feat: re-parse after running --fix

Related to, but does not fully address, #19.

> NOTE: this PR also enables the vitest plugin

Adds a step to the test stage that re-runs oxlint after it was previously run
with `--fix`. This should help us catch syntax errors created by buggy fixes.

We should also consider running `typecheck` scripts for repositories that have
them.

.github/workflows/ecosystem-ci.yml

@@ -119,10 +119,21 @@ jobs:
 
       - name: Check Panic
         working-directory: ${{ matrix.path }}
+        # NOTE (@DonIsaac): should we enable import-plugin?
+        # NOTE (@DonIsaac): should we test nursery rules too?
         run: |
           set +e # disable exit on run
 
-          ./oxlint --fix --jest-plugin --jsdoc-plugin --jsx-a11y-plugin --nextjs-plugin --react-perf-plugin -D all --silent
+          ./oxlint \
+            --fix \
+            --jsdoc-plugin \
+            --jest-plugin \
+            --vitest-plugin \
+            --jsx-a11y-plugin \
+            --nextjs-plugin \
+            --react-perf-plugin \
+            -D all \
+            --silent
 
           EXIT_CODE=$?
 
@@ -135,6 +146,23 @@ jobs:
           echo "exitcode=0" >> $GITHUB_OUTPUT
           exit 0
 
+      # Check for invalid syntax after lint rule violations have been fixed.
+      # All lint rules are set to `warning` here, so an exit code > 0 is either
+      # a parse error or a panic.
+      - name: Re-Parse After `--fix`
+        working-directory: ${{ matrix.path }}
+        run: |
+          ./oxlint \
+            --fix \
+            --jsdoc-plugin \
+            --jest-plugin \
+            --vitest-plugin \
+            --jsx-a11y-plugin \
+            --nextjs-plugin \
+            --react-perf-plugin \
+            -W all \
+            --silent
+
   comment:
     needs: test
     if: ${{ always() }}