@velvetmonkey/vault-core 2.0.89 → 2.0.91

package/dist/astProtectedZones.d.ts

@@ -0,0 +1,16 @@
+/**
+ * AST-based protected zone detection
+ *
+ * Uses mdast tree to compute ProtectedZone[] more accurately than regex,
+ * especially for nested callouts, tables, and multi-line HTML comments.
+ *
+ * Supplemental regex handles Obsidian-specific syntax not in mdast:
+ * [[wikilinks]], bare URLs, #hashtags, %%comments%%
+ */
+import type { Root } from 'mdast';
+import type { ProtectedZone } from './types.js';
+/**
+ * Extract protected zones from an mdast tree.
+ */
+export declare function getProtectedZonesFromAst(tree: Root, content: string): ProtectedZone[];
+//# sourceMappingURL=astProtectedZones.d.ts.map

package/dist/astProtectedZones.js

@@ -0,0 +1,129 @@
+/**
+ * AST-based protected zone detection
+ *
+ * Uses mdast tree to compute ProtectedZone[] more accurately than regex,
+ * especially for nested callouts, tables, and multi-line HTML comments.
+ *
+ * Supplemental regex handles Obsidian-specific syntax not in mdast:
+ * [[wikilinks]], bare URLs, #hashtags, %%comments%%
+ */
+import { visit, SKIP } from 'unist-util-visit';
+function getOffset(node) {
+    const pos = node.position;
+    if (pos?.start &&
+        pos?.end &&
+        typeof pos.start.offset === 'number' &&
+        typeof pos.end.offset === 'number') {
+        return { start: pos.start.offset, end: pos.end.offset };
+    }
+    return null;
+}
+function zone(start, end, type) {
+    return { start, end, type };
+}
+/**
+ * Check if a blockquote node is an Obsidian callout (> [!type] syntax).
+ * If so, protect the entire blockquote — fixes nested callout bugs.
+ */
+function isObsidianCallout(node) {
+    if (node.type !== 'blockquote')
+        return false;
+    const children = node.children;
+    if (!children?.length)
+        return false;
+    const firstChild = children[0];
+    if (firstChild.type !== 'paragraph')
+        return false;
+    const paraChildren = firstChild.children;
+    if (!paraChildren?.length)
+        return false;
+    const firstInline = paraChildren[0];
+    if (firstInline.type !== 'text')
+        return false;
+    const value = firstInline.value;
+    if (typeof value !== 'string')
+        return false;
+    return /^\[![\w-]+\]/.test(value);
+}
+/**
+ * Extract protected zones from an mdast tree.
+ */
+export function getProtectedZonesFromAst(tree, content) {
+    const zones = [];
+    visit(tree, (node) => {
+        const off = getOffset(node);
+        if (!off)
+            return;
+        switch (node.type) {
+            case 'yaml':
+                // Frontmatter: extend zone to include the --- delimiters and trailing newline
+                {
+                    const fmStart = 0;
+                    const closingIdx = content.indexOf('---', off.end);
+                    const fmEnd = closingIdx !== -1 ? closingIdx + 3 : off.end;
+                    const afterEnd = fmEnd < content.length && content[fmEnd] === '\n' ? fmEnd + 1 : fmEnd;
+                    zones.push(zone(fmStart, afterEnd, 'frontmatter'));
+                }
+                break;
+            case 'code':
+                zones.push(zone(off.start, off.end, 'code_block'));
+                break;
+            case 'inlineCode':
+                zones.push(zone(off.start, off.end, 'inline_code'));
+                break;
+            case 'link':
+                zones.push(zone(off.start, off.end, 'markdown_link'));
+                break;
+            case 'heading':
+                zones.push(zone(off.start, off.end, 'header'));
+                break;
+            case 'html':
+                zones.push(zone(off.start, off.end, 'html_tag'));
+                break;
+            case 'inlineMath':
+            case 'math':
+                zones.push(zone(off.start, off.end, 'math'));
+                break;
+            case 'table':
+                zones.push(zone(off.start, off.end, 'table'));
+                break;
+            case 'blockquote':
+                if (isObsidianCallout(node)) {
+                    zones.push(zone(off.start, off.end, 'obsidian_callout'));
+                    return SKIP; // Don't recurse into callout children
+                }
+                break;
+        }
+    });
+    // Supplemental regex for Obsidian-specific syntax not in mdast
+    addRegexZones(content, zones);
+    // Sort by start position
+    zones.sort((a, b) => a.start - b.start);
+    return zones;
+}
+/**
+ * Add zones for Obsidian-specific syntax that mdast doesn't parse:
+ * - [[wikilinks]]
+ * - Bare URLs
+ * - #hashtags
+ * - %%comments%%
+ */
+function addRegexZones(content, zones) {
+    const patterns = [
+        [/\[\[[^\]]+\]\]/g, 'wikilink'],
+        [/https?:\/\/[^\s\)\]]+(?:\([^\)]+\))?[^\s\)\]]*/g, 'url'],
+        [/#[\w-]+/g, 'hashtag'],
+        [/%%.*?%%/gs, 'obsidian_comment'],
+    ];
+    for (const [pattern, type] of patterns) {
+        let match;
+        while ((match = pattern.exec(content)) !== null) {
+            zones.push({
+                start: match.index,
+                end: match.index + match[0].length,
+                type,
+            });
+        }
+    }
+}
+//# sourceMappingURL=astProtectedZones.js.map

