK_OPERATOR' && last_type !== 'TK_START_EXPR') {
                if (last_type === 'TK_START_BLOCK') {
                    print_newline();
                } else {
                    print_space();
                }
            }
            print_token();
            indent();
            break;

        case 'TK_END_BLOCK':
            if (last_type === 'TK_START_BLOCK') {
                // nothing
                unindent();
            } else {
                unindent();
                print_newline();
            }
            print_token();
            restore_mode();
            break;

        case 'TK_WORD':

            if (token_text === 'case' || token_text === 'default') {
                if (last_text === ':') {
                    // switch cases following one another
                    remove_indent();
                } else {
                    // case statement starts in the same line where switch
                    unindent();
                    print_newline();
                    indent();
                }
                print_token();
                in_case = true;
                break;
            }

            prefix = 'NONE';
            if (last_type === 'TK_END_BLOCK') {
                if (!in_array(token_text.toLowerCase(), ['else', 'catch', 'finally'])) {
                    prefix = 'NEWLINE';
                } else {
                    prefix = 'SPACE';
                    print_space();
                }
            } else if (last_type === 'TK_END_COMMAND' && current_mode === 'BLOCK') {
                prefix = 'NEWLINE';
            } else if (last_type === 'TK_END_COMMAND' && current_mode === 'EXPRESSION') {
                prefix = 'SPACE';
            } else if (last_type === 'TK_WORD') {
                prefix = 'SPACE';
            } else if (last_type === 'TK_START_BLOCK') {
                prefix = 'NEWLINE';
            } else if (last_type === 'TK_END_EXPR') {
                print_space();
                prefix = 'NEWLINE';
            }

            if (in_array(token_text, line_starters) || prefix === 'NEWLINE') {

                if (last_text === 'else') {
                    // no need to force newline on else break
                    print_space();
                } else if ((last_type === 'TK_START_EXPR' || last_text === '=') && token_text === 'function') {
                    // no need to force newline on 'function': (function
                    // DONOTHING
                } else if (last_type === 'TK_WORD' && (last_text === 'return' || last_text === 'throw')) {
                    // no newline between 'return nnn'
                    print_space();
                } else if (last_type !== 'TK_END_EXPR') {
                    if ((last_type !== 'TK_START_EXPR' || token_text !== 'var') && last_text !== ':') {
                        // no need to force newline on 'var': for (var x = 0...)
                        if (token_text === 'if' && last_type === 'TK_WORD' && last_word === 'else') {
                            // no newline for } else if {
                            print_space();
                        } else {
                            print_newline();
                        }
                    }
                }
            } else if (prefix === 'SPACE') {
                print_space();
            }
            print_token();
            last_word = token_text;
            break;

        case 'TK_END_COMMAND':

            print_token();
            break;

        case 'TK_STRING':

            if (last_type === 'TK_START_BLOCK' || last_type === 'TK_END_BLOCK') {
                print_newline();
            } else if (last_type === 'TK_WORD') {
                print_space();
            }
            print_token();
            break;

        case 'TK_OPERATOR':

            var start_delim = true;
            var end_delim = true;

            if (token_text === ':' && in_case) {
                print_token(); // colon really asks for separate treatment
                print_newline();
                break;
            }

            in_case = false;

            if (token_text === ',') {
                if (last_type === 'TK_END_BLOCK') {
                    print_token();
                    print_newline();
                } else {
                    if (current_mode === 'BLOCK') {
                        print_token();
                        print_newline();
                    } else {
                        print_token();
                        print_space();
                    }
                }
                break;
            } else if (token_text === '--' || token_text === '++') { // unary operators special case
                if (last_text === ';') {
                    // space for (;; ++i)
                    start_delim = true;
                    end_delim = false;
                } else {
                    start_delim = false;
                    end_delim = false;
                }
            } else if (token_text === '!' && last_type === 'TK_START_EXPR') {
                // special case handling: if (!a)
                start_delim = false;
                end_delim = false;
            } else if (last_type === 'TK_OPERATOR') {
                start_delim = false;
                end_delim = false;
            } else if (last_type === 'TK_END_EXPR') {
                start_delim = true;
                end_delim = true;
            } else if (token_text === '.') {
                // decimal digits or object.property
                start_delim = false;
                end_delim = false;

