feat(logs): Retrieve and print logs of triggered workflow

Upgrade all dependencies

Author: aurelien-baudet

.github/workflows/build-test.yaml

@@ -163,6 +163,38 @@ jobs:
         expected: failure
         actual: ${{ steps.timeout-workflow.outcome }}
 
+  print-workflow-logs-test:
+    needs: [build]
+    runs-on: ubuntu-latest
+    name: "print-workflow-logs-test [trigger+wait|by workflow filename|print logs|shoud report timed_out]"
+    steps:
+    - name: Check out repository
+      uses: actions/checkout@v3
+    - name: Download dist
+      uses: actions/download-artifact@v3
+      with:
+        name: build
+        path: dist
+    - name: Invoke 'retrieve-logs' workflow and wait for result using this action
+      id: print-workflow-logs
+      uses: ./
+      with:
+        workflow: retrieve-logs.yml
+        token: ${{ secrets.PERSONAL_TOKEN }}
+        wait-for-completion-interval: 10s
+        wait-for-completion-timeout: 120s
+        workflow-logs: print
+      continue-on-error: true
+    - uses: nick-invision/assert-action@v1
+      with:
+        expected: timed_out
+        actual: ${{ steps.print-workflow-logs.outputs.workflow-conclusion }}
+    - uses: nick-invision/assert-action@v1
+      with:
+        expected: failure
+        actual: ${{ steps.print-workflow-logs.outcome }}
+      # TODO: add assertions on logs
+
     # - name: Invoke external workflow using this action
     #   uses: ./
     #   with:
@@ -172,7 +204,13 @@ jobs:
     #     ref: master 
 
   deploy:
-    needs: [echo-1-test, echo-2-test, long-running-test, failing-test, timeout-test]
+    needs:
+      - echo-1-test
+      - echo-2-test
+      - long-running-test
+      - failing-test
+      - timeout-test
+      - print-workflow-logs-test
     runs-on: ubuntu-latest
     steps:
     - name: Check out repository

.github/workflows/retrieve-logs.yml

@@ -0,0 +1,27 @@
+name: Retrieve logs
+
+on: 
+  workflow_dispatch:
+
+jobs:
+  echo:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Echo message
+        run: for i in {1..5000}; do; echo "Hello $i"; sleep 0.1; done    
+  
+  timeout:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Sleep
+        run: sleep 1200s
+  
+  more-real-example:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/setup-node@v3
+      with:
+        node-version: '16'
+    - name: Some command with npx
+      run: |
+        npx create-react-app example-app

README.md

@@ -15,56 +15,70 @@ _This action is a fork of `benc-uk/workflow-dispatch` to add support for waiting
 
 ## Inputs
 ### `workflow`
-**Required.** The name or the filename or ID of the workflow to trigger and run.
+> **Required.** The name or the filename or ID of the workflow to trigger and run.
 
 ### `token`
 
-**Required.** A GitHub access token (PAT) with write access to the repo in question. **NOTE.** The automatically provided token e.g. `${{ secrets.GITHUB_TOKEN }}` can not be used, GitHub prevents this token from being able to fire the  `workflow_dispatch` and `repository_dispatch` event. [The reasons are explained in the docs](https://docs.github.com/en/actions/reference/events-that-trigger-workflows#triggering-new-workflows-using-a-personal-access-token).  
-
-The solution is to manually create a PAT and store it as a secret e.g. `${{ secrets.PERSONAL_TOKEN }}`
+> **Required.** A GitHub access token (PAT) with write access to the repo in question. 
+> 
+> **NOTE.** The automatically provided token e.g. `${{ secrets.GITHUB_TOKEN }}` can not be used, GitHub prevents this token from being able to fire the  `workflow_dispatch` and `repository_dispatch` event. [The reasons are explained in the docs](https://docs.github.com/en/actions/reference/events-that-trigger-workflows#triggering-new-workflows-using-a-personal-access-token).  
+> The solution is to manually create a PAT and store it as a secret e.g. `${{ secrets.PERSONAL_TOKEN }}`
 
 ### `inputs`
-**Optional.** The inputs to pass to the workflow (if any are configured), this must be a JSON encoded string, e.g. `{ "myInput": "foobar" }`.
-
-All values must be strings (even if they are used as booleans or numbers in the triggered workflow). The triggered workflow should use `fromJson` function to get the right type
+> **Optional.** The inputs to pass to the workflow (if any are configured), this must be a JSON encoded string, e.g. `{ "myInput": "foobar" }`.
+>
+> All values must be strings (even if they are used as booleans or numbers in the triggered workflow). The triggered workflow should use `fromJson` function to get the right type
 
 ### `ref`
-**Optional.** The Git reference used with the triggered workflow run. The reference can be a branch, tag, or a commit SHA. If omitted the context ref of the triggering workflow is used. If you want to trigger on pull requests and run the target workflow in the context of the pull request branch, set the ref to `${{ github.event.pull_request.head.ref }}`
+> **Optional.** The Git reference used with the triggered workflow run. The reference can be a branch, tag, or a commit SHA. If omitted the context ref of the triggering workflow is used. If you want to trigger on pull requests and run the target workflow in the context of the pull request branch, set the ref to `${{ github.event.pull_request.head.ref }}`
 
 ### `repo`
-**Optional.** The default behavior is to trigger workflows in the same repo as the triggering workflow, if you wish to trigger in another GitHub repo "externally", then provide the owner + repo name with slash between them e.g. `microsoft/vscode`
+> **Optional.** The default behavior is to trigger workflows in the same repo as the triggering workflow, if you wish to trigger in another GitHub repo "externally", then provide the owner + repo name with slash between them e.g. `microsoft/vscode`
 
 ### `run-name` (since 3.0.0)
-**Optional.** The default behavior is to get the remote run ID based on the latest workflow name and date, if you have multiple of the same workflow running at the same time it can point to an incorrect run id.
-You can specify the run name to fetch the run ID based on the actual run name.
+> **Optional.** The default behavior is to get the remote run ID based on the latest workflow name and date, if you have multiple of the same workflow running at the same time it can point to an incorrect run id.
+> You can specify the run name to fetch the run ID based on the actual run name.
 