package/dist/index.d.ts

@@ -7,7 +7,9 @@
 export type { EntityIndex, EntityCategory, EntityWithAliases, Entity, EntityWithType, ScanOptions, WikilinkOptions, WikilinkResult, ImplicitEntityConfig, ExtendedWikilinkOptions, ImplicitEntityMatch, ResolveAliasOptions, ProtectedZone, ProtectedZoneType, } from './types.js';
 export { scanVaultEntities, getAllEntities, getAllEntitiesWithTypes, getEntityName, getEntityAliases, loadEntityCache, saveEntityCache, ENTITY_CACHE_VERSION, } from './entities.js';
 export { applyWikilinks, processWikilinks, resolveAliasWikilinks, suggestWikilinks, detectImplicitEntities, findEntityMatches, IMPLICIT_EXCLUDE_WORDS, } from './wikilinks.js';
-export { getProtectedZones, isInProtectedZone, rangeOverlapsProtectedZone, } from './protectedZones.js';
+export { getProtectedZones, getProtectedZonesRegex, isInProtectedZone, rangeOverlapsProtectedZone, } from './protectedZones.js';
+export { parseMarkdown } from './parseMarkdown.js';
+export { getProtectedZonesFromAst } from './astProtectedZones.js';
 export { OperationLogger, createLoggerFromConfig, generateSessionId, getSessionId, setSessionId, } from './logging/index.js';
 export type { OperationLogEntry, SessionMetrics, AggregatedMetrics, LoggingConfig, ProductId, } from './logging/index.js';
 export { openStateDb, deleteStateDb, stateDbExists, searchEntities, searchEntitiesPrefix, getEntityByName, getEntitiesByAlias, getAllEntitiesFromDb, getEntityIndexFromDb, recordEntityMention, getEntityRecency, getAllRecency, setWriteState, getWriteState, deleteWriteState, setFlywheelConfig, getFlywheelConfig, getAllFlywheelConfig, saveFlywheelConfigToDb, loadFlywheelConfigFromDb, getStateDbMetadata, recordMergeDismissal, getDismissedMergePairs, saveVaultIndexCache, loadVaultIndexCache, getVaultIndexCacheInfo, rebuildEntitiesFts, loadContentHashes, saveContentHashBatch, renameContentHash, SCHEMA_VERSION, STATE_DB_FILENAME, FLYWHEEL_DIR, } from './sqlite.js';

