The Glean CLI tool

The Glean CLI tool (glean) can create and write data into databases, perform a variety of admin tasks on databases, and also do one-off queries.

The glean tool accepts all the common options to specify how to connect to access the databases.

The available commands are as follows:

glean create

Create a new database.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database
• --finish
  Also mark the DB as complete
• --stacked REPO
  Create a stacked database on top of REPO.
• --property NAME=VALUE
  Set properties when creating a DB
• FILE..
  File(s) of facts to write into the database (JSON). See Writing data to Glean.

glean write

Write facts to a database.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database
• FILE..
  File(s) of facts to write into the database (JSON). See Writing data to Glean.
• --finish
  Also mark the DB as complete

glean finish

Notify server that a database is complete.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database

glean dump

Dump the contents of the specified database into a file.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database
• FILE
  File to write the facts into

glean delete

Delete a database.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database

glean derive

Derive and store a predicate. See Derived Predicates.

• --page-bytes BYTES
  Maximum number of bytes per page

• --page-facts FACTS
  Maximum number of facts per page

• PREDICATE
  Predicates to derive

glean index

Index some source code using one of the known indexers.

The form of the command in general is

glean index LANGUAGE DIR --repo NAME/HASH

There may also be additional options accepted for each LANGUAGE; try glean index LANGUAGE --help to find out.

For information on each indexer, see Indexers.

glean query

Execute an Angle query and print the results, or write them to a file.

• --repo NAME/HASH or --repo-name NAME
  Specifies the database to query

• --page-bytes BYTES
  Maximum number of bytes per page

• --page-facts FACTS
  Maximum number of facts per page

• --recursive
  Fetch nested facts (slower)

• --limit FACTS
  Maximum number of facts to query

• -o,--output FILE
  Output the facts to a file

• --stats FILE
  Output stats to a file ('-' for stdout)

• --timeout MILLISECONDS Override the default query timeout

• QUERY query to execute (@file to read from file, - for stdin)

• --omit-results Don't print results; use with --stat to get a count of results

glean restore

Restore a database from backup.

• LOCATOR
  DB location, see :list-all in glean shell.

Alternatively the DB to restore can be specified by:

• --repo NAME/HASH or --repo-name NAME and (--repo-hash HASH or --date YYY-MM-DD)

glean validate

Perform checks on the internal integrity of a database. This is for testing and debugging Glean itself.

a local database

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database

• --no-typecheck
  Don't typecheck facts.

• --no-keys
  Don't verify key uniqueness

• --limit N
  Only validate the first N facts

glean validate-schema

Validate a schema. Checks that a new schem does not modify any of the predicates in the existing schema, which could lead to problems.

• FILE
  Name of schema file

glean stats

Get fact counts and sizes. Like the :statistics command in the shell.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database

glean unfinish

Unfinish a local database (turn it from complete to incomplete state). This is for testing and development and not for routine use: once a database is marked complete it could be replicated, so we shouldn't be modifying it.

• --repo NAME/HASH or --repo-name NAME --repo-hash HASH
  Specifies the name and hash of the database