mediawiki-extensions-Visual.../modules/parser/mediawiki.parser.environment.js

447 lines
12 KiB
JavaScript
Raw Normal View History

var title = require('./mediawiki.Title.js'),
Title = title.Title,
Namespace = title.Namespace;
var wikipedias = "en|de|fr|nl|it|pl|es|ru|ja|pt|zh|sv|vi|uk|ca|no|fi|cs|hu|ko|fa|id|tr|ro|ar|sk|eo|da|sr|lt|ms|eu|he|sl|bg|kk|vo|war|hr|hi|et|az|gl|simple|nn|la|th|el|new|roa-rup|oc|sh|ka|mk|tl|ht|pms|te|ta|be-x-old|ceb|br|be|lv|sq|jv|mg|cy|lb|mr|is|bs|yo|an|hy|fy|bpy|lmo|pnb|ml|sw|bn|io|af|gu|zh-yue|ne|nds|ku|ast|ur|scn|su|qu|diq|ba|tt|my|ga|cv|ia|nap|bat-smg|map-bms|wa|kn|als|am|bug|tg|gd|zh-min-nan|yi|vec|hif|sco|roa-tara|os|arz|nah|uz|sah|mn|sa|mzn|pam|hsb|mi|li|ky|si|co|gan|glk|ckb|bo|fo|bar|bcl|ilo|mrj|fiu-vro|nds-nl|tk|vls|se|gv|ps|rue|dv|nrm|pag|koi|pa|rm|km|kv|udm|csb|mhr|fur|mt|wuu|lij|ug|lad|pi|zea|sc|bh|zh-classical|nov|ksh|or|ang|kw|so|nv|xmf|stq|hak|ay|frp|frr|ext|szl|pcd|ie|gag|haw|xal|ln|rw|pdc|pfl|krc|crh|eml|ace|gn|to|ce|kl|arc|myv|dsb|vep|pap|bjn|as|tpi|lbe|wo|mdf|jbo|kab|av|sn|cbk-zam|ty|srn|kbd|lo|ab|lez|mwl|ltg|ig|na|kg|tet|za|kaa|nso|zu|rmy|cu|tn|chr|got|sm|bi|mo|bm|iu|chy|ik|pih|ss|sd|pnt|cdo|ee|ha|ti|bxr|om|ks|ts|ki|ve|sg|rn|dz|cr|lg|ak|tum|fj|st|tw|ch|ny|ff|xh|ng|ii|cho|mh|aa|kj|ho|mus|kr|hz";
var interwikiMap = {};
wikipedias.split('|').forEach( function (prefix) {
interwikiMap[prefix] = 'http://' + prefix + '.wikipedia.org/w';
});
var MWParserEnvironment = function(opts) {
var options = {
tagHooks: {},
parserFunctions: {},
pageCache: {}, // @fixme use something with managed space
debug: false,
trace: false,
wgScriptPath: "/wiki/",
wgScript: "/wiki/index.php",
wgUploadPath: "/wiki/images",
wgScriptExtension: ".php",
fetchTemplates: false,
maxDepth: 40,
pageName: 'Main page',
interwikiMap: interwikiMap,
interwikiRegexp: Object.keys(interwikiMap).join('|')
};
// XXX: this should be namespaced
$.extend(options, opts);
$.extend(this, options);
};
MWParserEnvironment.prototype.addInterwiki = function (prefix, wgScript) {
this.interwikiMap[prefix] = wgScript;
this.interwikiRegexp = Object.keys(this.interwikiMap).join('|');
};
// Outstanding page requests (for templates etc)
// Class-static
MWParserEnvironment.prototype.requestQueue = {};
MWParserEnvironment.prototype.lookupKV = function ( kvs, key ) {
if ( ! kvs ) {
return null;
}
var kv;
for ( var i = 0, l = kvs.length; i < l; i++ ) {
kv = kvs[i];
Big token transform framework overhaul part 2 * Tokens are now immutable. The progress of transformations is tracked on chunks instead of tokens. Tokenizer output is cached and can be directly returned without a need for cloning. Transforms are required to clone or newly create tokens they are modifying. * Expansions per chunk are now shared between equivalent frames via a cache stored on the chunk itself. Equivalence of frames is not yet ideal though, as right now a hash tree of *unexpanded* arguments is used. This should be switched to a hash of the fully expanded local parameters instead. * There is now a vastly improved maybeSyncReturn wrapper for async transforms that either forwards processing to the iterative transformTokens if the current transform is still ongoing, or manages a recursive transformation if needed. * Parameters for parser functions are now wrapped in abstract Params and ParserValue objects, which support some handy on-demand *value* expansions. Keys are always expanded. Parser functions are converted to use these interfaces, and now properly expand their values in the correct frame. Making this expansion lazier is certainly possible, but would complicate transformTokens and other token-handling machinery. Need to investigate if it would really be worth it. Dead branch elimination is certainly a bigger win overall. * Complex recursive asynchronous expansions should now be closer to correct for both the iterative (transformTokens) and recursive (maybeSyncReturn after transformTokens has returned) code paths. * Performance degraded slightly. There are no micro-optimizations done yet and the shared expansion cache still has a low hit rate. The progress tracking on chunks is not yet perfect, so there are likely a lot of unneeded re-expansions that can be easily eliminated. There is also more debug tracing right now. Obama currently expands in 54 seconds on my laptop. Change-Id: I4a603f3d3c70ca657ebda9fbb8570269f943d6b6
2012-05-10 08:04:24 +00:00
if ( kv.k.trim() === key ) {
// found, return it.
return kv;
}
}
// nothing found!
return null;
};
MWParserEnvironment.prototype.lookup = function ( kvs, key ) {
if ( ! kvs ) {
return null;
}
var kv;
for ( var i = 0, l = kvs.length; i < l; i++ ) {
kv = kvs[i];
if ( kv.k.trim() === key ) {
// found, return it.
return kv.v;
}
}
// nothing found!
return null;
};
MWParserEnvironment.prototype.lookupValue = function ( kvs, key ) {
if ( ! kvs ) {
return null;
}
var kv;
for ( var i = 0, l = kvs.length; i < l; i++ ) {
kv = kvs[i];
if ( kv.v === key ) {
// found, return it.
return kv;
}
}
// nothing found!
return null;
};
/**
* Trim space and newlines from leading and trailing text tokens.
*/
MWParserEnvironment.prototype.tokenTrim = function ( tokens ) {
var l = tokens.length,
i, token;
// strip leading space
for ( i = 0; i < l; i++ ) {
token = tokens[i];
if ( token.constructor === String ) {
token = token.replace( /^\s+/, '' );
tokens[i] = token;
if ( token !== '' ) {
break;
}
} else {
break;
}
}
// strip trailing space
for ( i = l - 1; i >= 0; i-- ) {
token = tokens[i];
if ( token.constructor === String ) {
token = token.replace( /\s+$/, '' );
tokens[i] = token;
if ( token !== '' ) {
break;
}
} else {
break;
}
}
return tokens;
};
/**
* Convert an array of key-value pairs into a hash of keys to values. For
* duplicate keys, the last entry wins.