package/dist/index.js

@@ -9,7 +9,10 @@ export { scanVaultEntities, getAllEntities, getAllEntitiesWithTypes, getEntityNa
 // Wikilinks
 export { applyWikilinks, processWikilinks, resolveAliasWikilinks, suggestWikilinks, detectImplicitEntities, findEntityMatches, IMPLICIT_EXCLUDE_WORDS, } from './wikilinks.js';
 // Protected zones
-export { getProtectedZones, isInProtectedZone, rangeOverlapsProtectedZone, } from './protectedZones.js';
+export { getProtectedZones, getProtectedZonesRegex, isInProtectedZone, rangeOverlapsProtectedZone, } from './protectedZones.js';
+// AST parsing
+export { parseMarkdown } from './parseMarkdown.js';
+export { getProtectedZonesFromAst } from './astProtectedZones.js';
 // Logging (unified cross-product logging)
 export { OperationLogger, createLoggerFromConfig, generateSessionId, getSessionId, setSessionId, } from './logging/index.js';
 // SQLite State Database

package/dist/parseMarkdown.d.ts

@@ -0,0 +1,17 @@
+/**
+ * Markdown AST parser for protected zone detection
+ *
+ * Uses mdast (fromMarkdown) with GFM, frontmatter, and math extensions.
+ * Returns null on parse error to trigger regex fallback.
+ */
+import type { Root } from 'mdast';
+export interface ParseMarkdownOptions {
+    /** Skip AST parsing entirely, return null */
+    forceRegex?: boolean;
+}
+/**
+ * Parse markdown content into an AST tree.
+ * Returns null on parse error (triggers regex fallback).
+ */
+export declare function parseMarkdown(content: string, options?: ParseMarkdownOptions): Root | null;
+//# sourceMappingURL=parseMarkdown.d.ts.map

package/dist/parseMarkdown.js

@@ -0,0 +1,32 @@
+/**
+ * Markdown AST parser for protected zone detection
+ *
+ * Uses mdast (fromMarkdown) with GFM, frontmatter, and math extensions.
+ * Returns null on parse error to trigger regex fallback.
+ */
+import { fromMarkdown } from 'mdast-util-from-markdown';
+import { gfmFromMarkdown } from 'mdast-util-gfm';
+import { frontmatterFromMarkdown } from 'mdast-util-frontmatter';
+import { mathFromMarkdown } from 'mdast-util-math';
+import { gfm } from 'micromark-extension-gfm';
+import { frontmatter } from 'micromark-extension-frontmatter';
+import { math } from 'micromark-extension-math';
+/**
+ * Parse markdown content into an AST tree.
+ * Returns null on parse error (triggers regex fallback).
+ */
+export function parseMarkdown(content, options) {
+    if (options?.forceRegex) {
+        return null;
+    }
+    try {
+        return fromMarkdown(content, {
+            extensions: [gfm(), frontmatter(['yaml']), math()],
+            mdastExtensions: [gfmFromMarkdown(), frontmatterFromMarkdown(['yaml']), mathFromMarkdown()],
+        });
+    }
+    catch {
+        return null;
+    }
+}
+//# sourceMappingURL=parseMarkdown.js.map

package/dist/protectedZones.d.ts

@@ -1,6 +1,9 @@
 /**
  * Protected zones detection for wikilink application
  *
+ * AST-first approach: uses mdast for accurate zone detection (nested callouts,
+ * tables, multi-line HTML), with regex fallback on parse failure.
+ *
  * These are areas in markdown content where wikilinks should NOT be applied:
  * - YAML frontmatter
  * - Code blocks (``` ... ```)
@@ -13,7 +16,8 @@
  * - Obsidian comments (%% ... %%)
  * - Math expressions ($ ... $ and $$ ... $$)
  * - Markdown headers (# to ###### at line start)
- * - Obsidian callouts (> [!type] syntax)
+ * - Obsidian callouts (> [!type] syntax — entire block, including nested)
+ * - Tables (GFM pipe tables)
  */
 import type { ProtectedZone } from './types.js';
 /**
@@ -22,7 +26,15 @@ import type { ProtectedZone } from './types.js';
  */
 export declare function findFrontmatterEnd(content: string): number;
 /**
- * Get all protected zones in content where wikilinks should not be applied
+ * Get all protected zones using regex-only detection (legacy/fallback).
+ * Exported for testing and explicit fallback use.
+ */
+export declare function getProtectedZonesRegex(content: string): ProtectedZone[];
+/**
+ * Get all protected zones in content where wikilinks should not be applied.
+ *
+ * AST-first: parses markdown into AST for accurate detection of nested
+ * callouts, tables, and HTML comments. Falls back to regex on parse failure.
  */
 export declare function getProtectedZones(content: string): ProtectedZone[];
 /**

