How to use VSCode

VSCode (or better the open source version VSCodium) may be used to edit files.

Here are some useful tricks.

Perlcritic

One way to have perlcritic work is the following:

• install the perlcritic extension
• add a perlcritic.sh at the root of your project with following content:

  #!/usr/bin/env bash
  . .envrc >/dev/null 2>&1
  docker compose run --rm --no-deps backend perlcritic "$@" 2>/dev/null

  the second line is useful only if you use direnv
• chmod +x perlcritic.sh
• patch perlcritic by editing its files, following sfodje/perlcritic issue #26
• edit the perlcritic configuration in workspace to set the value (after adjusting the pathname for your own OFF setup):
  • Executable: /path/to/your/own/OFF-server/perlcritic.sh

Perl Language Server

The extension Language Server and Debugger is harder to work with!

Note: This setup does not work yet, but might be getting close. It is probably due to https://github.com/richterger/Perl-LanguageServer/issues/131.

• install the extension
• add a script shell-into-appserver.sh in the project:

  #!/usr/bin/env bash
  declare -x PATH=$PATH:/usr/local/bin/
  source .envrc
  COMMAND=$(echo "$@" | sed 's/^.*perl /perl /')
  >&2 echo "launching $COMMAND"
  docker compose run --rm --no-deps -T -p 127.0.0.  1:13603:13603 backend $COMMAND

  Note: the second line is useful only if you use direnv
• chmod +x shell-into-appserver.sh
• Edit workspace settings to have those settings:

      "perl": {
          "enable": true,
          "perlInc": ["/opt/product-opener/lib", "/opt/perl/local/lib/perl5"],
          "ignoreDirs": ["/opt/perl/local/lib/perl5", ".  vscode"],
          "fileFilter": [".pm", ".pl", ".t"],
          "sshAddr": "dummy",
          "sshUser": "dummy",
          "sshCmd": "./shell-into-appserver.sh",
          "sshWorkspaceRoot": "/opt/product-opener",
          "logLevel": 2
      },

Remote container ?

Note: at the moment we do not support the Remote Container extension. While we can consider using it, it has some drawbacks because not all the project is contained within the "backend" container. For example everything related to nodejs is in the "frontend" container. So it means making a quite complete Docker image on its own with all the tooling necessary.