Add psycopg example on Direct Database Connections (#174)

Author: Kludex

docs/guides/advanced/direct_database_connections.md

@@ -12,13 +12,17 @@ To connect, you'll first need to generate generate database credentials from you
 
 The credentials generated are a PostgreSQL URI which can be used as a connection string for compatible tools. These will only be shown by the Logfire platform once, so save them to a secure location for future use!
 
-## Example: pgcli
+## Connecting to the database
+
+As mentioned, you can use the generated credentials to connect to the database using any tool that supports PostgreSQL connections.
+
+### Using `pgcli`
 
 [`pgcli`](https://www.pgcli.com/) is a command-line tool to access PostgreSQL databases.
 
 Using the credentials generated in the previous step as the argument to `pgcli`, you can connect directly to **Logfire**:
 
-```
+```bash
 $ pgcli postgresql://<user>:<password>@db.logfire.dev:5432/proj_david-test  # REDACTED
 Version: 4.0.1
 Home: http://pgcli.com
@@ -31,7 +35,7 @@ proj_david-test> select start_timestamp, message from records limit 10;
 | 2024-04-28 10:50:41.665576+00 | GET /contact/                          |
 | 2024-04-28 10:50:41.711119+00 | GET /contact/ http send response.start |
 | 2024-04-28 10:50:41.709458+00 | response 500                           |
-| 2024-04-28 10:50:38.50534+00  | action=view-cart size=517 i=0          |
+| 2024-04-28 10:50:38.505343+00 | action=view-cart size=517 i=0          |
 | 2024-04-28 10:50:39.446668+00 | action=view-faq size=637 i=2           |
 | 2024-04-28 10:50:38.681198+00 | action=view-terms size=216 i=3         |
 | 2024-04-28 10:50:39.416706+00 | action=view-product size=380 i=0       |
@@ -41,4 +45,40 @@ SELECT 10
 Time: 0.218s
 ```
 
-With the flexibility of PostgreSQL access available to you, we can't wait to hear what you do with the Logfire platform!
+### Using `psycopg`
+
+[`psycopg`](https://www.psycopg.org/) is a Python library that allows you to connect to PostgreSQL databases.
+
+You can use the credentials generated in the previous step to connect to **Logfire**:
+
+```python title='main.py'
+import os
+
+import psycopg
+
+POSTGRES_URI = os.getenv('POSTGRES_URI')
+
+with psycopg.connect(POSTGRES_URI) as conn:
+    cursor = conn.cursor()
+    cursor.execute('SELECT start_timestamp, message FROM records LIMIT 10')
+    for start_timestamp, message in cursor.fetchall():
+        print(f'{start_timestamp}: {message}')
+```
+
+Set the `export POSTGRES_URI=...` environment variable to the URI generated in the previous step, and run the
+script with `python main.py`:
+
+```bash
+2024-04-28 10:50:41.681886+00: action=view-faq size=549 i=0
+2024-04-28 10:50:41.711747+00: GET /contact/ http send response.body
+2024-04-28 10:50:41.665576+00: GET /contact/
+2024-04-28 10:50:41.711119+00: GET /contact/ http send response.start
+2024-04-28 10:50:41.709458+00: response 500
+2024-04-28 10:50:38.505343+00: action=view-cart size=517 i=0
+2024-04-28 10:50:39.446668+00: action=view-faq size=637 i=2
+2024-04-28 10:50:38.681198+00: action=view-terms size=216 i=3
+2024-04-28 10:50:39.416706+00: action=view-product size=380 i=0
+2024-04-28 10:50:38.394237+00: sub-sub-sub-action=logout
+```
+
+With the flexibility of PostgreSQL access available to you, we can't wait to hear what you do with **Logfire**! :fire: