@blackwell-systems/gcf 0.6.1 → 1.0.0

package/dist/decode_generic.js

@@ -1,18 +1,18 @@
 import { decode } from './decode.js';
+import { parseScalar, parseQuotedString, splitRespectingQuotes, splitFieldDecl, MISSING, ATTACHMENT, } from './scalar.js';
 /**
- * Decode any GCF text (tabular or graph profile) back into a JS value.
- * Returns objects, arrays, and primitives matching the original structure.
- *
- * If the input starts with "GCF " (graph profile), falls back to decode()
- * and returns the Payload as a plain object.
+ * Decode GCF v2.0 generic or graph profile text into a JS value.
  */
 export function decodeGeneric(input) {
     input = input.trimEnd();
     if (!input)
-        return null;
+        throw new Error('missing_header: empty input');
     const lines = input.split('\n');
-    // Graph profile fallback.
-    if (lines[0].startsWith('GCF ')) {
+    const header = lines[0].replace(/\r$/, '');
+    if (!header.startsWith('GCF '))
+        throw new Error('missing_header: first line does not begin with GCF');
+    const profile = parseHeaderProfile(header);
+    if (profile === 'graph') {
         const p = decode(input);
         return {
             tool: p.tool,
@@ -30,99 +30,139 @@ export function decodeGeneric(input) {
                 source: e.source,
                 target: e.target,
                 edgeType: e.edgeType,
-                ...(e.status ? { status: e.status } : {}),
+                status: e.status ?? '',
             })),
         };
     }
+    if (profile !== 'generic')
+        throw new Error(`unknown_profile: ${profile}`);
+    // Filter body.
+    const contentLines = [];
+    let summaryLine = '';
+    let deferredSectionCount = 0;
+    for (let i = 1; i < lines.length; i++) {
+        const l = lines[i].replace(/\r$/, '');
+        if (l === '')
+            continue;
+        // Tab check.
+        for (let j = 0; j < l.length; j++) {
+            if (l[j] === '\t')
+                throw new Error('tab_indentation: tabs in leading whitespace');
+            if (l[j] !== ' ')
+                break;
+        }
+        const trimmed = l.trimStart();
+        if (trimmed.startsWith('# '))
+            continue;
+        if (trimmed.startsWith('##! ')) {
+            summaryLine = trimmed;
+            continue;
+        }
+        if (trimmed.startsWith('## ') && trimmed.includes('[?]'))
+            deferredSectionCount++;
+        contentLines.push(l);
+    }
+    // Validate ##! summary counts.
+    if (summaryLine && deferredSectionCount > 0) {
+        validateSummaryCounts(summaryLine, deferredSectionCount, contentLines);
+    }
+    if (contentLines.length === 0)
+        return {};
+    const first = contentLines[0].trimStart();
+    // Root scalar.
+    if (first.startsWith('=')) {
+        if (contentLines.length > 1)
+            throw new Error('trailing_characters: extra lines after root scalar');
+        return parseScalar(first.slice(1), false);
+    }
+    // Root array.
+    if (first.startsWith('## [')) {
+        const [arr] = parseArrayFromHeader(contentLines, 0, 0, first.slice(3));
+        return arr;
+    }
+    // Root object.
     const result = {};
-    parseObject(lines, 0, 0, result);
+    parseObjectBody(contentLines, 0, 0, result);
     return result;
 }
-function parseObject(lines, start, depth, out) {
-    const indent = '  '.repeat(depth);
+function parseHeaderProfile(header) {
+    const parts = header.split(/\s+/);
+    if (parts.length < 2)
+        throw new Error('missing_profile');
+    const seen = new Set();
+    let profile = '';
+    for (let i = 1; i < parts.length; i++) {
+        const eq = parts[i].indexOf('=');
+        if (eq < 0)
+            throw new Error(`malformed_header_field: ${parts[i]}`);
+        const key = parts[i].slice(0, eq);
+        if (seen.has(key))
+            throw new Error(`duplicate_header_field: ${key}`);
+        seen.add(key);
+        if (key === 'profile')
+            profile = parts[i].slice(eq + 1);
+    }
+    if (!profile)
+        throw new Error('missing_profile');
+    return profile;
+}
+function parseObjectBody(lines, start, depth, out) {
+    const ind = '  '.repeat(depth);
     let i = start;
     while (i < lines.length) {
-        const raw = lines[i].replace(/\r$/, '');
-        if (raw === '' || raw.startsWith('# ')) {
-            i++;
-            continue;
-        }
-        if (depth > 0 && !raw.startsWith(indent))
+        const line = lines[i];
+        if (depth > 0 && !line.startsWith(ind))
             break;
-        const content = depth > 0 ? raw.slice(indent.length) : raw;
-        // Skip _summary.