*/
MWParserEnvironment.prototype.KVtoHash = function ( kvs ) {
if ( ! kvs ) {
console.warn( "Invalid kvs!: " + JSON.stringify( kvs, null, 2 ) );
return {};
}
var res = {};
for ( var i = 0, l = kvs.length; i < l; i++ ) {
var kv = kvs[i],
key = this.tokensToString( kv.k ).trim();
//if( res[key] === undefined ) {
res[key.toLowerCase()] = this.tokenTrim( kv.v );
//}
}
//console.warn( 'KVtoHash: ' + JSON.stringify( res ));
return res;
};
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
// Strip 'end' tokens and trailing newlines
MWParserEnvironment.prototype.stripEOFTkfromTokens = function ( tokens ) {
Big token transform framework overhaul part 2 * Tokens are now immutable. The progress of transformations is tracked on chunks instead of tokens. Tokenizer output is cached and can be directly returned without a need for cloning. Transforms are required to clone or newly create tokens they are modifying. * Expansions per chunk are now shared between equivalent frames via a cache stored on the chunk itself. Equivalence of frames is not yet ideal though, as right now a hash tree of *unexpanded* arguments is used. This should be switched to a hash of the fully expanded local parameters instead. * There is now a vastly improved maybeSyncReturn wrapper for async transforms that either forwards processing to the iterative transformTokens if the current transform is still ongoing, or manages a recursive transformation if needed. * Parameters for parser functions are now wrapped in abstract Params and ParserValue objects, which support some handy on-demand *value* expansions. Keys are always expanded. Parser functions are converted to use these interfaces, and now properly expand their values in the correct frame. Making this expansion lazier is certainly possible, but would complicate transformTokens and other token-handling machinery. Need to investigate if it would really be worth it. Dead branch elimination is certainly a bigger win overall. * Complex recursive asynchronous expansions should now be closer to correct for both the iterative (transformTokens) and recursive (maybeSyncReturn after transformTokens has returned) code paths. * Performance degraded slightly. There are no micro-optimizations done yet and the shared expansion cache still has a low hit rate. The progress tracking on chunks is not yet perfect, so there are likely a lot of unneeded re-expansions that can be easily eliminated. There is also more debug tracing right now. Obama currently expands in 54 seconds on my laptop. Change-Id: I4a603f3d3c70ca657ebda9fbb8570269f943d6b6
2012-05-10 08:04:24 +00:00
this.dp( 'stripping end or whitespace tokens' );
if ( tokens.constructor !== Array ) {
tokens = [ tokens ];
}
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
if ( ! tokens.length ) {
return tokens;
}
// Strip 'end' tokens and trailing newlines
var l = tokens[tokens.length - 1];
Big token transform framework overhaul part 2 * Tokens are now immutable. The progress of transformations is tracked on chunks instead of tokens. Tokenizer output is cached and can be directly returned without a need for cloning. Transforms are required to clone or newly create tokens they are modifying. * Expansions per chunk are now shared between equivalent frames via a cache stored on the chunk itself. Equivalence of frames is not yet ideal though, as right now a hash tree of *unexpanded* arguments is used. This should be switched to a hash of the fully expanded local parameters instead. * There is now a vastly improved maybeSyncReturn wrapper for async transforms that either forwards processing to the iterative transformTokens if the current transform is still ongoing, or manages a recursive transformation if needed. * Parameters for parser functions are now wrapped in abstract Params and ParserValue objects, which support some handy on-demand *value* expansions. Keys are always expanded. Parser functions are converted to use these interfaces, and now properly expand their values in the correct frame. Making this expansion lazier is certainly possible, but would complicate transformTokens and other token-handling machinery. Need to investigate if it would really be worth it. Dead branch elimination is certainly a bigger win overall. * Complex recursive asynchronous expansions should now be closer to correct for both the iterative (transformTokens) and recursive (maybeSyncReturn after transformTokens has returned) code paths. * Performance degraded slightly. There are no micro-optimizations done yet and the shared expansion cache still has a low hit rate. The progress tracking on chunks is not yet perfect, so there are likely a lot of unneeded re-expansions that can be easily eliminated. There is also more debug tracing right now. Obama currently expands in 54 seconds on my laptop. Change-Id: I4a603f3d3c70ca657ebda9fbb8570269f943d6b6
2012-05-10 08:04:24 +00:00
if ( l.constructor === EOFTk || l.constructor === NlTk ||
( l.constructor === String && l.match( /^\s+$/ ) ) ) {
var origTokens = tokens;
tokens = origTokens.slice();
tokens.rank = origTokens.rank;
while ( tokens.length &&
(( l.constructor === EOFTk || l.constructor === NlTk ) ||
( l.constructor === String && l.match( /^\s+$/ ) ) ) )
{
this.dp( 'stripping end or whitespace tokens' );
tokens.pop();
l = tokens[tokens.length - 1];
}
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
}
return tokens;
};
// Does this need separate UI/content inputs?
