n3 1.26.0 → 2.0.0-beta.2

package/lib/IRIs.js

@@ -20,7 +20,9 @@ var _default = exports.default = {
     nil: `${RDF}nil`,
     first: `${RDF}first`,
     rest: `${RDF}rest`,
-    langString: `${RDF}langString`
+    langString: `${RDF}langString`,
+    dirLangString: `${RDF}dirLangString`,
+    reifies: `${RDF}reifies`
   },
   owl: {
     sameAs: 'http://www.w3.org/2002/07/owl#sameAs'
@@ -30,7 +32,6 @@ var _default = exports.default = {
     forAll: `${SWAP}reify#forAll`
   },
   log: {
-    implies: `${SWAP}log#implies`,
-    isImpliedBy: `${SWAP}log#isImpliedBy`
+    implies: `${SWAP}log#implies`
   }
 };

package/lib/N3DataFactory.js

@@ -99,8 +99,17 @@ class Literal extends Term {
     // Find the last quotation mark (e.g., '"abc"@en-us')
     const id = this.id;
     let atPos = id.lastIndexOf('"') + 1;
+    const dirPos = id.lastIndexOf('--');
     // If "@" it follows, return the remaining substring; empty otherwise
-    return atPos < id.length && id[atPos++] === '@' ? id.substr(atPos).toLowerCase() : '';
+    return atPos < id.length && id[atPos++] === '@' ? (dirPos > atPos ? id.substr(0, dirPos) : id).substr(atPos).toLowerCase() : '';
+  }
+
+  // ### The direction of this literal
+  get direction() {
+    // Find the last double dash (e.g., '"abc"@en-us--ltr')
+    const id = this.id;
+    const atPos = id.lastIndexOf('--') + 2;
+    return atPos > 1 && atPos < id.length ? id.substr(atPos).toLowerCase() : '';
   }
 
   // ### The datatype IRI of this literal
@@ -116,8 +125,8 @@ class Literal extends Term {
     const char = dtPos < id.length ? id[dtPos] : '';
     // If "^" it follows, return the remaining substring
     return char === '^' ? id.substr(dtPos + 2) :
-    // If "@" follows, return rdf:langString; xsd:string otherwise
-    char !== '@' ? xsd.string : rdf.langString;
+    // If "@" follows, return rdf:langString or rdf:dirLangString; xsd:string otherwise
+    char !== '@' ? xsd.string : id.indexOf('--', dtPos) > 0 ? rdf.dirLangString : rdf.langString;
   }
 
   // ### Returns whether this object represents the same term as the other
@@ -126,13 +135,14 @@ class Literal extends Term {
     // equality can be computed through ids
     if (other instanceof Literal) return this.id === other.id;
     // Otherwise, compare term type, value, language, and datatype
-    return !!other && !!other.datatype && this.termType === other.termType && this.value === other.value && this.language === other.language && this.datatype.value === other.datatype.value;
+    return !!other && !!other.datatype && this.termType === other.termType && this.value === other.value && this.language === other.language && (this.direction === other.direction || this.direction === '' && !other.direction) && this.datatype.value === other.datatype.value;
   }
   toJSON() {
     return {
       termType: this.termType,
       value: this.value,
       language: this.language,
+      direction: this.direction,
       datatype: {
         termType: 'NamedNode',
         value: this.datatypeString
@@ -225,7 +235,20 @@ function termFromId(id, factory, nested) {
       if (id[id.length - 1] === '"') return factory.literal(id.substr(1, id.length - 2));
       // Literal with datatype or language
       const endPos = id.lastIndexOf('"', id.length - 1);
-      return factory.literal(id.substr(1, endPos - 1), id[endPos + 1] === '@' ? id.substr(endPos + 2) : factory.namedNode(id.substr(endPos + 3)));
+      let languageOrDatatype;
+      if (id[endPos + 1] === '@') {
+        languageOrDatatype = id.substr(endPos + 2);
+        const dashDashIndex = languageOrDatatype.lastIndexOf('--');
+        if (dashDashIndex > 0 && dashDashIndex < languageOrDatatype.length) {
+          languageOrDatatype = {
+            language: languageOrDatatype.substr(0, dashDashIndex),
+            direction: languageOrDatatype.substr(dashDashIndex + 2)
+          };
+        }
+      } else {
+        languageOrDatatype = factory.namedNode(id.substr(endPos + 3));
+      }
+      return factory.literal(id.substr(1, endPos - 1), languageOrDatatype);
     case '[':
       id = JSON.parse(id);
       break;
@@ -258,7 +281,7 @@ function termToId(term, nested) {
     case 'DefaultGraph':
       return '';
     case 'Literal':
-      return `"${term.value}"${term.language ? `@${term.language}` : term.datatype && term.datatype.value !== xsd.string ? `^^${term.datatype.value}` : ''}`;
+      return `"${term.value}"${term.language ? `@${term.language}${term.direction ? `--${term.direction}` : ''}` : term.datatype && term.datatype.value !== xsd.string ? `^^${term.datatype.value}` : ''}`;
     case 'Quad':
       const res = [termToId(term.subject, true), termToId(term.predicate, true), termToId(term.object, true)];
       if (term.graph && term.graph.termType !== 'DefaultGraph') {
@@ -339,6 +362,11 @@ function literal(value, languageOrDataType) {
   // Create a language-tagged string
   if (typeof languageOrDataType === 'string') return new Literal(`"${value}"@${languageOrDataType.toLowerCase()}`);
 
+  // Create a language-tagged string with base direction
+  if (languageOrDataType !== undefined && !('termType' in languageOrDataType)) {
+    return new Literal(`"${value}"@${languageOrDataType.language.toLowerCase()}--${languageOrDataType.direction.toLowerCase()}`);
+  }
+
   // Automatically determine datatype for booleans and numbers
   let datatype = languageOrDataType ? languageOrDataType.value : '';
   if (datatype === '') {

package/lib/N3Lexer.js

@@ -5,6 +5,7 @@ Object.defineProperty(exports, "__esModule", {
 });
 exports.default = void 0;
 var _buffer = require("buffer");
+var _queueMicrotask = _interopRequireDefault(require("queue-microtask"));
 var _IRIs = _interopRequireDefault(require("./IRIs"));
 function _interopRequireDefault(e) { return e && e.__esModule ? e : { default: e }; }
 // **N3Lexer** tokenizes N3 documents.
@@ -50,6 +51,7 @@ const lineModeRegExps = {
   _unescapedIri: true,
   _simpleQuotedString: true,
   _langcode: true,
+  _dircode: true,
   _blank: true,
   _newline: true,
   _comment: true,
@@ -67,15 +69,16 @@ class N3Lexer {
     this._unescapedIri = /^<([^\x00-\x20<>\\"\{\}\|\^\`]*)>[ \t]*/; // IRI without escape sequences; no unescaping
     this._simpleQuotedString = /^"([^"\\\r\n]*)"(?=[^"])/; // string without escape sequences
     this._simpleApostropheString = /^'([^'\\\r\n]*)'(?=[^'])/;
-    this._langcode = /^@([a-z]+(?:-[a-z0-9]+)*)(?=[^a-z0-9\-])/i;
+    this._langcode = /^@([a-z]+(?:-[a-z0-9]+)*)(?=[^a-z0-9])/i;
+    this._dircode = /^--(ltr)|(rtl)/;
     this._prefix = /^((?:[A-Za-z\xc0-\xd6\xd8-\xf6\xf8-\u02ff\u0370-\u037d\u037f-\u1fff\u200c\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])(?:\.?[\-0-9A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])*)?:(?=[#\s<])/;
     this._prefixed = /^((?:[A-Za-z\xc0-\xd6\xd8-\xf6\xf8-\u02ff\u0370-\u037d\u037f-\u1fff\u200c\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])(?:\.?[\-0-9A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])*)?:((?:(?:[0-:A-Z_a-z\xc0-\xd6\xd8-\xf6\xf8-\u02ff\u0370-\u037d\u037f-\u1fff\u200c\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff]|%[0-9a-fA-F]{2}|\\[!#-\/;=?\-@_~])(?:(?:[\.\-0-:A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff]|%[0-9a-fA-F]{2}|\\[!#-\/;=?\-@_~])*(?:[\-0-:A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff]|%[0-9a-fA-F]{2}|\\[!#-\/;=?\-@_~]))?)?)(?:[ \t]+|(?=\.?[,;!\^\s#()\[\]\{\}"'<>]))/;
     this._variable = /^\?(?:(?:[A-Z_a-z\xc0-\xd6\xd8-\xf6\xf8-\u02ff\u0370-\u037d\u037f-\u1fff\u200c\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])(?:[\-0-:A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])*)(?=[.,;!\^\s#()\[\]\{\}"'<>])/;
     this._blank = /^_:((?:[0-9A-Z_a-z\xc0-\xd6\xd8-\xf6\xf8-\u02ff\u0370-\u037d\u037f-\u1fff\u200c\u200d\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])(?:\.?[\-0-9A-Z_a-z\xb7\xc0-\xd6\xd8-\xf6\xf8-\u037d\u037f-\u1fff\u200c\u200d\u203f\u2040\u2070-\u218f\u2c00-\u2fef\u3001-\ud7ff\uf900-\ufdcf\ufdf0-\ufffd]|[\ud800-\udb7f][\udc00-\udfff])*)(?:[ \t]+|(?=\.?[,;:\s#()\[\]\{\}"'<>]))/;
     this._number = /^[\-+]?(?:(\d+\.\d*|\.?\d+)[eE][\-+]?|\d*(\.)?)\d+(?=\.?[,;:\s#()\[\]\{\}"'<>])/;
     this._boolean = /^(?:true|false)(?=[.,;\s#()\[\]\{\}"'<>])/;
-    this._keyword = /^@[a-z]+(?=[\s#<:])/i;
-    this._sparqlKeyword = /^(?:PREFIX|BASE|GRAPH)(?=[\s#<])/i;
+    this._atKeyword = /^@[a-z]+(?=[\s#<:])/i;
+    this._keyword = /^(?:PREFIX|BASE|VERSION|GRAPH)(?=[\s#<])/i;
     this._shortPredicates = /^a(?=[\s#()\[\]\{\}"'<>])/;
     this._newline = /^[ \t]*(?:#[^\n\r]*)?(?:\r\n|\n|\r)[ \t]*/;
     this._comment = /#([^\n\r]*)/;
@@ -83,9 +86,6 @@ class N3Lexer {
     this._endOfFile = /^(?:#[^\n\r]*)?$/;
     options = options || {};
 
-    // Whether the log:isImpliedBy predicate is supported
-    this._isImpliedBy = options.isImpliedBy;
-
     // In line mode (N-Triples or N-Quads), only simple features may be parsed
     if (this._lineMode = !!options.lineMode) {
       this._n3Mode = false;
@@ -178,15 +178,15 @@ class N3Lexer {
             if (value === null || illegalIriChars.test(value)) return reportSyntaxError(this);
             type = 'IRI';
           }
-          // Try to find a nested triple
-          else if (input.length > 1 && input[1] === '<') type = '<<', matchLength = 2;
+          // Try to find a triple term
+          else if (input.length > 2 && input[1] === '<' && input[2] === '(') type = '<<(', matchLength = 3;
+          // Try to find a reified triple
+          else if (!this._lineMode && input.length > (inputFinished ? 1 : 2) && input[1] === '<') type = '<<', matchLength = 2;
           // Try to find a backwards implication arrow
-          else if (this._n3Mode && input.length > 1 && input[1] === '=') {
-            matchLength = 2;
-            if (this._isImpliedBy) type = 'abbreviation', value = '<';else type = 'inverse', value = '>';
-          }
+          else if (this._n3Mode && input.length > 1 && input[1] === '=') type = 'inverse', matchLength = 2, value = '>';
           break;
         case '>':
+          // Try to find a reified triple
           if (input.length > 1 && input[1] === '>') type = '>>', matchLength = 2;
           break;
         case '_':
@@ -235,9 +235,9 @@ class N3Lexer {
           break;
         case '@':
           // Try to find a language code
-          if (this._previousMarker === 'literal' && (match = this._langcode.exec(input))) type = 'langcode', value = match[1];
+          if (this._previousMarker === 'literal' && (match = this._langcode.exec(input)) && match[1] !== 'version') type = 'langcode', value = match[1];
           // Try to find a keyword
-          else if (match = this._keyword.exec(input)) type = match[0];
+          else if (match = this._atKeyword.exec(input)) type = match[0];
           break;
         case '.':
           // Try to find a dot as punctuation
@@ -260,6 +260,12 @@ class N3Lexer {
         case '9':
         case '+':
         case '-':
+          if (input[1] === '-') {
+            // Try to find a direction code
+            if (this._previousMarker === 'langcode' && (match = this._dircode.exec(input))) type = 'dircode', matchLength = 2, value = match[1] || match[2], matchLength = value.length + 2;
+            break;
+          }
+
           // Try to find a number. Since it can contain (but not end with) a dot,
           // we always need a non-dot character before deciding it is a number.
           // Therefore, try inserting a space if we're at the end of the input.
@@ -274,8 +280,10 @@ class N3Lexer {
         case 'P':
         case 'G':
         case 'g':
+        case 'V':
+        case 'v':
           // Try to find a SPARQL-style keyword
-          if (match = this._sparqlKeyword.exec(input)) type = match[0].toUpperCase();else inconclusive = true;
+          if (match = this._keyword.exec(input)) type = match[0].toUpperCase();else inconclusive = true;
           break;
         case 'f':
         case 't':
@@ -295,13 +303,23 @@ class N3Lexer {
           break;
         case '!':
           if (!this._n3Mode) break;
+        case ')':
+          if (!inputFinished && (input.length === 1 || input.length === 2 && input[1] === '>')) {
+            // Don't consume yet, as it *could* become a triple term end.
+            break;
+          }
+          // Try to find a triple term
+          if (input.length > 2 && input[1] === '>' && input[2] === '>') {
+            type = ')>>', matchLength = 3;
+            break;
+          }
         case ',':
         case ';':
         case '[':
         case ']':
         case '(':
-        case ')':
         case '}':
+        case '~':
           if (!this._lineMode) {
             matchLength = 1;
             type = firstChar;
@@ -474,7 +492,7 @@ class N3Lexer {
     if (typeof input === 'string') {
       this._input = this._readStartingBom(input);
       // If a callback was passed, asynchronously call it
-      if (typeof callback === 'function') queueMicrotask(() => this._tokenizeToEnd(callback, true));
+      if (typeof callback === 'function') (0, _queueMicrotask.default)(() => this._tokenizeToEnd(callback, true));
       // If no callback was passed, tokenize synchronously and return
       else {
         const tokens = [];