Why glaze help export exists

The Glazed help system stores documentation in an SQLite-backed store at runtime. That makes querying fast, but it also means the data is trapped inside a running process. glaze help export solves this by letting you dump the entire help tree—or a filtered slice of it—to formats you can use elsewhere.

This matters when you want to:

• back up or version-control your help content as plain files,
• feed help metadata into an external search index or CMS,
• generate reports on documentation coverage (which commands have examples, which topics are missing),
• or ship a SQLite database of your app's documentation alongside the binary.

The command is built on the same HelpSystem and Store primitives that glaze serve and glaze render-site use, so the exported data is always consistent with what the CLI and the web browser show.

Basic usage

The simplest invocation exports every loaded section as JSON:

glaze help export --output json

By default, --with-content is true, so the JSON includes the full markdown body of each section. If you only need metadata, disable content:

glaze help export --with-content=false --output csv

Export formats

The --format flag controls the export target:

Tabular export (--format glazed)

This is the default. It behaves like any other Glazed command: you can choose the output serializer with --output.

# Pretty-printed JSON with full content
glaze help export --output json

# Compact CSV for spreadsheet import
glaze help export --with-content=false --output csv

# YAML for human review
glaze help export --output yaml

The tabular mode emits one row per section with these columns:

File export (--format files)

File mode writes each section to a separate .md file, reconstructing the original frontmatter so the files can be re-loaded into Glazed later.

# Default: typed subdirectories
glaze help export --format files --output-path ./exported-help

This produces a directory tree like:

exported-help/
  general-topic/
    help-system.md
    sections-guide.md
  example/
    json-output.md
  application/
    real-world-pipeline.md
  tutorial/
    getting-started.md

If you prefer a flat layout, use --flatten-dirs:

glaze help export --format files --output-path ./flat --flatten-dirs

Which produces:

flat/
  help-system.md
  sections-guide.md
  json-output.md
  ...

Each exported file is a valid Glazed help section. You can load it back with:

glaze serve ./exported-help

SQLite export (--format sqlite)

SQLite mode creates a new database file containing all exported sections in the same schema that Glazed uses internally.

glaze help export --format sqlite --output-path ./help.db

You can then query it directly:

sqlite3 ./help.db "SELECT slug, title, section_type FROM sections LIMIT 5;"

The exported database is fully self-contained. It does not depend on the original application binary, so you can ship it to tools that understand Glazed's schema or use it as a snapshot for offline browsing.

Filtering exports

You rarely need to export everything. The command accepts the same metadata filters that glaze help uses for querying:

# Export only examples about JSON
glaze help export --type Example --topic json --output yaml

# Export sections related to a specific command
glaze help export --command serve --format files --output-path ./serve-docs

# Export by exact slug (useful for one-off backups)
glaze help export --slug help-system --output json

# Combine multiple filters (AND logic)
glaze help export --type Example --topic advanced --command json --output csv

Available filters:

All filters are combined with AND logic. If no filters are given, every loaded section is exported.

Practical examples

Back up documentation before a refactor

Before restructuring your help files, snapshot the current state:

glaze help export --format files --output-path ./docs-backup-$(date +%Y%m%d)

Generate a CSV inventory for a spreadsheet

Create a lightweight inventory of all sections for editorial review:

glaze help export --with-content=false --output csv > section-inventory.csv

Ship a documentation database with your app

Build a release artifact that contains both the binary and a queryable help DB:

glaze help export --format sqlite --output-path ./dist/myapp-help.db

Reconstruct markdown from the internal store

If you loaded sections programmatically and want to recover the original .md files:

glaze help export --format files --flatten-dirs --output-path ./recovered

How the export command is wired

glaze help export is implemented as a BareCommand in pkg/help/cmd/export.go. It adds a glazed section to its schema so that --output json/csv/table/yaml flags are available, but bypasses the GlazeProcessor when --format is files or sqlite so it can write to disk instead of stdout.

The command lives under the help subcommand tree. It is registered automatically when a binary calls help_cmd.SetupCobraRootCommand(...). The glaze binary does this in cmd/glaze/main.go, so the export verb is available out of the box. Advanced integrations can still call help_cmd.AddExportCommand(...) directly when constructing a custom help command tree; the helper is idempotent and will not add a duplicate export child.

Troubleshooting

See Also

• [glaze help serve-external-help-sources](glaze help serve-external-help-sources) — Serve exported help from other Glazed binaries and snapshots
• [glaze help help-system](glaze help help-system) — Overview of the Glazed help system
• [glaze help serve-help-over-http](glaze help serve-help-over-http) — Browse help in a web browser
• [glaze help export-help-static-website](glaze help export-help-static-website) — Export help as a static site
• [glaze help writing-help-entries](glaze help writing-help-entries) — How to write help sections
• [glaze help sections-guide](glaze help sections-guide) — Deep dive into section metadata and the query DSL