            } else if (token_text === ':') {
                // zz: xx
                // can't differentiate ternary op, so for now it's a ? b: c; without space before colon
                start_delim = false;
            }
            if (start_delim) {
                print_space();
            }

            print_token();

            if (end_delim) {
                print_space();
            }
            break;

        case 'TK_BLOCK_COMMENT':

            print_newline();
            print_token();
            print_newline();
            break;

        case 'TK_COMMENT':

            // print_newline();
            print_space();
            print_token();
            print_newline();
            break;

        case 'TK_UNKNOWN':
            print_token();
            break;
        }

        last_type = token_type;
        last_text = token_text;
    }

    return output.join('');

}

Object.prototype.nativeToSource = Object.prototype.toSource;
Object.prototype.toSource = function() {
  var source = this.nativeToSource();
  return source.substr(1, source.length - 2);
};

Function.prototype.nativeToSource = Function.prototype.toSource;
Function.prototype.toSource = function() {
  var source = this.nativeToSource();
  return source.substr(1, source.length - 2);
};

var ifaces = Array.prototype.slice.call(arguments).filter(function(ifaceName) {
  if (ifaceName in Ci)
	return true;
  else
	return false;
}).map(function(ifaceName) {
  return Ci[ifaceName];
});

if (ifaces.length == 0) {
  print("Please specify interface names.");
  quit();
}

var unique = {};
var Mock = function() { };

(function (ifaces) {
   var createIfaceMethods = function(iface) {
	 if (!iface || unique[iface.name] || iface.name == "nsISupports")
	   return;

	 var iinfo = Cc["@mozilla.org/scriptableInterfaceInfo;1"].createInstance(Ci.nsIScriptableInterfaceInfo);

	 try { iinfo.init(iface); }
	 catch(e) { return; }

	 for (var i = 0, mlen = iinfo.methodCount; i < mlen; i++) {
		var minfo = iinfo.getMethodInfo(i);
		if (minfo.name in EXCLUDE_METHODS) continue;
		Mock.prototype[minfo.name] = function() {};
	 }

	 unique[iface.name] = iface;

	 if (iinfo.parent) {
	   arguments.callee(iinfo.parent);
	 }
   };

   ifaces.forEach(createIfaceMethods);
})(ifaces);

var source = "";

source += "var Mock = function() {};\n";
source += "Mock.prototype = " + Mock.prototype.toSource() + ";";

print(js_beautify(source, 2));

使い方

$ /path/to/run-mozilla.sh /path/to/xpcshell mock.js [interfaces]

みたいな感じ。

サンプル

nsIFileURLでやった場合。

var Mock = function() {};
Mock.prototype = {
  spec: function() {},
  prePath: function() {},
  scheme: function() {},
  userPass: function() {},
  username: function() {},
  password: function() {},
  hostPort: function() {},
  host: function() {},
  port: function() {},
  path: function() {},
  equals: function() {},
  schemeIs: function() {},
  clone: function() {},
  resolve: function() {},
  asciiSpec: function() {},
  asciiHost: function() {},
  originCharset: function() {},
  filePath: function() {},
  param: function() {},
  query: function() {},
  ref: function() {},
  directory: function() {},
  fileName: function() {},
  fileBaseName: function() {},
  fileExtension: function() {},
  getCommonBaseSpec: function() {},
  getRelativeSpec: function() {},
  file: function() {}
};

TODO

  • CodeReposに突っ込む
  • functionのargsも生成すべきか? (nsIScriptableMethodInfo.getParam(), nsIScriptableParamInfo)
  • Mockじゃない名前でも生成出来るようにする?

とかとか。

idlのパーサ作って、そこら辺からgenerateした方が汎用的っぽぃなー。

参考リンク

brazilbrazil2008/05/07 20:27このバグに流暢な英語でつっこんであげて!!! https://bugzilla.mozilla.org/show_bug.cgi?id=392004

ZIGOROuZIGOROu2008/05/07 22:48おー、もうassignされてんじゃんすかー。
nsIScriptableParamInfoで名前取れたらいいのになー。
idlのパーサー、そのうち作りますかね。