Skip to content

Set up web server

If your configuration be ready to fake, it's time to set up a web server to fake them and provides this service to outside!

Run by CLI

It's very easy to set up and run the web server by PyFake-API-Server. With default setting, namely using default value auto of option --app-type, it would automatically detect which the Python web framework it can use in the current runtime environment. Therefore, the running server log message would be different with different Python web framework.

>>> fake rest-server run -p <configuration path>
[2023-06-06 21:55:53 +0800] [78900] [INFO] Starting gunicorn 20.1.0
[2023-06-06 21:55:53 +0800] [78900] [INFO] Listening at: http://127.0.0.1:9672 (78900)
[2023-06-06 21:55:53 +0800] [78900] [INFO] Using worker: sync
[2023-06-06 21:55:53 +0800] [78915] [INFO] Booting worker with pid: 78915

>>> fake rest-server run -p <configuration path>
INFO:     Started server process [78594]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:9672 (Press CTRL+C to quit)

After running the web server, let's try to send HTTP request by command line curl to the API and verify its feature.

>>> curl -X GET http://127.0.0.1:9672/foo
"This is Foo home API."%

Run by Docker

Although Docker is easy, but we should be careful to use it with options. It has three options are necessary you must use:

  • -v

    Mount the configuration of fake APIs into container. By Docker's rule, you need to use absolute path here. For the target path it would mount to, it must be in directory /mit-fake-api-server/ because where is the command line works at.

  • -e

    Set environment variable into container. It has some parameters it provides to use.

    • CONFIG_PATH

      The configuration path. By default, it would try to use file api.yaml in the root directory where it works at. In the other words, it would try to find the file /mit-fake-api-server/api.yaml in Docker container in default.

    • WEB_FRAMEWORK

      Which Python web framework it should use to set up server. By default, it would use auto to do it.

    • WORKERS

      How many workers it should use to process HTTP requests. By default, its value is 1.

    • LOG_LEVEL

      The log level of web server it should export. By default, its value is info.

  • -p

    It needs to export the port number 9672 of web server so that it can provide the service to fake APIs for outside.

Let's give you a sample command line to set up fake server by Docker:

>>> docker run --name fake-server \
               -v <configuration root directory>:/mit-fake-api-server/<configuration root directory> \
               -e CONFIG_PATH=<configuration path>
               -p 9672:9672 \
               fake-api-server:v0.1.0
Hint: Still being confused about the configuration path setting? Let's demonstrate some usage scenarios to you.

Assume that we have below files tree:

local root directory (/User/foo/fake-api-demo)
├── file1
├── file2
├── api.yaml
├── folder1
│   ├── file3
│   ├── file4
│   └── beta-api.yaml
  • Scenario 1: Use the configuration api.yaml

Your command line would be like below:

>>> docker run --name fake-server \
               -v /User/foo/fake-api-demo:/mit-fake-api-server \
               -p 9672:9672 \
               fake-api-server:v0.1.0

You can mount the all files in folder /User/foo/fake-api-demo into folder /mit-fake-api-server of container. And the file tree in container would be as below:

container root directory (/)
├── mit-fake-api-server
│   ├── file1
│   ├── file2
│   ├── api.yaml
│   └── folder1
|       ├── file3
|       ├── file4
│       └── beta-api.yaml

Do you remember the command line working directory is mit-fake-api-server and the default value is api.yaml of environment CONFIG_PATH? And it exactly has a file at path mit-fake-api-server/api.yaml so it would set up server successfully!

  • Scenario 2: Use the configuration beta-api.yaml

If you try to set up server by configuration beta-api.yaml, your command line would be like below:

>>> docker run --name fake-server \
               -v /User/foo/fake-api-demo:/mit-fake-api-server \
               -e CONFIG_PATH=./folder1/beta-api.yaml
               -p 9672:9672 \
               fake-api-server:v0.1.0

The mount setting is the same as previous scenario so the file tree also be the same as previous one. And the command line working directory is mit-fake-api-server. So we just need to give it a relative path in working directory. Therefore, the value would be ./folder1/beta-api.yaml.

What the detail usage of PyFake-API-Server by Docker?

Please refer to PyFake-API-Server's image overview in Docker hub to get more details.

Great! Now the web server for faking your API has done, and it would start to provide this service for other projects or services.