eyeling 1.24.30 → 1.24.31

package/lib/lexer.js

@@ -356,6 +356,24 @@ function stripQuotes(lex) {
 // This keeps all downstream parsing/reasoning N3-only.
 const LOG_NAME_OF_IRI = '<http://www.w3.org/2000/10/swap/log#nameOf>';
 const RDF_REIFIES_IRI = '<http://www.w3.org/1999/02/22-rdf-syntax-ns#reifies>';
+const RDF_TYPE_IRI = '<http://www.w3.org/1999/02/22-rdf-syntax-ns#type>';
+const XSD_INTEGER_IRI = '<http://www.w3.org/2001/XMLSchema#integer>';
+const EYMSG_NS = 'https://eyereasoner.github.io/eyeling/vocab/message#';
+const EYMSG = Object.freeze({
+  RDFMessageStream: `<${EYMSG_NS}RDFMessageStream>`,
+  MessageEnvelope: `<${EYMSG_NS}MessageEnvelope>`,
+  envelope: `<${EYMSG_NS}envelope>`,
+  firstEnvelope: `<${EYMSG_NS}firstEnvelope>`,
+  lastEnvelope: `<${EYMSG_NS}lastEnvelope>`,
+  orderedEnvelopes: `<${EYMSG_NS}orderedEnvelopes>`,
+  messageCount: `<${EYMSG_NS}messageCount>`,
+  offset: `<${EYMSG_NS}offset>`,
+  nextEnvelope: `<${EYMSG_NS}nextEnvelope>`,
+  payloadGraph: `<${EYMSG_NS}payloadGraph>`,
+  payloadKind: `<${EYMSG_NS}payloadKind>`,
+  empty: `<${EYMSG_NS}empty>`,
+  nonEmpty: `<${EYMSG_NS}nonEmpty>`,
+});
 
 function normalizeRdfCompatibility(inputText) {
   let text = String(inputText ?? '');
@@ -366,10 +384,11 @@ function normalizeRdfCompatibility(inputText) {
   // plausible top-level TriG named graph block.
   const hasTripleTerms = text.includes('<<');
   const hasVersionDirective = /^\s*(?:@version|VERSION)\s+(["'])1\.2\1\s*\.?\s*(?:#.*)?$/im.test(text);
+  const hasMessageVersionDirective = /^\s*(?:@version|VERSION)\s+(["'])(?:1\.1|1\.2|1\.2-basic)-messages\1\s*\.?\s*(?:#.*)?$/im.test(text);
   const hasNamedGraphCandidate = /(?:^|[.\r\n])\s*(?:GRAPH\s+)?(?:<[^>\r\n]*>|_:[A-Za-z][A-Za-z0-9_-]*|[A-Za-z][A-Za-z0-9_-]*:[^\s{};,.()[\]]*)\s*\{/m.test(text);
   const hasAnnotationSyntax = /(?:^|\s)~\s*(?:<|_:[A-Za-z]|[A-Za-z][A-Za-z0-9_-]*:|\{\|)|\{\|/.test(text);
 
-  if (!hasTripleTerms && !hasVersionDirective && !hasNamedGraphCandidate && !hasAnnotationSyntax) return text;
+  if (!hasTripleTerms && !hasVersionDirective && !hasMessageVersionDirective && !hasNamedGraphCandidate && !hasAnnotationSyntax) return text;
 
   function isWordChar(ch) {
     return ch != null && /[A-Za-z0-9_:-]/.test(ch);
@@ -723,7 +742,7 @@ function normalizeRdfCompatibility(inputText) {
   }
 
   function stripVersionDirectives(s) {
-    return s.replace(/^\s*(?:@version|VERSION)\s+(["'])1\.2\1\s*\.?\s*(?:#.*)?$/gim, '');
+    return s.replace(/^\s*(?:@version|VERSION)\s+(["'])(?:1\.1|1\.2|1\.2-basic)(?:-messages)?\1\s*\.?\s*(?:#.*)?$/gim, '');
   }
 
   function skipWsAndComments(s, at) {
@@ -877,10 +896,270 @@ function normalizeRdfCompatibility(inputText) {
     return out;
   }
 