MWParserEnvironment.prototype.formatNum = function( num ) {
return num + '';
};
MWParserEnvironment.prototype.getVariable = function( varname, options ) {
//XXX what was the original author's intention?
//something like this?:
// return this.options[varname];
return this[varname];
};
MWParserEnvironment.prototype.setVariable = function( varname, value, options ) {
this[varname] = value;
};
/**
* @return MWParserFunction
*/
MWParserEnvironment.prototype.getParserFunction = function( name ) {
if (name in this.parserFunctions) {
return new this.parserFunctions[name]( this );
} else {
return null;
}
};
/**
* @return MWParserTagHook
*/
MWParserEnvironment.prototype.getTagHook = function( name ) {
if (name in this.tagHooks) {
return new this.tagHooks[name](this);
} else {
return null;
}
};
MWParserEnvironment.prototype.makeTitleFromPrefixedText = function ( text ) {
text = this.normalizeTitle( text );
var nsText = text.split( ':', 1 )[0];
if ( nsText && nsText !== text ) {
var _ns = new Namespace(0);
var ns = _ns._defaultNamespaceIDs[ nsText.toLowerCase() ];
//console.warn( JSON.stringify( [ nsText, ns ] ) );
if ( ns !== undefined ) {
return new Title( text.substr( nsText.length + 1 ), ns, nsText, this );
} else {
return new Title( text, 0, '', this );
}
} else {
return new Title( text, 0, '', this );
}
};
// XXX: move to Title!
MWParserEnvironment.prototype.normalizeTitle = function( name ) {
if (typeof name !== 'string') {
throw new Error('nooooooooo not a string');
}
var forceNS, self = this;
if ( name.substr( 0, 1 ) === ':' ) {
forceNS = ':';
name = name.substr(1);
} else {
forceNS = '';
}
name = name.trim().replace(/[\s_]+/g, '_');
// Implement int: as alias for MediaWiki:
if ( name.substr( 0, 4 ) === 'int:' ) {
name = 'MediaWiki:' + name.substr( 4 );
}
// FIXME: Generalize namespace case normalization
if ( name.substr( 0, 10 ).toLowerCase() === 'mediawiki:' ) {
name = 'MediaWiki:' + name.substr( 10 );
}
function upperFirst( s ) { return s.substr(0, 1).toUpperCase() + s.substr(1); }
function splitNS ( ) {
var nsMatch = name.match( /([a-zA-Z-]+):/ ),
ns = nsMatch && nsMatch[1] || '';
if( ns !== '' && ns !== name ) {
if ( self.interwikiMap[ns.toLowerCase()] ) {
forceNS += ns + ':';
name = name.substr( nsMatch[0].length );
splitNS();
} else {
name = upperFirst( ns ) + ':' + upperFirst( name.substr( ns.length + 1 ) );
}
} else {
name = upperFirst( name );
}
}
splitNS();
//name = name.split(':').map( upperFirst ).join(':');
//if (name === '') {
// throw new Error('Invalid/empty title');
//}
return forceNS + name;
};
/**
* @fixme do this for real eh
*/
MWParserEnvironment.prototype.resolveTitle = function( name, namespace ) {
// Resolve subpages
if ( name.length && name[0] === '/' ) {
name = this.normalizeTitle( this.pageName ) + name;
}
// FIXME: match against proper list of namespaces
if ( name.indexOf(':') == -1 && namespace ) {
// hack hack hack
name = namespace + ':' + this.normalizeTitle( name );
}
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
// Strip leading ':'
if (name[0] === ':') {
name = name.substr( 1 );
}
return name;
};
MWParserEnvironment.prototype.tokensToString = function ( tokens, strict ) {
var out = [];
//console.warn( 'MWParserEnvironment.tokensToString, tokens: ' + JSON.stringify( tokens ) );
// XXX: quick hack, track down non-array sources later!