-        if (content.startsWith('## _summary')) {
-            i++;
-            continue;
+        const content = depth > 0 ? line.slice(ind.length) : line;
+        if (content.length > 0 && content[0] === ' ') {
+            throw new Error('invalid_indent: indentation increases by more than one level');
         }
-        // Tabular array or section header.
+        // Array section.
         if (content.startsWith('## ')) {
-            const header = content.slice(3);
-            const bracketIdx = header.indexOf(' [');
-            if (bracketIdx >= 0) {
-                const name = header.slice(0, bracketIdx);
-                const rest = header.slice(bracketIdx + 2);
-                const closeBracket = rest.indexOf(']');
-                if (closeBracket >= 0) {
-                    const afterBracket = rest.slice(closeBracket + 1);
-                    if (afterBracket.startsWith('{')) {
-                        // Tabular with fields.
-                        const fieldEnd = afterBracket.indexOf('}');
-                        if (fieldEnd >= 0) {
-                            const fields = afterBracket.slice(1, fieldEnd).split(',');
-                            i++;
-                            const [rows, consumed] = parseTabularRows(lines, i, depth, fields);
-                            out[name] = rows;
-                            i += consumed;
-                            continue;
-                        }
-                    }
-                    else {
-                        const countStr = rest.slice(0, closeBracket);
-                        if (countStr === '0') {
-                            out[name] = [];
-                            i++;
-                            continue;
-                        }
-                        // Non-uniform array.
-                        i++;
-                        const [items, consumed] = parseNonUniformArray(lines, i, depth);
-                        out[name] = items;
-                        i += consumed;
-                        continue;
-                    }
-                }
+            const hdr = content.slice(3);
+            const bi = hdr.indexOf(' [');
+            if (bi >= 0) {
+                const name = parseKeyFromHeader(hdr.slice(0, bi));
+                checkDup(out, name);
+                const [arr, consumed] = parseArrayFromHeader(lines, i, depth, hdr.slice(bi));
+                out[name] = arr;
+                i += consumed;
+                continue;
             }
-            // Plain section header.
-            let name = header;
-            const bi = name.indexOf(' [');
-            if (bi >= 0)
-                name = name.slice(0, bi);
+            const name = parseKeyFromHeader(hdr);
+            checkDup(out, name);
             i++;
             const nested = {};
-            const consumed = parseObject(lines, i, depth + 1, nested);
+            const consumed = parseObjectBody(lines, i, depth + 1, nested);
             out[name] = nested;
             i += consumed;
             continue;
         }
-        // Inline primitive array: name[N]: val1,val2,...
-        const bracketIdx = content.indexOf('[');
-        if (bracketIdx > 0) {
-            const colonIdx = content.indexOf(']: ');
-            if (colonIdx > bracketIdx) {
-                const name = content.slice(0, bracketIdx);
-                const valsStr = content.slice(colonIdx + 3);
-                out[name] = valsStr.split(',').map(v => parseValue(v.trim()));
-                i++;
-                continue;
+        // Inline array.
+        if (!content.startsWith('@') && !content.startsWith('##')) {
+            const bracketIdx = content.indexOf('[');
+            if (bracketIdx > 0) {
+                const rest = content.slice(bracketIdx);
+                const closeIdx = rest.indexOf(']');
+                if (closeIdx >= 0) {
+                    const after = rest.slice(closeIdx + 1);
+                    if (after.startsWith(': ') || after === ':') {
+                        const name = parseKeyFromHeader(content.slice(0, bracketIdx));
+                        checkDup(out, name);
+                        const [arr] = parseArrayFromHeader(lines, i, depth, rest);
+                        out[name] = arr;
+                        i++;
+                        continue;
+                    }
+                }
             }
         }
         // Key=value.
