extend compatibility to pydantic v1 and v2

Author: francesco086

CHANGELOG.md

@@ -10,6 +10,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 Anything MAY change at any time. The public API SHOULD NOT be considered stable.").
 While in this phase, we will denote breaking changes with a minor increase.
 
+## 0.3.2
+
+### Changed
+
+* Extend compatibility to pydantic v1. Now `dac` works with both v1 and v2 of pydantic
+
 ## 0.3.1
 
 ### Changed

pyproject.toml

@@ -23,7 +23,7 @@ classifiers = [
 
 dependencies = [
   "build~=0.9",
-  "pydantic~=2.1",
+  "pydantic>=1.10,<3.0",
   "toml~=0.10",
   "typer[all]~=0.7",
   "wheel~=0.38"

src/dac/_input/_config_pydantic_v1.py

@@ -0,0 +1,110 @@
+import importlib
+import inspect
+import sys
+from pathlib import Path
+from typing import Dict, Optional
+
+import pandera as pa
+from pydantic import BaseModel, root_validator, validator
+
+from dac._file_helper import temporarily_copied_file
+from dac._input.pyproject import PyProjectConfig
+
+
+class PackConfig(BaseModel):
+    data_path: Optional[Path] = None
+    load_path: Path
+    schema_path: Path
+    wheel_dir: Path
+    pyproject: PyProjectConfig
+
+    @validator("data_path", "load_path", "schema_path", "wheel_dir")
+    def path_exists(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        if path is not None and not path.exists():
+            raise ValueError((f"Path {path.as_posix()} is not valid"))
+        return path
+
+    @validator("load_path")
+    def load_contains_expected_function(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        try:
+            sys.path.append(path.parent.as_posix())
+            pkg = importlib.import_module(name=path.stem)
+        except Exception as e:
+            raise ValueError(
+                (
+                    f"{path.as_posix()} is not a path to a python module that can be imported, "
+                    "because of the following error:"
+                    "\n"
+                    f"{e}"
+                )
+            ) from e
+
+        try:
+            signature = inspect.getfullargspec(pkg.load)
+            assert signature.args == []
+        except Exception as e:
+            raise ValueError((f"{path.as_posix()} does not contain the required `def load()`")) from e
+        return path
+
+    @validator("schema_path")
+    def schema_contains_expected_class(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        try:
+            sys.path.append(path.parent.as_posix())
+            pkg = importlib.import_module(name=path.stem)
+        except Exception as e:
+            raise ValueError(
+                (
+                    f"{path.as_posix()} is not a path to a python module that can be imported, "
+                    "because of the following error:"
+                    "\n"
+                    f"{e}"
+                )
+            ) from e
+
+        try:
+            issubclass(pkg.Schema, pa.SchemaModel)
+        except Exception as e:
+            raise ValueError((f"{path.as_posix()} does not contain the required `class Schema(pa.SchemaModel)`")) from e
+        return path
+
+    @root_validator
+    def schema_match_data(  # pylint: disable=no-self-argument,no-self-use
+        cls, values: Dict[str, Path]
+    ) -> Dict[str, Path]:
+        try:
+            sys.path.append(values["load_path"].parent.as_posix())
+            load_module = importlib.import_module(name=values["load_path"].stem)
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed because the load module could not be imported"
+            ) from e
+
+        try:
+            if values.get("data_path", None) is not None:
+                with temporarily_copied_file(
+                    src=values["data_path"], dst=values["load_path"].parent / values["data_path"].name
+                ):
+                    data = load_module.load()
+            else:
+                data = load_module.load()
+        except Exception as e:
+            raise ValueError("`load()` failed due to the following error:" "\n" f"{e}") from e
+
+        try:
+            sys.path.append(values["schema_path"].parent.as_posix())
+            schema_module = importlib.import_module(name=values["schema_path"].stem)
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed because the schema module could not be imported"
+            ) from e
+
+        try:
+            schema_module.Schema.validate(data, lazy=True)
+        except pa.errors.SchemaErrors as e:
+            raise ValueError("Validation of the schema against the data has failed:" "\n" f"{e.failure_cases}") from e
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed for unexpected reasons:" "\n" f"{e}"
+            ) from e
+
+        return values

