Examples

The examples provided are all generated from the PostgreSQL database for an instance of Apache Airflow, run locally with docker compose as described here .

The command to generate the first three "All" examples looked as follows, substituting -d 1 with the depth used in each example:

db-diagram \
    -d 1 \
    -t 'Airflow PostgreSQL Database Diagrams' \
    -md docs/examples/airflow/depth=1/diagrams.md \
    postgresql+psycopg://airflow:[email protected]:5432/airflow

The command to generate the filtered subset examples looked as follows, substituting -d 1 with the depth used in each example, and substituting -i 'ab_*' for the appropriate inclusion or exclusion filter(s):

db-diagram \
    -i 'ab_*' \
    -d 1 \
    -t 'Airflow PostgreSQL Database Diagrams' \
    -md docs/examples/airflow/depth=1/diagrams.md \
    postgresql+psycopg://airflow:[email protected]:5432/airflow

...the "other" examples' commands looked as follows (they just had several -e options to filter out all other subsets):

```bash
db-diagram \
    -e 'ab_*' \
    -e 'asset_*' \
    -e 'dag_*' \
    -e 'task_*' \
    -d 1 \
    -t 'Airflow PostgreSQL Database Diagrams' \
    -md docs/examples/airflow/depth=1/diagrams.md \
    postgresql+psycopg://airflow:[email protected]:5432/airflow

The command to generate the second three examples, for which images are pre-rendered as SVG (scalable vector graphics), looked as follows—substituting -d 1 with the depth used in each example:

db-diagram \
    -d 1 \
    -t 'Airflow PostgreSQL Database Diagrams' \
    -svg docs/examples/airflow/depth=1/svg \
    -t dark \
    -md docs/examples/airflow/depth=1/diagrams-svg.md \
    postgresql+psycopg://airflow:[email protected]:5432/airflow

As you can see above, in the second set of examples, we specify an SVG image directory. This causes db-diagram to render each diagram as an SVG image, using the specified theme (such as -t dark). Optionally, a CSS background color can be provided (by default this is -bg transparent), or for greater control a JSON mermaid config file can be specified using the -cf option.

Hopefully these examples will provide you with insight concerning the effect of db-diagram options, particularly concerning depth, image rendering, and include/exclude filtering.

• Depth: Note that, when you specify a depth greater than one—you start to generate diagrams which may be difficult to read. Usually, the default depth (1) is optimal, but it will only show all relationships for a single table in each diagram.
• Pre-Rendering Images: When you pre-render images—your pages load more quickly, but diagrams are not uniformly sized throughout the document. When you don't pre-render images, for databases with many tables—there will be a noticeable delay before diagrams are rendered.
• Filtering: Usually your best option, if possible, is to break down your documentation into subsets of the tables in your database, thereby getting fairly responsive page load times, but with uniform sizing of elements in your diagrams.