if ( ! $.isArray( tokens ) ) {
tokens = [ tokens ];
}
for ( var i = 0, l = tokens.length; i < l; i++ ) {
var token = tokens[i];
if ( token === undefined ) {
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
if ( this.debug ) { console.trace(); }
this.tp( 'MWParserEnvironment.tokensToString, invalid token: ' +
token, ' tokens:', tokens);
continue;
}
if ( token.constructor === String ) {
out.push( token );
} else if ( token.constructor === CommentTk || token.constructor === NlTk ) {
// strip comments and newlines
} else {
if ( strict ) {
return [out.join(''), tokens.slice( i )];
}
var tstring = JSON.stringify( token );
this.dp ( 'MWParserEnvironment.tokensToString, non-text token: ',
tstring, tokens);
Biggish token transform system refactoring * All parser pipelines including tokenizer and DOM stuff are now constructed from a 'recipe' data structure in a ParserPipelineFactory. * All sub-pipelines of these can now be cached * Event registrations to a pipeline are directly forwarded to the last pipeline member to save relatively expensive event forwarding. * Some APIs for on-demand expansion / format conversion of parameters from parser functions are added: param.to('tokens/expanded', cb) param.to('text/wiki', cb) (this does not work yet) All parameters are additionally wrapped into a Param object that provides method for positional parameter naming (.named() or conversion to a dict (.dict()). * The async token transform manager is now separated from a frame object, with the frame holding arguments, an on-demand expansion method and loop checks. * Only keys of template parameters are now expanded. Parser functions or template arguments trigger an expansion on-demand. This (unsurprisingly) makes a big performance difference with typical switch-heavy template systems. * Return values from async transforms are no longer used in favor of plain callbacks. This saves the complication of having to maintain two code paths. A trick in transformTokens still avoids the construction of unneeded TokenAccumulators. * The results of template expansions are no longer buffered. * 301 parser tests are passing Known issues: * Cosmetic cleanup remains to do * Some parser functions do not support async expansions yet, and need to be modified. Change-Id: I1a7690baffbe8141cadf67270904a1b2e1df879a
2012-04-25 14:35:59 +00:00
if ( this.debug ) { console.trace(); }
//out.push( tstring );
}
}
Big token transform framework overhaul part 2 * Tokens are now immutable. The progress of transformations is tracked on chunks instead of tokens. Tokenizer output is cached and can be directly returned without a need for cloning. Transforms are required to clone or newly create tokens they are modifying. * Expansions per chunk are now shared between equivalent frames via a cache stored on the chunk itself. Equivalence of frames is not yet ideal though, as right now a hash tree of *unexpanded* arguments is used. This should be switched to a hash of the fully expanded local parameters instead. * There is now a vastly improved maybeSyncReturn wrapper for async transforms that either forwards processing to the iterative transformTokens if the current transform is still ongoing, or manages a recursive transformation if needed. * Parameters for parser functions are now wrapped in abstract Params and ParserValue objects, which support some handy on-demand *value* expansions. Keys are always expanded. Parser functions are converted to use these interfaces, and now properly expand their values in the correct frame. Making this expansion lazier is certainly possible, but would complicate transformTokens and other token-handling machinery. Need to investigate if it would really be worth it. Dead branch elimination is certainly a bigger win overall. * Complex recursive asynchronous expansions should now be closer to correct for both the iterative (transformTokens) and recursive (maybeSyncReturn after transformTokens has returned) code paths. * Performance degraded slightly. There are no micro-optimizations done yet and the shared expansion cache still has a low hit rate. The progress tracking on chunks is not yet perfect, so there are likely a lot of unneeded re-expansions that can be easily eliminated. There is also more debug tracing right now. Obama currently expands in 54 seconds on my laptop. Change-Id: I4a603f3d3c70ca657ebda9fbb8570269f943d6b6