-        const eqIdx = content.indexOf('=');
+        const eqIdx = findKeyValueSplit(content);
         if (eqIdx > 0) {
-            const key = content.slice(0, eqIdx);
-            const val = content.slice(eqIdx + 1);
-            out[key] = parseValue(val);
+            const name = parseKeyFromHeader(content.slice(0, eqIdx));
+            checkDup(out, name);
+            out[name] = parseScalar(content.slice(eqIdx + 1), false);
             i++;
             continue;
         }
@@ -130,107 +170,317 @@ function parseObject(lines, start, depth, out) {
     }
     return i - start;
 }
-function parseTabularRows(lines, start, depth, fields) {
-    const indent = '  '.repeat(depth);
+function findKeyValueSplit(s) {
+    if (!s.length)
+        return -1;
+    if (s[0] === '"') {
+        for (let i = 1; i < s.length; i++) {
+            if (s[i] === '\\') {
+                i++;
+                continue;
+            }
+            if (s[i] === '"')
+                return (i + 1 < s.length && s[i + 1] === '=') ? i + 1 : -1;
+        }
+        return -1;
+    }
+    return s.indexOf('=');
+}
+function parseKeyFromHeader(s) {
+    s = s.trim();
+    if (s.length >= 2 && s[0] === '"')
+        return parseQuotedString(s);
+    return s;
+}
+function checkDup(obj, key) {
+    if (key in obj)
+        throw new Error(`duplicate_key: ${key}`);
+}
+function parseArrayFromHeader(lines, headerLine, depth, bracketPart) {
+    const bp = bracketPart.trimStart();
+    if (!bp.startsWith('['))
+        throw new Error('invalid_count');
+    const closeIdx = bp.indexOf(']');
+    if (closeIdx < 0)
+        throw new Error('invalid_count');
+    const countStr = bp.slice(1, closeIdx);
+    const afterBracket = bp.slice(closeIdx + 1);
+    let count = -1;
+    if (countStr !== '?')
+        count = parseCount(countStr);
+    if (count === 0 && !afterBracket.startsWith('{') && !afterBracket.startsWith(':')) {
+        return [[], 1];
+    }
+    // Inline.
+    if (afterBracket.startsWith(': ') || afterBracket === ':') {
+        const valsStr = afterBracket.startsWith(': ') ? afterBracket.slice(2) : '';
+        if (!valsStr) {
+            if (count >= 0 && count !== 0)
+                throw new Error(`count_mismatch: declared ${count}, got 0`);
+            return [[], 1];
+        }
+        const vals = splitRespectingQuotes(valsStr, ',');
+        if (count >= 0 && vals.length !== count)
+            throw new Error(`count_mismatch: declared ${count}, got ${vals.length}`);
+        return [vals.map(v => parseScalar(v.trim(), false)), 1];
+    }
+    // Tabular.
+    if (afterBracket.startsWith('{')) {
+        const braceEnd = findClosingBrace(afterBracket);
+        if (braceEnd < 0)
+            throw new Error('invalid field declaration');
+        const fields = splitFieldDecl(afterBracket.slice(0, braceEnd + 1));
+        const [rows, consumed] = parseTabularBody(lines, headerLine + 1, depth, fields, count);
+        if (count >= 0 && rows.length !== count)
+            throw new Error(`count_mismatch: declared ${count}, got ${rows.length}`);
+        return [rows, consumed + 1];
+    }
+    // Expanded.
+    const [items, consumed] = parseExpandedBody(lines, headerLine + 1, depth);
+    if (count >= 0 && items.length !== count)
+        throw new Error(`count_mismatch: declared ${count}, got ${items.length}`);
+    return [items, consumed + 1];
+}
+function findClosingBrace(s) {
+    let inQuote = false, escaped = false;
+    for (let i = 0; i < s.length; i++) {
+        if (escaped) {
+            escaped = false;
+            continue;
+        }
+        if (s[i] === '\\' && inQuote) {
+            escaped = true;
+            continue;
+        }
+        if (s[i] === '"') {
+            inQuote = !inQuote;
+            continue;
+        }
+        if (s[i] === '}' && !inQuote)
+            return i;
+    }
+    return -1;
+}
+function parseTabularBody(lines, start, depth, fields, expectedCount) {
+    const ind = '  '.repeat(depth);
     const rows = [];
     let i = start;
     while (i < lines.length) {
-        const raw = lines[i].replace(/\r$/, '');
-        if (raw === '') {
-            i++;
-            continue;
-        }
-        const content = depth > 0 ? (raw.startsWith(indent) ? raw.slice(indent.length) : null) : raw;
+        const line = lines[i];
+        const content = depth > 0 ? (line.startsWith(ind) ? line.slice(ind.length) : null) : line;
         if (content === null)
             break;
-        if (content.startsWith('## '))
+        if (content.startsWith('## ') || content.startsWith('##!'))
+            break;
+        if (content.length > 0 && content[0] === ' ') {
+            const trimmed = content.trimStart();
+            if (trimmed.startsWith('.'))