+
+  function isOnlyWhitespaceAndComments(s) {
+    return skipWsAndComments(s, 0) >= s.length;
+  }
+
+
+  function skipOldStyleDirective(s, at) {
+    let i = at;
+    while (i < s.length) {
+      const ch = s[i];
+      if (ch === '"' || ch === "'") {
+        i = readStringAt(s, i).end;
+        continue;
+      }
+      if (ch === '<') {
+        i = readIriAt(s, i).end;
+        continue;
+      }
+      if (ch === '#') {
+        while (i < s.length && s[i] !== '\n' && s[i] !== '\r') i += 1;
+        continue;
+      }
+      i += 1;
+      if (ch === '.') return i;
+    }
+    return i;
+  }
+
+  function stripDirectivesAndCommentsForEmptiness(s) {
+    let out = '';
+    let i = 0;
+    let statementStart = true;
+    while (i < s.length) {
+      const ch = s[i];
+      if (ch === '"' || ch === "'") {
+        const str = readStringAt(s, i);
+        out += str.text;
+        i = str.end;
+        statementStart = false;
+        continue;
+      }
+      if (ch === '<') {
+        const iri = readIriAt(s, i);
+        out += iri.text;
+        i = iri.end;
+        statementStart = false;
+        continue;
+      }
+      if (ch === '#') {
+        while (i < s.length && s[i] !== '\n' && s[i] !== '\r') i += 1;
+        statementStart = true;
+        continue;
+      }
+      if (statementStart) {
+        const start = skipWsAndComments(s, i);
+        out += s.slice(i, start);
+        i = start;
+        if (startsWordAt(s, 'PREFIX', i) || startsWordAt(s, 'BASE', i) || startsWordAt(s, 'VERSION', i)) {
+          while (i < s.length && s[i] !== '\n' && s[i] !== '\r') i += 1;
+          statementStart = true;
+          continue;
+        }
+        if (s[i] === '@') {
+          const lower = s.slice(i, i + 9).toLowerCase();
+          if (lower.startsWith('@prefix') || lower.startsWith('@base') || lower.startsWith('@version')) {
+            i = skipOldStyleDirective(s, i);
+            statementStart = true;
+            continue;
+          }
+        }
+      }
+      out += ch;
+      if (ch === '.' || ch === '}' || ch === '\n' || ch === '\r') statementStart = true;
+      else if (!/\s/.test(ch)) statementStart = false;
+      i += 1;
+    }
+    return out;
+  }
+
+  function simpleHashText(s) {
+    let h = 0x811c9dc5;
+    for (let i = 0; i < s.length; i += 1) {
+      h ^= s.charCodeAt(i);
+      h = Math.imul(h, 0x01000193) >>> 0;
+    }
+    return h.toString(16).padStart(8, '0');
+  }
+
+  function rewriteMessageBlankLabels(s, messageIndex) {
+    let out = '';
+    let i = 0;
+    const prefix = `_:eyeling_m${String(messageIndex).padStart(3, '0')}_`;
+    while (i < s.length) {
+      const ch = s[i];
+      if (ch === '"' || ch === "'") {
+        const str = readStringAt(s, i);
+        out += str.text;
+        i = str.end;
+        continue;
+      }
+      if (ch === '<' && !s.startsWith('<<', i)) {
+        const iri = readIriAt(s, i);
+        out += iri.text;
+        i = iri.end;
+        continue;
+      }
+      if (ch === '#') {
+        while (i < s.length) {
+          const c = s[i++];
+          out += c;
+          if (c === '\n' || c === '\r') break;
+        }
+        continue;
+      }
+      if (s.startsWith('_:', i)) {
+        let j = i + 2;
+        while (j < s.length && !/\s/.test(s[j]) && !'{}[](),;.'.includes(s[j])) j += 1;
+        const label = s.slice(i + 2, j);
+        if (label) {
+          out += prefix + label.replace(/[^A-Za-z0-9_]/g, '_');
+          i = j;
+          continue;
+        }
+      }
+      out += ch;
+      i += 1;
+    }
+    return out;
+  }
+
+  function findMessageDirectiveAt(s, at) {
+    if (startsWordAt(s, 'MESSAGE', at)) return { start: at, end: at + 'MESSAGE'.length };
+    if (s.slice(at, at + 8).toLowerCase() === '@message' && !isWordChar(s[at + 8])) {
+      let end = at + 8;
+      end = skipWsAndComments(s, end);
+      if (s[end] === '.') end += 1;
+      return { start: at, end };
+    }
+    return null;
+  }
+
+  function splitRdfMessageLog(s) {
+    const chunks = [];
+    let i = 0;
+    let start = 0;
+    let braceDepth = 0;
+    let bracketDepth = 0;
+    let parenDepth = 0;
+    let statementStart = true;
+    let sawDelimiter = false;
+
+    while (i < s.length) {
+      const ch = s[i];
+      if (ch === '"' || ch === "'") {
+        i = readStringAt(s, i).end;
+        statementStart = false;
+        continue;
+      }
+      if (ch === '<' && !s.startsWith('<<', i)) {
+        i = readIriAt(s, i).end;
+        statementStart = false;
+        continue;
+      }
+      if (ch === '#') {
+        while (i < s.length && s[i] !== '\n' && s[i] !== '\r') i += 1;
+        statementStart = true;
+        continue;
+      }
+      if (statementStart && braceDepth === 0 && bracketDepth === 0 && parenDepth === 0) {
+        const termStart = skipWsAndComments(s, i);
+        const msg = findMessageDirectiveAt(s, termStart);
+        if (msg) {
+          chunks.push(s.slice(start, termStart));
+          start = msg.end;
+          i = msg.end;
+          statementStart = true;
+          sawDelimiter = true;
+          continue;
+        }
+        if (termStart !== i) {
+          i = termStart;
+          continue;
+        }
+      }
+      if (ch === '{') braceDepth += 1;
+      else if (ch === '}' && braceDepth > 0) braceDepth -= 1;
+      else if (ch === '[') bracketDepth += 1;
+      else if (ch === ']' && bracketDepth > 0) bracketDepth -= 1;
+      else if (ch === '(') parenDepth += 1;
+      else if (ch === ')' && parenDepth > 0) parenDepth -= 1;
+
+      if (ch === '.' && braceDepth === 0 && bracketDepth === 0 && parenDepth === 0) statementStart = true;
+      else if (ch === '\n' || ch === '\r') statementStart = true;
+      else if (!/\s/.test(ch)) statementStart = false;
+      i += 1;
+    }
+
+    const tail = s.slice(start);
+    if (!sawDelimiter || !isOnlyWhitespaceAndComments(tail)) chunks.push(tail);
+    return chunks;
+  }
+
+  function normalizeMessageChunk(chunk, messageIndex) {
+    let body = String(chunk || '');
+    if (hasTripleTerms || body.includes('<<')) body = convertTripleTerms(body);
+    if (hasAnnotationSyntax || /(?:^|\s)~\s*(?:<|_:[A-Za-z]|[A-Za-z][A-Za-z0-9_-]*:|\{\|)|\{\|/.test(body)) {
+      body = convertAnnotations(body);
+    }
+    body = normalizeNamedGraphs(body);
+    body = rewriteMessageBlankLabels(body, messageIndex);
+    return body.trim();
+  }
+
+  function messageChunkHasRdf(body) {
+    return !isOnlyWhitespaceAndComments(stripDirectivesAndCommentsForEmptiness(body));
+  }
+
+  function normalizeRdfMessageLog(s) {
+    const withoutVersion = stripVersionDirectives(s);
+    const chunks = splitRdfMessageLog(withoutVersion);
+    const hash = simpleHashText(s);
+    const base = `urn:eyeling:message-log:${hash}`;
+    const stream = `<${base}#stream>`;
+    const envelopeIris = chunks.map((unused, idx) => `<${base}#m${String(idx + 1).padStart(3, '0')}>`);
+    const payloadIris = chunks.map((unused, idx) => `<${base}#m${String(idx + 1).padStart(3, '0')}/payload>`);
+    const out = [];
+
+    out.push(`${stream} ${RDF_TYPE_IRI} ${EYMSG.RDFMessageStream} .`);
+    out.push(`${stream} ${EYMSG.messageCount} "${chunks.length}"^^${XSD_INTEGER_IRI} .`);
+    if (envelopeIris.length) {
+      out.push(`${stream} ${EYMSG.orderedEnvelopes} (${envelopeIris.join(' ')}) .`);
+      out.push(`${stream} ${EYMSG.firstEnvelope} ${envelopeIris[0]} .`);
+      out.push(`${stream} ${EYMSG.lastEnvelope} ${envelopeIris[envelopeIris.length - 1]} .`);
+    }
+
+    for (let idx = 0; idx < chunks.length; idx += 1) {
+      const n = idx + 1;
+      const envelope = envelopeIris[idx];
+      const payload = payloadIris[idx];
+      const body = normalizeMessageChunk(chunks[idx], n);
+      const hasBody = messageChunkHasRdf(body);
+
+      out.push(`${stream} ${EYMSG.envelope} ${envelope} .`);
+      out.push(`${envelope} ${RDF_TYPE_IRI} ${EYMSG.MessageEnvelope} .`);
+      out.push(`${envelope} ${EYMSG.offset} "${n}"^^${XSD_INTEGER_IRI} .`);
+      out.push(`${envelope} ${EYMSG.payloadKind} ${hasBody ? EYMSG.nonEmpty : EYMSG.empty} .`);
+      if (idx + 1 < envelopeIris.length) out.push(`${envelope} ${EYMSG.nextEnvelope} ${envelopeIris[idx + 1]} .`);
+      if (hasBody) {
+        out.push(`${envelope} ${EYMSG.payloadGraph} ${payload} .`);
+        out.push(`${payload} ${LOG_NAME_OF_IRI} {`);
+        out.push(body);
+        out.push(`} .`);
+      }
+    }
+
+    return out.join('\n') + '\n';
+  }
+
+  if (hasMessageVersionDirective) return normalizeRdfMessageLog(text);
+
   if (hasTripleTerms) text = convertTripleTerms(text);
   if (hasAnnotationSyntax) text = convertAnnotations(text);
-  if (hasVersionDirective) text = stripVersionDirectives(text);
-  if (hasVersionDirective || hasNamedGraphCandidate) text = normalizeNamedGraphs(text);
+  if (hasVersionDirective || hasMessageVersionDirective) text = stripVersionDirectives(text);
+  if (hasVersionDirective || hasMessageVersionDirective || hasNamedGraphCandidate) text = normalizeNamedGraphs(text);
   return text;
 }
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "eyeling",
-  "version": "1.24.30",
+  "version": "1.24.31",
   "description": "A minimal Notation3 (N3) reasoner in JavaScript.",
   "main": "./index.js",
   "keywords": [

