chore: move v8 to master branch (#2549)

v8 of the SDK has lived on `pysdk-release-v8` for some time. We now move
it back to master, while separately, moving v7 (ex-master) to
`pysdk-release-v7`.

Author: haakonvt

.coveragerc

@@ -1,5 +1,7 @@
 [run]
-source=cognite
+source =
+    cognite
+    tests
 omit =
     cognite/client/_proto/*
 

.gemini/styleguide.md

@@ -5,30 +5,31 @@
 - **Strong Typing**: Use type hints extensively with MyPy. Avoid `Any` when possible
 - **Type Safety**: Use dataclasses models for complex data
   structures instead of untyped dictionaries
-- **CogniteResource**: Use `CogniteResource` for DTO (Data Transfer Object) that are request or response
-  objects to/from the Cognite API
-- **CogniteObject**: Use `CogniteObject` for objects that are within `CogniteResource` objects.
+- **CogniteResource**: Use `CogniteResource` for DTO (Data Transfer Object) that are request or
+  response objects to/from the Cognite API
 - **IO Safety**: Always use typed data structures for file operations and data parsing
 - **Readability**: Code should be immediately understandable
 - **Maintainability**: Write code that is easy to modify and extend
 - **Consistency**: Follow established patterns across the codebase
+- **Code Comments**: Use comments sparingly. They should always add necessary context, or
+  explain unintuitive code and never feel like a repeat of the code.
 
 ## Principles on doing pull request reviews
 
-- **Main point first.** Start with the key feedback or required action.
-- **Be concise.** Use short, direct comments. Avoid unnecessary explanations.
-- **Actionable suggestions.** If something needs fixing, state exactly what and how.
-- **One issue per comment.** Separate unrelated feedback for clarity.
-- **Code, not prose.** Prefer code snippets or examples over long text.
-- **Background only if needed.** Add context only if the main point isn't obvious.
+- **Main point first** Start with the key feedback or required action.
+- **Be concise** Use short, direct comments. Avoid unnecessary explanations.
+- **Actionable suggestions** If something needs fixing, state exactly what and how.
+- **One issue per comment** Separate unrelated feedback for clarity.
+- **Code, not prose** Prefer code snippets or examples over long text.
+- **Background only if needed** Add context only if the main point isn't obvious.
 
 ## How to do pull request summaries
 
-- **Short recap.** Summarize the main point of the PR in one or two sentences.
-- **Don't repeat the PR description.** Only add new or clarifying information.
-- **Be brief unless needed.** Only write a longer summary if the PR description
+- **Short recap** Summarize the main point of the PR in one or two sentences.
+- **Don't repeat the PR description** Only add new or clarifying information.
+- **Be brief unless needed** Only write a longer summary if the PR description
   is missing crucial details.
-- **Extend, don't duplicate.** If more detail is needed, clearly state what is
+- **Extend, don't duplicate** If more detail is needed, clearly state what is
   missing from the PR description and add only the necessary context.
 
 ## Line Length and Formatting
@@ -73,7 +74,7 @@ import logging
 from pathlib import Path
 from typing import TYPE_CHECKING
 
-from requests import Response
+from httpx import Response
 
 from cognite.client.data_classes import Asset
 
@@ -91,7 +92,8 @@ if TYPE_CHECKING:
 
 ## Docstrings
 
-Use concise docstrings with Args/Returns in google-style format. Based on repository patterns:
+Use concise docstrings with Args/Returns in google-style format. 'Raises' can be used, but is not
+required. Based on repository patterns:
 
 ```python
 def render_header(header: str) -> str:
@@ -123,7 +125,8 @@ def walk_sdk_documentation(content: Tag, parser: Parser[T]) -> Iterable[T]:
 ## Error Handling
 
 - **Specific exceptions**: Avoid broad `Exception` catches
-- **Graceful handling**: Provide meaningful error messages
+- **Graceful handling**: Provide meaningful error messages, for example "got type foo,
+  expected type bar".
 - **Type safety**: Return `None` or use Union types for fallible operations
 
 ```python
@@ -138,27 +141,6 @@ def validate_data_modeling_identifier(space: str | None, external_id: str | None
 
 - **Pre-commit**: `pre-commit run --all-files` for comprehensive checks
 
-
-## Data Structures
-
-**Prefer typed structures**:
-
-```python
-# Good
-@dataclass
-class FunctionError:
-    function_name: str
-    message: str
-
-# Good
-class QueryCompletion(BaseModel):
-    query: str
-    variables: dict[str, Any]
-
-# Avoid
-error_data = {"function_name": "foo", "message": "bar"}
-```
-
 ## Logging
 
 - Use the `logging` module with appropriate levels
@@ -192,6 +174,4 @@ error_data = {"function_name": "foo", "message": "bar"}
   guarantee against regression.
 - Counter-intuitively, for data classes in the SDK, testing their load/dump
   methods should not be done as this is done automatically
-  (see `tests/tests_unit/test_base.py`). This rule applies to any subclass of
-  `CogniteObject`, `CogniteResource`, `CogniteUpdate`,
-   `WriteableCogniteResource` or `WriteableCogniteResourceList`.
+  (see `tests/tests_unit/test_base.py`). This rule applies to any subclass of `CogniteResource` or `CogniteUpdate`.

.github/workflows/build.yml

@@ -4,7 +4,6 @@ name: build
 on:
   merge_group:
   pull_request:
-    branches: [master]
 
 concurrency:
   group: pr-${{ github.workflow }}-${{ github.head_ref }}
@@ -21,6 +20,16 @@ jobs:
       - name: Linting and static code checks
         run: pre-commit run --all-files --show-diff-on-failure
 
+  verify-sync-codegen:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - uses: ./.github/actions/setup
+        with:
+          extras: "-E pandas"
+      - name: Ensures sync code is up-to-date, with no stale files
+        run: python scripts/sync_client_codegen/main.py verify
+
   build_docs:
     runs-on: ubuntu-latest
     steps:
@@ -43,15 +52,15 @@ jobs:
       - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # v6
       - uses: ./.github/actions/setup
       - name: Test core
-        run: pytest tests/tests_unit -n8 --dist loadscope --maxfail 10 -m 'not dsl' --test-deps-only-core
+        run: pytest tests/tests_unit -n6 --dist loadscope --maxfail 10 -m 'not dsl' --test-deps-only-core
 
   test_full:
     runs-on: ${{ matrix.os }}
     strategy:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, windows-latest]
-        python-version: ["3.10", "3.11", "3.12", "3.13"]
+        python-version: ["3.10", "3.11", "3.12", "3.13", "3.14"]
     steps:
       - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # v6
       - uses: ./.github/actions/setup
@@ -73,6 +82,7 @@ jobs:
         run: pytest --durations=10 --cov --cov-report term --cov-report xml:coverage.xml -n8 --dist loadscope --reruns 2 --maxfail 20
 
       - uses: codecov/codecov-action@18283e04ce6e62d37312384ff67231eb8fd56d24 # v5
+        # Report code coverage from windows runner; xdist + ubuntu gives incorrect results
         if: matrix.os == 'windows-latest' && matrix.python-version == '3.10'
         with:
           token: ${{ secrets.CODECOV_TOKEN }}

.github/workflows/manual-pre-release.yaml

@@ -0,0 +1,38 @@
+name: Release
+
+on:
+  push:
+    branches:
+      - 'pysdk-release-v8'
+
+permissions:
+  contents: read
+
+jobs:
+  test_full_build_and_release:
+    runs-on: ubuntu-latest
+    environment: CD
+    steps:
+      - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8
+
+      - uses: ./.github/actions/setup
+
+      - name: Ensure version is a v8-release
+        run: |
+          VERSION=$(poetry version --short)
+          echo "Detected version: $VERSION"
+          if [[ "$VERSION" =~ ^8\.[0-9]+\.[0-9]+$ ]]; then
+            echo "✅ Version format is valid for a v8-release."
+          else
+            echo "❌ Error: Version '$VERSION' is not a valid v8-release (must be 8.x.y format)."
+            exit 1
+          fi
+
+      - name: Build package
+        run: poetry build
+
+      - name: Release to PyPI
+        env:
+          TWINE_USERNAME: __token__
+          TWINE_PASSWORD: ${{ secrets.PYPI_API_TOKEN }}
+        run: twine upload --skip-existing --disable-progress-bar --verbose dist/*