+                throw new Error(`orphan_attachment: ${trimmed}`);
             break;
-        if (content.startsWith('# ')) {
-            i++;
-            continue;
         }
         let rowData = content;
-        let hasNested = false;
+        let rowHasID = false;
         if (rowData.startsWith('@')) {
             const sp = rowData.indexOf(' ');
             if (sp > 0) {
                 rowData = rowData.slice(sp + 1);
-                hasNested = true;
+                rowHasID = true;
             }
         }
-        const vals = rowData.split('|');
+        const vals = splitRespectingQuotes(rowData, '|');
+        if (vals.length !== fields.length)
+            throw new Error(`row_width_mismatch: expected ${fields.length}, got ${vals.length}`);
         const row = {};
+        const attachmentFields = [];
         for (let j = 0; j < fields.length; j++) {
-            row[fields[j]] = j < vals.length ? parseValue(vals[j]) : null;
+            const parsed = parseScalar(vals[j], true);
+            if (parsed === MISSING)
+                continue;
+            if (parsed === ATTACHMENT) {
+                attachmentFields.push(fields[j]);
+                continue;
+            }
+            row[fields[j]] = parsed;
         }
         i++;
-        if (hasNested) {
-            const nestedIndent = indent + '  ';
+        if (rowHasID && attachmentFields.length > 0) {
+            const attIndent = ind + '  ';
+            const resolved = new Set();
             while (i < lines.length) {
-                const nl = lines[i].replace(/\r$/, '');
-                if (!nl.startsWith(nestedIndent))
+                const al = lines[i];
+                if (!al.startsWith(attIndent))
                     break;
-                const nc = nl.slice(nestedIndent.length);
-                if (nc.startsWith('.')) {
-                    const fieldName = nc.slice(1);
-                    i++;
-                    const nested = {};
-                    const consumed = parseObject(lines, i, depth + 2, nested);
-                    row[fieldName] = nested;
-                    i += consumed;
-                }
-                else {
+                const ac = al.slice(attIndent.length);
+                if (!ac.startsWith('.'))
                     break;
-                }
+                const [name, val, consumed] = parseAttachment(lines, i, ac.slice(1), depth + 2);
+                if (resolved.has(name))
+                    throw new Error(`duplicate_attachment: ${name}`);
+                resolved.add(name);
+                row[name] = val;
+                i += consumed;
+            }
+            for (const f of attachmentFields) {
+                if (!resolved.has(f))
+                    throw new Error(`missing_attachment: ${f}`);
+            }
+        }
+        if (!rowHasID || attachmentFields.length === 0) {
+            const attIndent = ind + '  ';
+            if (i < lines.length && lines[i].startsWith(attIndent)) {
+                const peek = lines[i].slice(attIndent.length);
+                if (peek.startsWith('.'))
+                    throw new Error(`orphan_attachment: ${peek}`);
             }
         }
         rows.push(row);
+        if (expectedCount >= 0 && rows.length >= expectedCount)
+            break;
     }
     return [rows, i - start];
 }
-function parseNonUniformArray(lines, start, depth) {
-    const indent = '  '.repeat(depth);
+function parseAttachment(lines, lineIdx, rest, depth) {
+    let name;
+    let afterName;
+    if (rest[0] === '"') {
+        let closeIdx = -1;
+        for (let j = 1; j < rest.length; j++) {
+            if (rest[j] === '\\') {
+                j++;
+                continue;
+            }
+            if (rest[j] === '"') {
+                closeIdx = j;
+                break;
+            }
+        }
+        if (closeIdx < 0)
+            throw new Error('unterminated_quote');
+        name = parseQuotedString(rest.slice(0, closeIdx + 1));
+        afterName = rest.slice(closeIdx + 1).trimStart();
+    }
+    else {
+        const sp = rest.indexOf(' ');
+        if (sp < 0)
+            throw new Error(`invalid attachment: ${rest}`);
+        name = rest.slice(0, sp);
+        afterName = rest.slice(sp).trimStart();
+    }
+    if (afterName.startsWith('{}')) {
+        const nested = {};
+        const consumed = parseObjectBody(lines, lineIdx + 1, depth, nested);
+        return [name, nested, consumed + 1];
+    }
+    if (afterName.startsWith('[')) {
+        const [arr, consumed] = parseArrayFromHeader(lines, lineIdx, depth, afterName);
+        return [name, arr, consumed];
+    }
+    throw new Error(`invalid attachment form: ${afterName}`);
+}
+function parseExpandedBody(lines, start, depth) {
+    const ind = '  '.repeat(depth);
     const items = [];
     let i = start;
     while (i < lines.length) {
-        const raw = lines[i].replace(/\r$/, '');
-        if (raw === '') {
-            i++;
-            continue;
-        }
-        const content = depth > 0 ? (raw.startsWith(indent) ? raw.slice(indent.length) : null) : raw;
+        const line = lines[i];
+        const content = depth > 0 ? (line.startsWith(ind) ? line.slice(ind.length) : null) : line;
         if (content === null)
             break;
-        if (content.startsWith('## '))
+        if (content.startsWith('## ') || content.startsWith('##!'))