 ### `wait-for-completion`
-**Optional.** If `true`, this action will actively poll the workflow run to get the result of the triggered workflow. It is enabled by default. If the triggered workflow fails due to either `failure`, `timed_out` or `cancelled` then the step that has triggered the other workflow will be marked as failed too.
+> **Optional.** If `true`, this action will actively poll the workflow run to get the result of the triggered workflow. It is enabled by default. If the triggered workflow fails due to either `failure`, `timed_out` or `cancelled` then the step that has triggered the other workflow will be marked as failed too.
 
 ### `wait-for-completion-timeout`
-**Optional.** The time to wait to mark triggered workflow has timed out. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `wait-for-completion` is `false`. Default is `1h`
+> **Optional.** The time to wait to mark triggered workflow has timed out. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `wait-for-completion` is `false`. Default is `1h`
 
 ### `wait-for-completion-interval`
-**Optional.** The time to wait between two polls for getting run status. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `wait-for-completion` is `false`. Default is `1m`.
-**/!\ Do not use a value that is too small to avoid `API Rate limit exceeded`**
+> **Optional.** The time to wait between two polls for getting run status. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `wait-for-completion` is `false`. Default is `1m`.
+> 
+> **/!\ Do not use a value that is too small to avoid `API Rate limit exceeded`**
 
 ### `display-workflow-run-url`
-**Optional.** If `true`, it displays in logs the URL of the triggered workflow. It is useful to follow the progress of the triggered workflow. It is enabled by default.
+> **Optional.** If `true`, it displays in logs the URL of the triggered workflow. It is useful to follow the progress of the triggered workflow. It is enabled by default.
 
 ### `display-workflow-run-url-timeout`
-**Optional.** The time to wait for getting the workflow run URL. If the timeout is reached, it doesn't fail and continues. Displaying the workflow URL is just for information purpose. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `display-workflow-run-url` is `false`. Default is `10m`
+> **Optional.** The time to wait for getting the workflow run URL. If the timeout is reached, it doesn't fail and continues. Displaying the workflow URL is just for information purpose. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `display-workflow-run-url` is `false`. Default is `10m`
 
 ### `display-workflow-run-url-interval`
-**Optional.** The time to wait between two polls for getting workflow run URL. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `display-workflow-run-url` is `false`. Default is `1m`.
-**/!\ Do not use a value that is too small to avoid `API Rate limit exceeded`**
+> **Optional.** The time to wait between two polls for getting workflow run URL. The time must be suffixed by the time unit e.g. `10m`. Time unit can be `s` for seconds, `m` for minutes and `h` for hours. It has no effect if `display-workflow-run-url` is `false`. Default is `1m`.
+> 
+> **/!\ Do not use a value that is too small to avoid `API Rate limit exceeded`**
+
+### `workflow-logs`
+> **Optional.** Indicate what to do with logs of the triggered workflow:
+> 
+> * `print`: Retrieve the logs for each job of the triggered workflow and print into the logs of the job that triggered the workflow.
+> * `ignore`: Do not retrieve log of triggered workflow at all (default).
+> 
+> Only available when `wait-for-completion` is `true`. 
+> 
+> Default is `ignore`.
+
 
 ## Outputs
 ### `workflow-url`
-The URL of the workflow run that has been triggered. It may be undefined if the URL couldn't be retrieved (timeout reached) or if `wait-for-completion` and `display-workflow-run-url` are both `false`
+> The URL of the workflow run that has been triggered. It may be undefined if the URL couldn't be retrieved (timeout reached) or if `wait-for-completion` and `display-workflow-run-url` are > both `false`
 
 ### `workflow-conclusion`
-The result of the triggered workflow. May be one of `success`, `failure`, `cancelled`, `timed_out`, `skipped`, `neutral`, `action_required`. The step in your workflow will fail if the triggered workflow completes with `failure`, `cancelled` or `timed_out`. Other workflow conlusion are considered success.
-Only available if `wait-for-completion` is `true`
+> The result of the triggered workflow. May be one of `success`, `failure`, `cancelled`, `timed_out`, `skipped`, `neutral`, `action_required`. The step in your workflow will fail if the triggered workflow completes with `failure`, `cancelled` or `timed_out`. Other workflow conlusion are considered success.
+> Only available if `wait-for-completion` is `true`
 
 
 ## Example usage
@@ -115,3 +129,14 @@ Only available if `wait-for-completion` is `true`
   if: always()
   run: echo "Another Workflow conclusion: ${{ steps.trigger-step.outputs.workflow-conclusion }}"
 ```
+
+
+## Contributions
+
+Thanks to:
+
+* [samirergaibi](https://github.com/samirergaibi)
+* [rui-ferreira](https://github.com/rui-ferreira)
+* [robbertvdg](https://github.com/robbertvdg)
+* [samit2040](https://github.com/samit2040)
+* [jonas-schievink](https://github.com/jonas-schievink)

action.yaml

@@ -44,6 +44,10 @@ inputs:
     description: 'Time to wait (+unit) between two polls to get run status'
     required: false
     default: 1m
+  workflow-logs:
+    description: 'Indicate what to do with logs of the triggered workflow. `ignore` do not retrieve logs from tiggered workflow. `print` retrieves logs from triggered workflow and print in the workflow that triggered the other workflow.'
+    required: false
+    default: ignore
 
 runs:
   using: 'node16'