src/dac/_input/_config_pydantic_v2.py

@@ -0,0 +1,109 @@
+import importlib
+import inspect
+import sys
+from pathlib import Path
+from typing import Optional
+
+import pandera as pa
+from pydantic import BaseModel, field_validator, model_validator
+
+from dac._file_helper import temporarily_copied_file
+from dac._input.pyproject import PyProjectConfig
+
+
+class PackConfig(BaseModel):
+    data_path: Optional[Path] = None
+    load_path: Path
+    schema_path: Path
+    wheel_dir: Path
+    pyproject: PyProjectConfig
+
+    @field_validator("data_path", "load_path", "schema_path", "wheel_dir")
+    @classmethod
+    def path_exists(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        if path is not None and not path.exists():
+            raise ValueError((f"Path {path.as_posix()} is not valid"))
+        return path
+
+    @field_validator("load_path")
+    @classmethod
+    def load_contains_expected_function(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        try:
+            sys.path.append(path.parent.as_posix())
+            pkg = importlib.import_module(name=path.stem)
+        except Exception as e:
+            raise ValueError(
+                (
+                    f"{path.as_posix()} is not a path to a python module that can be imported, "
+                    "because of the following error:"
+                    "\n"
+                    f"{e}"
+                )
+            ) from e
+
+        try:
+            signature = inspect.getfullargspec(pkg.load)
+            assert signature.args == []
+        except Exception as e:
+            raise ValueError((f"{path.as_posix()} does not contain the required `def load()`")) from e
+        return path
+
+    @field_validator("schema_path")
+    @classmethod
+    def schema_contains_expected_class(cls, path: Path) -> Path:  # pylint: disable=no-self-argument,no-self-use
+        try:
+            sys.path.append(path.parent.as_posix())
+            pkg = importlib.import_module(name=path.stem)
+        except Exception as e:
+            raise ValueError(
+                (
+                    f"{path.as_posix()} is not a path to a python module that can be imported, "
+                    "because of the following error:"
+                    "\n"
+                    f"{e}"
+                )
+            ) from e
+
+        try:
+            issubclass(pkg.Schema, pa.SchemaModel)
+        except Exception as e:
+            raise ValueError((f"{path.as_posix()} does not contain the required `class Schema(pa.SchemaModel)`")) from e
+        return path
+
+    @model_validator(mode="after")
+    def schema_match_data(self) -> "PackConfig":
+        try:
+            sys.path.append(self.load_path.parent.as_posix())
+            load_module = importlib.import_module(name=self.load_path.stem)
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed because the load module could not be imported"
+            ) from e
+
+        try:
+            if self.data_path is not None:
+                with temporarily_copied_file(src=self.data_path, dst=self.load_path.parent / self.data_path.name):
+                    data = load_module.load()
+            else:
+                data = load_module.load()
+        except Exception as e:
+            raise ValueError("`load()` failed due to the following error:" "\n" f"{e}") from e
+
+        try:
+            sys.path.append(self.schema_path.parent.as_posix())
+            schema_module = importlib.import_module(name=self.schema_path.stem)
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed because the schema module could not be imported"
+            ) from e
+
+        try:
+            schema_module.Schema.validate(data, lazy=True)
+        except pa.errors.SchemaErrors as e:
+            raise ValueError("Validation of the schema against the data has failed:" "\n" f"{e.failure_cases}") from e
+        except Exception as e:
+            raise ValueError(
+                "Validation of the schema against the data has failed for unexpected reasons:" "\n" f"{e}"
+            ) from e
+
+        return self

src/dac/_input/_pyproject_pydantic_v1.py

@@ -0,0 +1,32 @@
+from typing import List
+
+import toml  # type: ignore
+from pydantic import BaseModel, validator
+
+
+class PyProjectConfig(BaseModel):
+    project_name: str
+    project_version: str
+    project_dependencies: str
+
+    @validator("project_name")
+    def valid_project_name(cls, name: str) -> str:  # pylint: disable=no-self-argument,no-self-use
+        if not name.isidentifier() or "\xb7" in name:
+            raise ValueError(f"Invalid project name: {name} (hint: only '_' are allowed, no '-')")
+        return name
+
+    def generate_pyproject_toml(self) -> str:
+        return toml.dumps(
+            {
+                "project": {
+                    "name": self.project_name,
+                    "version": self.project_version,
+                    "dependencies": self._get_list_of_project_dependencies(),
+                }
+            }
+        )
+
+    def _get_list_of_project_dependencies(self) -> List[str]:
+        splitted_by_newline = self.project_dependencies.splitlines()
+        splitted_by_newline_or_comma = [s for ss in splitted_by_newline for s in ss.split(";")]
+        return sorted(map(lambda x: x.strip(), splitted_by_newline_or_comma))

src/dac/_input/_pyproject_pydantic_v2.py

@@ -0,0 +1,33 @@
+from typing import List
+
+import toml  # type: ignore
+from pydantic import BaseModel, field_validator
+
+
+class PyProjectConfig(BaseModel):
+    project_name: str
+    project_version: str
+    project_dependencies: str
+
+    @field_validator("project_name")
+    @classmethod
+    def valid_project_name(cls, name: str) -> str:  # pylint: disable=no-self-argument,no-self-use
+        if not name.isidentifier() or "\xb7" in name:
+            raise ValueError(f"Invalid project name: {name} (hint: only '_' are allowed, no '-')")
+        return name
+
+    def generate_pyproject_toml(self) -> str:
+        return toml.dumps(
+            {
+                "project": {
+                    "name": self.project_name,
+                    "version": self.project_version,
+                    "dependencies": self._get_list_of_project_dependencies(),
+                }
+            }
+        )
+
+    def _get_list_of_project_dependencies(self) -> List[str]:
+        splitted_by_newline = self.project_dependencies.splitlines()
+        splitted_by_newline_or_comma = [s for ss in splitted_by_newline for s in ss.split(";")]
+        return sorted(map(lambda x: x.strip(), splitted_by_newline_or_comma))