             break;
-        if (content.startsWith('@')) {
-            const sp = content.indexOf(' ');
-            if (sp > 0) {
-                items.push(parseValue(content.slice(sp + 1)));
-            }
+        if (!content.startsWith('@'))
+            break;
+        const sp = content.indexOf(' ');
+        if (sp < 0)
+            break;
+        const idStr = content.slice(1, sp);
+        const id = parseInt(idStr, 10);
+        if (!isNaN(id) && id !== items.length) {
+            throw new Error(`invalid_item_id: expected @${items.length}, got @${idStr}`);
+        }
+        const marker = content.slice(sp + 1);
+        if (marker.startsWith('=')) {
+            items.push(parseScalar(marker.slice(1), false));
             i++;
+            continue;
         }
-        else {
-            break;
+        if (marker.startsWith('{}')) {
+            const nested = {};
+            i++;
+            const consumed = parseObjectBody(lines, i, depth + 1, nested);
+            items.push(nested);
+            i += consumed;
+            continue;
+        }
+        if (marker.startsWith('[')) {
+            const [arr, consumed] = parseArrayFromHeader(lines, i, depth + 1, marker);
+            items.push(arr);
+            i += consumed;
+            continue;
         }
+        break;
     }
     return [items, i - start];
 }
-function parseValue(s) {
-    if (s === '-')
-        return null;
-    if (s === 'true')
-        return true;
-    if (s === 'false')
-        return false;
-    if (s === '""')
-        return '';
-    if (s.length >= 2 && s[0] === '"' && s[s.length - 1] === '"') {
-        return s.slice(1, -1).replace(/\\"/g, '"').replace(/\\\\/g, '\\');
-    }
-    const n = Number(s);
-    if (!isNaN(n) && s !== '')
-        return n;
-    return s;
+function parseCount(s) {
+    if (s === '0')
+        return 0;
+    if (!s.length || s[0] === '0')
+        throw new Error(`invalid_count: ${s}`);
+    const n = parseInt(s, 10);
+    if (isNaN(n) || String(n) !== s)
+        throw new Error(`invalid_count: ${s}`);
+    return n;
+}
+function validateSummaryCounts(summaryLine, deferredCount, contentLines) {
+    // Parse counts from "##! summary counts=N,M,..."
+    const parts = summaryLine.split(/\s+/);
+    let countsStr = '';
+    for (const p of parts) {
+        if (p.startsWith('counts=')) {
+            countsStr = p.slice(7);
+            break;
+        }
+    }
+    if (!countsStr)
+        return;
+    const countVals = countsStr.split(',');
+    if (countVals.length !== deferredCount) {
+        throw new Error(`count_mismatch: summary has ${countVals.length} count entries but ${deferredCount} deferred sections`);
+    }
+    // Count actual items per deferred section.
+    const actualCounts = [];
+    let inDeferred = false;
+    let currentCount = 0;
+    for (const l of contentLines) {
+        const trimmed = l.trimStart();
+        if (trimmed.startsWith('## ') && trimmed.includes('[?]')) {
+            if (inDeferred)
+                actualCounts.push(currentCount);
+            inDeferred = true;
+            currentCount = 0;
+            continue;
+        }
+        if (trimmed.startsWith('## ')) {
+            if (inDeferred) {
+                actualCounts.push(currentCount);
+                inDeferred = false;
+            }
+            continue;
+        }
+        if (inDeferred && !trimmed.startsWith(' ') && !trimmed.startsWith('.')) {
+            currentCount++;
+        }
+    }
+    if (inDeferred)
+        actualCounts.push(currentCount);
+    for (let i = 0; i < countVals.length; i++) {
+        const declared = parseInt(countVals[i], 10);
+        if (isNaN(declared))
+            throw new Error(`count_mismatch: invalid count value "${countVals[i]}"`);
+        if (i < actualCounts.length && declared !== actualCounts[i]) {
+            throw new Error(`count_mismatch: section ${i} declared ${declared} in summary, actual ${actualCounts[i]}`);
+        }
+    }
 }
 //# sourceMappingURL=decode_generic.js.map