Running GSQL

GSQL provides several modes of execution, ranging from an interactive shell to automated command files.

Language Basics

• Identifiers: Case-sensitive names for variables and types. Must start with a letter or underscore (no leading digits).
• Keywords: Semantic words like VERTEX or EDGE. These are not case-sensitive but are often reserved.
• Comments:
  • Single-line: # or //
  • Multi-line: /* ... */

The GSQL Shell

Enter the shell by typing gsql in your terminal.

Authentication

bashterminal
Copy
# Standard login (defaults to tigergraph user)
$ gsql -u <username> -p <password>

# Token-based login (Modern/Recommended)
$ gsql -t <token_value>

Multi-Line Mode

In the interactive shell, use BEGIN and END to wrap multi-line statements:

gsqlterminal
Copy
GSQL > BEGIN
GSQL > SELECT member_id, last_name
GSQL > FROM Member
GSQL > WHERE age >= 21
GSQL > END

Running Scripts

You can automate GSQL by saving commands in a .gsql file.

Execution Methods:

1. From Linux Shell:

   bashterminal
   Copy
   $ gsql my_script.gsql

2. Inline Commands:

   bashterminal
   Copy
   $ gsql 'RUN QUERY my_query(10)'

3. Inside GSQL Shell:

   gsqlterminal
   Copy
   GSQL > @my_script.gsql

Essential Catalog Commands

• ls: Displays the entire catalog (Vertex types, Edge types, Graphs, Jobs, Queries).
• SHOW [VERTEX|EDGE|GRAPH] <name>: Displays detailed properties of a specific component.
• SET exit_on_error = FALSE: Allows a script to continue even if a semantic error occurs.

System Reset

Caution: The --reset option erases all data, schemas, and definitions.

bashterminal
Copy
$ gsql --reset