2012-05-10 08:04:24 +00:00
this.dp( 'MWParserEnvironment.tokensToString result: ', out );
return out.join('');
};
/**
* Perform a shallow clone of a chunk of tokens
*/
MWParserEnvironment.prototype.cloneTokens = function ( chunk ) {
var out = [],
token, tmpToken;
for ( var i = 0, l = chunk.length; i < l; i++ ) {
token = chunk[i];
if ( token.constructor === String ) {
out.push( token );
} else {
tmpToken = $.extend( {}, token );
tmpToken.rank = 0;
out.push(tmpToken);
}
}
return out;
};
MWParserEnvironment.prototype.decodeURI = function ( s ) {
return s.replace( /%[0-9a-f][0-9a-f]/g, function( m ) {
try {
return decodeURI( m );
} catch ( e ) {
return m;
}
} );
};
MWParserEnvironment.prototype.sanitizeURI = function ( s ) {
var host = s.match(/^[a-zA-Z]+:\/\/[^\/]+(?:\/|$)/),
path = s,
anchor = null;
//console.warn( 'host: ' + host );
if ( host ) {
path = s.substr( host[0].length );
host = host[0];
} else {
host = '';
}
var bits = path.split('#');
if ( bits.length > 1 ) {
anchor = bits[bits.length - 1];
path = path.substr(0, path.length - anchor.length - 1);
}
host = host.replace( /%(?![0-9a-fA-F][0-9a-fA-F])|[#|]/g, function ( m ) {
return encodeURIComponent( m );
} );
path = path.replace( /%(?![0-9a-fA-F][0-9a-fA-F])|[ \[\]#|]/g, function ( m ) {
return encodeURIComponent( m );
} );
s = host + path;
if ( anchor !== null ) {
s += '#' + anchor;
}
return s;
};
/**
* Simple debug helper
*/
MWParserEnvironment.prototype.dp = function ( ) {
if ( this.debug ) {
if ( arguments.length > 1 ) {
try {
console.warn( JSON.stringify( arguments, null, 2 ) );
} catch ( e ) {
Big token transform framework overhaul part 2 * Tokens are now immutable. The progress of transformations is tracked on chunks instead of tokens. Tokenizer output is cached and can be directly returned without a need for cloning. Transforms are required to clone or newly create tokens they are modifying. * Expansions per chunk are now shared between equivalent frames via a cache stored on the chunk itself. Equivalence of frames is not yet ideal though, as right now a hash tree of *unexpanded* arguments is used. This should be switched to a hash of the fully expanded local parameters instead. * There is now a vastly improved maybeSyncReturn wrapper for async transforms that either forwards processing to the iterative transformTokens if the current transform is still ongoing, or manages a recursive transformation if needed. * Parameters for parser functions are now wrapped in abstract Params and ParserValue objects, which support some handy on-demand *value* expansions. Keys are always expanded. Parser functions are converted to use these interfaces, and now properly expand their values in the correct frame. Making this expansion lazier is certainly possible, but would complicate transformTokens and other token-handling machinery. Need to investigate if it would really be worth it. Dead branch elimination is certainly a bigger win overall. * Complex recursive asynchronous expansions should now be closer to correct for both the iterative (transformTokens) and recursive (maybeSyncReturn after transformTokens has returned) code paths. * Performance degraded slightly. There are no micro-optimizations done yet and the shared expansion cache still has a low hit rate. The progress tracking on chunks is not yet perfect, so there are likely a lot of unneeded re-expansions that can be easily eliminated. There is also more debug tracing right now. Obama currently expands in 54 seconds on my laptop. Change-Id: I4a603f3d3c70ca657ebda9fbb8570269f943d6b6
2012-05-10 08:04:24 +00:00
console.trace();
console.warn( e );
}
} else {
console.warn( arguments[0] );
}
}
};
/**
* Even simpler debug helper that always prints..
*/
MWParserEnvironment.prototype.ap = function ( ) {
if ( arguments.length > 1 ) {
try {
console.warn( JSON.stringify( arguments, null, 2 ) );
} catch ( e ) {
console.warn( e );
}
} else {
console.warn( arguments[0] );
}
};
/**
* Simple debug helper, trace-only
*/
MWParserEnvironment.prototype.tp = function ( ) {
if ( this.debug || this.trace ) {
if ( arguments.length > 1 ) {
console.warn( JSON.stringify( arguments, null, 2 ) );
} else {
console.warn( arguments[0] );
}
}
};
if (typeof module == "object") {
module.exports.MWParserEnvironment = MWParserEnvironment;
}