package/dist/protectedZones.js

@@ -1,6 +1,9 @@
 /**
  * Protected zones detection for wikilink application
  *
+ * AST-first approach: uses mdast for accurate zone detection (nested callouts,
+ * tables, multi-line HTML), with regex fallback on parse failure.
+ *
  * These are areas in markdown content where wikilinks should NOT be applied:
  * - YAML frontmatter
  * - Code blocks (``` ... ```)
@@ -13,8 +16,11 @@
  * - Obsidian comments (%% ... %%)
  * - Math expressions ($ ... $ and $$ ... $$)
  * - Markdown headers (# to ###### at line start)
- * - Obsidian callouts (> [!type] syntax)
+ * - Obsidian callouts (> [!type] syntax — entire block, including nested)
+ * - Tables (GFM pipe tables)
  */
+import { parseMarkdown } from './parseMarkdown.js';
+import { getProtectedZonesFromAst } from './astProtectedZones.js';
 /**
  * Find where YAML frontmatter ends
  * @returns Character index after closing ---, or 0 if no frontmatter
@@ -58,9 +64,10 @@ function findPatternZones(content, pattern, type) {
     return zones;
 }
 /**
- * Get all protected zones in content where wikilinks should not be applied
+ * Get all protected zones using regex-only detection (legacy/fallback).
+ * Exported for testing and explicit fallback use.
  */
-export function getProtectedZones(content) {
+export function getProtectedZonesRegex(content) {
     const zones = [];
     // 1. YAML frontmatter (must be first)
     const frontmatterEnd = findFrontmatterEnd(content);
@@ -97,6 +104,19 @@ export function getProtectedZones(content) {
     zones.sort((a, b) => a.start - b.start);
     return zones;
 }
+/**
+ * Get all protected zones in content where wikilinks should not be applied.
+ *
+ * AST-first: parses markdown into AST for accurate detection of nested
+ * callouts, tables, and HTML comments. Falls back to regex on parse failure.
+ */
+export function getProtectedZones(content) {
+    const tree = parseMarkdown(content);
+    if (tree) {
+        return getProtectedZonesFromAst(tree, content);
+    }
+    return getProtectedZonesRegex(content);
+}
 /**
  * Check if a position is within any protected zone
  */

package/dist/types.d.ts

@@ -75,7 +75,7 @@ export interface ProtectedZone {
 /**
  * Types of protected zones
  */
-export type ProtectedZoneType = 'frontmatter' | 'code_block' | 'inline_code' | 'wikilink' | 'markdown_link' | 'url' | 'hashtag' | 'html_tag' | 'obsidian_comment' | 'math' | 'header' | 'obsidian_callout';
+export type ProtectedZoneType = 'frontmatter' | 'code_block' | 'inline_code' | 'wikilink' | 'markdown_link' | 'url' | 'hashtag' | 'html_tag' | 'obsidian_comment' | 'math' | 'header' | 'obsidian_callout' | 'table';
 /**
  * Options for entity scanning
  */

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@velvetmonkey/vault-core",
-  "version": "2.0.89",
+  "version": "2.0.91",
   "description": "Shared vault utilities for Flywheel ecosystem (entity scanning, wikilinks, protected zones)",
   "type": "module",
   "main": "dist/index.js",
@@ -36,7 +36,15 @@
     "prepublishOnly": "npm run build"
   },
   "dependencies": {
-    "better-sqlite3": "^11.0.0"
+    "better-sqlite3": "^11.0.0",
+    "mdast-util-from-markdown": "^2.0.0",
+    "mdast-util-gfm": "^3.0.0",
+    "mdast-util-frontmatter": "^2.0.0",
+    "mdast-util-math": "^3.0.0",
+    "micromark-extension-gfm": "^3.0.0",
+    "micromark-extension-frontmatter": "^2.0.0",
+    "micromark-extension-math": "^3.0.0",
+    "unist-util-visit": "^5.0.0"
   },
   "devDependencies": {
     "@types/better-sqlite3": "^7.6.8",