add pre-commit to verify license headers in python files (#68)

* commit license header

* add pre-commit to verify license headers in python files

* add empty line after license header

.pre-commit-config.yaml

@@ -8,3 +8,32 @@ repos:
       - id: ruff
       # Run the formatter.
       - id: ruff-format
+
+  - repo: https://github.com/johannsdg/pre-commit-license-headers
+    rev: v0.1.0
+    hooks:
+      - id: check-license-headers
+        args:
+          - "--template"
+          - |
+            Copyright (c) [YEARS] Chai Discovery, Inc.
+            This source code is licensed under the Chai Discovery Community License
+            Agreement (LICENSE.md) found in the root directory of this source tree.
+          # - "--owner=The Pre-Commit License Headers Authors"
+        include: .*py
+        exclude: |
+          (?x)^(
+            examples/.*|
+            .pre-commit-config.yaml|
+            .pre-commit-hooks.yaml|
+            .github/workflows/.*yml|
+            .github/dependabot.yml|
+            .github/pull_request_template.md|
+            Dockerfile.chailab|
+            LICENSE.md|
+            README.md|
+            requirements.in|
+            ruff.toml|
+            chai_lab/data/residue_constants.py|
+            pyproject.toml
+          )$

chai_lab/__init__.py

@@ -1 +1,5 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 __version__ = "0.0.1"

chai_lab/chai1.py

@@ -1,4 +1,8 @@
-# %%
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
+
 import math
 from collections import Counter
 from dataclasses import dataclass

chai_lab/data/__init__.py

@@ -0,0 +1,3 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.

chai_lab/data/collate/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/collate/collate.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import dataclasses
 import logging
 from typing import Any

chai_lab/data/collate/utils.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from dataclasses import dataclass
 
 from chai_lab.data.dataset.structure.all_atom_structure_context import (

chai_lab/data/dataset/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/dataset/all_atom_feature_context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import dataclass
 from typing import Any, Final

chai_lab/data/dataset/constraints/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/dataset/constraints/constraint_context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from dataclasses import asdict, dataclass
 from typing import Any
 

chai_lab/data/dataset/embeddings/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/dataset/embeddings/embedding_context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from dataclasses import asdict, dataclass
 
 import torch

chai_lab/data/dataset/embeddings/esm.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import os
 from contextlib import contextmanager
 

chai_lab/data/dataset/inference_dataset.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 import string
 from dataclasses import dataclass

chai_lab/data/dataset/msas/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/dataset/msas/msa_context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from dataclasses import dataclass
 
 import torch

chai_lab/data/dataset/structure/all_atom_residue_tokenizer.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import dataclass
 from itertools import chain

chai_lab/data/dataset/structure/all_atom_structure_context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import asdict, dataclass
 from functools import cached_property, partial

chai_lab/data/dataset/structure/chain.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from dataclasses import dataclass
 
 from chai_lab.data.dataset.structure.all_atom_structure_context import (

chai_lab/data/dataset/structure/utils.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from functools import lru_cache
 
 import torch

chai_lab/data/dataset/templates/context.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import asdict, dataclass
 

chai_lab/data/features/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/features/feature_factory.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 """Helper methods for generating model input features"""
 
 import logging

chai_lab/data/features/feature_type.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from enum import Enum
 
 

chai_lab/data/features/feature_utils.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 """Utility classes and functions for feature representations"""
 
 from chai_lab.utils.typing import typecheck

chai_lab/data/features/generators/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+

chai_lab/data/features/generators/atom_element.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from torch import Tensor
 

chai_lab/data/features/generators/atom_name.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from torch import Tensor
 
 from chai_lab.data.features.feature_type import FeatureType

chai_lab/data/features/generators/base.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 """Feature Generator ABC and Default implementation"""
 
 from abc import ABC

chai_lab/data/features/generators/blocked_atom_pair_distances.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from typing import Any, Literal
 
 import torch

chai_lab/data/features/generators/docking.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 import random
 from dataclasses import dataclass

chai_lab/data/features/generators/esm_generator.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from torch import Tensor
 
 from chai_lab.data.features.feature_type import FeatureType

chai_lab/data/features/generators/identity.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import rearrange
 from torch import Tensor

chai_lab/data/features/generators/is_cropped_chain.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from torch import Tensor
 

chai_lab/data/features/generators/missing_chain_contact.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from torch import Tensor
 

chai_lab/data/features/generators/msa.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from typing import Any
 
 import torch

chai_lab/data/features/generators/ref_pos.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 
 from chai_lab.data.features.feature_type import FeatureType

chai_lab/data/features/generators/relative_chain.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import rearrange
 from torch import Tensor

chai_lab/data/features/generators/relative_entity.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import rearrange
 from torch import Tensor

chai_lab/data/features/generators/relative_sep.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import rearrange
 from torch import Tensor

chai_lab/data/features/generators/relative_token.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import rearrange
 from torch import Tensor

chai_lab/data/features/generators/residue_type.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import numpy as np
 import torch
 from torch import Tensor

chai_lab/data/features/generators/structure_metadata.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import repeat
 from torch import Tensor

chai_lab/data/features/generators/templates.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 """
 Feature generators for templates. This includes the following:
 - Template mask (includes both the psuedo beta mask and backbone frame mask)

chai_lab/data/features/generators/token_dist_restraint.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import dataclass
 

chai_lab/data/features/generators/token_pair_distance.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 from typing import Any
 
 import torch

chai_lab/data/features/generators/token_pair_pocket_restraint.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import logging
 from dataclasses import dataclass
 

chai_lab/data/features/token_utils.py

@@ -1,3 +1,7 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+
 import torch
 from einops import repeat
 from torch import Tensor

chai_lab/data/io/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2024 Chai Discovery, Inc.
+# This source code is licensed under the Chai Discovery Community License
+# Agreement (LICENSE.md) found in the root directory of this source tree.
+