Computed Data

Extenote derives certain data dynamically from your vault content. This page explains what gets computed, when it’s computed, and how to choose between dynamic computation and persisted values.

Computed vs Persisted

Some data can be either:

• Computed on-demand: Calculated fresh each time it’s needed
• Persisted: Saved to frontmatter via CLI commands, then read back

When to Use Each Approach

Dynamic Computation

• Small vaults: Computation is fast enough
• Always fresh: No risk of stale data
• Read-only use: Viewing, searching, querying

Persisted Values

• Large vaults: Avoid re-computing on every load
• Static builds: Generate data once, use in builds
• Audit trails: Track what was computed and when

Computation Details

cited_in (Citation Tracking)

Scans all projects that include shared-references for Pandoc/Quarto citations ([@key]) and builds a reverse index.

Dynamic: computeCitedIn(objects, config) returns CitedInMap with:

• citedIn: Map<string, string[]> - citation_key → [project1, project2, …]
• totalCitations: number
• scannedProjects: string[]

Persist: extenote sync-citations writes cited_in field to each bibtex entry.

Cross-References (Wikilinks)

Parses [[wikilinks]] in document bodies and builds forward/backward reference maps.

Dynamic only: computeAllCrossRefs(objects) returns refs per object.

Uses parseWikiLinks() to extract links, buildObjectIndex() for ID lookups.

Tag Index

Builds a tree structure of all tags with their objects.

Dynamic only:

• getAllTags(objects) - flat list of all tags
• buildTagTree(objects) - hierarchical tree with counts

Tags support namespaces via : (e.g., collection:data-leverage).

Project Membership

Determines which objects belong to each project based on source configuration and includes.

Dynamic only: objectBelongsToProject(obj, projectName, config) checks:

1. Object’s sourceId matches project sources
2. Object type matches project includes

Vault Summary

Aggregates statistics about the vault.

Dynamic only: summarizeVault(vault) returns:

• Total object count by type
• Public/private/unlisted counts
• Issue counts by severity
• Project breakdowns

Web UI: Computed Data View

The web app’s status page shows:

• Last vault load time
• Computation times for each derived data type
• Memory usage
• Option to recompute specific data

Access via the “System” or “Status” menu item.

Performance Considerations

For vaults with 1000+ objects, consider persisting frequently-accessed computed data.

CLI Commands

# Persist cited_in to bibtex entries
extenote sync-citations

# Preview what would be updated
extenote sync-citations --dry-run

# View vault summary (computed)
extenote summary

API Reference

import {
  computeCitedIn,
  getCitedIn,
  computeAllCrossRefs,
  buildTagTree,
  summarizeVault
} from "@extenote/core";

// Compute cited_in dynamically
const citedInMap = computeCitedIn(vault.objects, vault.config);

// Get cited_in for specific entry (checks persisted first, falls back to computed)
const projects = getCitedIn(entry, citedInMap);

// Build full cross-reference graph
const crossRefs = computeAllCrossRefs(vault.objects);

// Build tag tree
const tagTree = buildTagTree(vault.objects);

// Get vault summary
const summary = summarizeVault(vault);