package/test/api.test.js

@@ -2864,6 +2864,64 @@ PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
       /:audit\s+\{[\s\S]*:workOrder\s+:basedOn\s+:factoryDataset\s*\.[\s\S]*\}/m,
     ],
   },
+  {
+    name: 'RDF mode replays RDF 1.2 message logs into eymsg envelopes',
+    opt: { proofComments: false, rdf: true },
+    input: {
+      sources: [
+        {
+          label: 'rules.n3',
+          text: `@prefix : <http://example.org/msgtest#>.
+@prefix eymsg: <https://eyereasoner.github.io/eyeling/vocab/message#>.
+@prefix log: <http://www.w3.org/2000/10/swap/log#>.
+@prefix list: <http://www.w3.org/2000/10/swap/list#>.
+
+{
+  ?S a eymsg:RDFMessageStream;
+    eymsg:messageCount ?Count;
+    eymsg:orderedEnvelopes ?Envelopes;
+    eymsg:firstEnvelope ?First.
+  ?Envelopes list:length ?Len.
+  ?First eymsg:nextEnvelope ?Second.
+  ?Second eymsg:nextEnvelope ?Third;
+    eymsg:payloadKind eymsg:empty.
+  ?First eymsg:payloadGraph ?P1.
+  ?P1 log:nameOf ?G1.
+  ?G1 log:includes { ?B1 :value 1. }.
+  ?Third eymsg:payloadGraph ?P3.
+  ?P3 log:nameOf ?G3.
+  ?G3 log:includes { ?B3 :value 2. }.
+} log:query {
+  :result :messageCount ?Count;
+    :listLength ?Len;
+    :empty ?Second;
+    :firstBlank ?B1;
+    :thirdBlank ?B3.
+}.
+`,
+        },
+        {
+          label: 'message-log.trig',
+          text: `VERSION "1.2-messages"
+@prefix : <http://example.org/msgtest#> .
+_:b :value 1 .
+MESSAGE
+# empty message
+MESSAGE
+_:b :value 2 .
+`,
+        },
+      ],
+    },
+    expect: [
+      /:result\s+:messageCount\s+"3"\^\^xsd:integer\s*\./m,
+      /:result\s+:listLength\s+3\s*\./m,
+      /:result\s+:empty\s+<urn:eyeling:message-log:[0-9a-f]+#m002>\s*\./m,
+      /:result\s+:firstBlank\s+_:src2_eyeling_m001_b\s*\./m,
+      /:result\s+:thirdBlank\s+_:src2_eyeling_m003_b\s*\./m,
+    ],
+  },
+
   {
     name: 'API empty input returns empty output',
     opt: { proofComments: false },