.github/workflows/release-pypi-v8.yaml

@@ -3,7 +3,7 @@ name: build
 
 on:
   pull_request:
-    branches: [master]
+    branches: [master, pysdk-release-v8]
 
 jobs:
   build_and_test_jupyter_pyodide:
@@ -23,11 +23,11 @@ jobs:
         with:
           node-version: "22"
       - name: Install dependencies
-        run: npm install pyodide@0.27.6 # JupyterLite currently using pyodide 0.27.6
+        run: npm install pyodide@0.29.0 # JupyterLite currently using pyodide 0.29.0
       - name: Install cognite-sdk in pyodide environment
         run: |
           whl_file=$(find dist -name "*.whl" | sed 's|^dist/||') # Find the built wheel file, remove dist/ prefix
           echo "Found built wheel file: $whl_file"
           SDK_FILE_PATH=$whl_file \
-          PACKAGES="[\"pyodide-http\", \"http://localhost:3000/dist/$whl_file\"]" \
+          PACKAGES="[\"anyio\", \"http://localhost:3000/dist/$whl_file\"]" \
           node scripts/test-pyodide.js

.github/workflows/verify-jupyter.yml

@@ -3,7 +3,7 @@ name: build
 
 on:
   pull_request:
-    branches: [master]
+    branches: [master, pysdk-release-v8]
 
 jobs:
   build_and_test_streamlit_pyodide:
@@ -29,5 +29,5 @@ jobs:
           whl_file=$(find dist -name "*.whl" | sed 's|^dist/||') # Find the built wheel file, remove dist/ prefix
           echo "Found built wheel file: $whl_file"
           SDK_FILE_PATH=$whl_file \
-          PACKAGES="[\"pyodide-http\", \"http://localhost:3000/dist/$whl_file\"]" \
+          PACKAGES="[\"ssl\", \"anyio\", \"http://localhost:3000/dist/$whl_file\"]" \
           node scripts/test-pyodide.js

.github/workflows/verify-streamlit.yml

@@ -13,6 +13,7 @@ repos:
           - --fix
           - --exit-non-zero-on-fix
           - --force-exclude
+
       - id: ruff-format
         name: ruff format
         language: python
@@ -32,22 +33,11 @@ repos:
     hooks:
       - id: mypy
         name: mypy
-      # Eventually we want to run mypy on the whole test code base. But for now we only enable it for data modeling, simulators and datapoints
-        entry: |
-          poetry run dmypy run -- cognite tests/tests_integration/test_api/test_data_modeling
-          tests/tests_unit/test_data_classes/test_data_models
-          tests/tests_unit/test_utils/test_time.py
-          tests/tests_integration/test_api/test_datapoints.py
-          tests/tests_integration/test_api/test_simulators
-          tests/tests_integration/test_api/test_principals.py
-          tests/tests_unit/test_api/test_simulators
-          tests/tests_unit/test_data_classes/test_simulators.py
+        entry: poetry run dmypy run -- cognite tests
         files: ^.*.(py|pyi)$
         language: system
         pass_filenames: false
 
-  - repo: local
-    hooks:
       - id: custom-checks
         name: custom repo checks
         entry: poetry run python scripts/run_checks.py
@@ -56,6 +46,14 @@ repos:
         pass_filenames: true
         require_serial: true  # avoid possible race conditions
 
+      - id: sync-client-codegen
+        name: run sync codegen on changed files
+        entry: poetry run python scripts/sync_client_codegen/main.py run --files
+        files: ^cognite/.*.(py|pyi)$
+        language: system
+        pass_filenames: true
+        require_serial: true
+
   - repo: https://github.com/jsh9/pydoclint  # Run after 'custom-checks' as these may auto-fix
     rev: 0.8.3
     hooks: