Separate gui and cli tests for speed

.github/workflows/build_and_test.yml

@@ -33,7 +33,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        test-type: [ 'performance-tests', 'unit-tests', 'ui-tests' ]
+        test-type: [ 'performance-tests', 'unit-tests', 'gui-tests', 'cli-tests' ]
         python-version: [ '3.8', '3.11', '3.12' ]
         os: [ ubuntu-latest ]
     uses: ./.github/workflows/test_ert.yml
@@ -58,7 +58,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        test-type: [ 'performance-tests', 'unit-tests', 'ui-tests' ]
+        test-type: [ 'performance-tests', 'unit-tests', 'gui-tests', 'cli-tests' ]
         python-version: [ '3.8', '3.12' ]
         os: [ 'macos-13', 'macos-14', 'macos-14-large']
         exclude:
@@ -80,7 +80,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        test-type: [ 'performance-tests', 'unit-tests', 'ui-tests' ]
+        test-type: [ 'performance-tests', 'unit-tests', 'gui-tests', 'cli-tests' ]
         python-version: [ '3.12' ]
         os: [ 'macos-latest' ]
     uses: ./.github/workflows/test_ert.yml

.github/workflows/test_ert.yml

@@ -42,11 +42,15 @@ jobs:
       run: |
         uv pip install ".[dev]"
 
-    - name: UI Test
-      if: inputs.test-type == 'ui-tests'
+    - name: GUI Test
+      if: inputs.test-type == 'gui-tests'
       run: |
         pytest --cov=ert --cov-report=xml:cov1.xml --junit-xml=junit.xml -v --mpl --benchmark-disable tests/ui_tests/gui
-        pytest --cov=ert --cov-report=xml:cov2.xml --junit-xml=junit.xml -n logical -v --benchmark-disable  --dist loadgroup tests/ui_tests/cli
+
+    - name: CLI Test
+      if: inputs.test-type == 'cli-tests'
+      run: |
+        pytest --cov=ert --cov-report=xml:cov1.xml --junit-xml=junit.xml -n logical -v --benchmark-disable  --dist loadgroup tests/ui_tests/cli
 
     - name: Unit Test
       if: inputs.test-type == 'unit-tests'

codecov.yml

@@ -4,4 +4,4 @@ fixes:
 comment:
     # The code coverage is made up of 4 test runs so only after all coverage
     # reports have been uploaded will the comparison be sane
-    after_n_builds: 12
+    after_n_builds: 16