etlplus 0.12.12__py3-none-any.whl → 0.14.3__py3-none-any.whl

etlplus/dag.py

@@ -0,0 +1,103 @@
+"""
+:mod:`etlplus.dag` module.
+
+Lightweight directed acyclic graph (DAG) helpers for ordering jobs based on
+``depends_on``.
+"""
+
+from __future__ import annotations
+
+from collections import deque
+from dataclasses import dataclass
+
+from .config.jobs import JobConfig
+
+# SECTION: EXPORTS ========================================================== #
+
+
+__all__ = [
+    'DagError',
+    'topological_sort_jobs',
+]
+
+
+# SECTION: ERRORS =========================================================== #
+
+
+@dataclass(slots=True)
+class DagError(ValueError):
+    """
+    Raised when the job dependency graph is invalid.
+
+    Attributes
+    ----------
+    message : str
+        Error message.
+    """
+
+    # -- Attributes -- #
+
+    message: str
+
+    # -- Magic Methods (Object Representation) -- #
+
+    def __str__(self) -> str:
+        return self.message
+
+
+# SECTION: FUNCTIONS ======================================================== #
+
+
+def topological_sort_jobs(
+    jobs: list[JobConfig],
+) -> list[JobConfig]:
+    """
+    Return jobs in topological order based on ``depends_on``.
+
+    Parameters
+    ----------
+    jobs : list[JobConfig]
+        List of job configurations to sort.
+
+    Returns
+    -------
+    list[JobConfig]
+        Jobs sorted in topological order.
+
+    Raises
+    ------
+    DagError
+        If a dependency is missing, self-referential, or when a cycle is
+        detected.
+    """
+    index = {job.name: job for job in jobs}
+    edges: dict[str, set[str]] = {name: set() for name in index}
+    indegree: dict[str, int] = {name: 0 for name in index}
+
+    for job in jobs:
+        for dep in job.depends_on:
+            if dep not in index:
+                raise DagError(
+                    f'Unknown dependency "{dep}" in job "{job.name}"',
+                )
+            if dep == job.name:
+                raise DagError(f'Job "{job.name}" depends on itself')
+            if job.name not in edges[dep]:
+                edges[dep].add(job.name)
+                indegree[job.name] += 1
+
+    queue = deque(sorted(name for name, deg in indegree.items() if deg == 0))
+    ordered: list[str] = []
+
+    while queue:
+        name = queue.popleft()
+        ordered.append(name)
+        for child in sorted(edges[name]):
+            indegree[child] -= 1
+            if indegree[child] == 0:
+                queue.append(child)
+
+    if len(ordered) != len(jobs):
+        raise DagError('Dependency cycle detected')
+
+    return [index[name] for name in ordered]

etlplus/enums.py

@@ -23,7 +23,6 @@ __all__ = [
     'AggregateName',
     'CoercibleStrEnum',
     'DataConnectorType',
-    'HttpMethod',
     'OperatorName',
     'PipelineStep',
 ]
@@ -200,37 +199,6 @@ class DataConnectorType(CoercibleStrEnum):
         }
 
 
-class HttpMethod(CoercibleStrEnum):
-    """Supported HTTP verbs that accept JSON payloads."""
-
-    # -- Constants -- #
-
-    CONNECT = 'connect'
-    DELETE = 'delete'
-    GET = 'get'
-    HEAD = 'head'
-    OPTIONS = 'options'
-    PATCH = 'patch'
-    POST = 'post'
-    PUT = 'put'
-    TRACE = 'trace'
-
-    # -- Getters -- #
-
-    @property
-    def allows_body(self) -> bool:
-        """
-        Whether the method typically allows a request body.
-
-        Notes
-        -----
-        - RFCs do not strictly forbid bodies on some other methods (e.g.,
-            ``DELETE``), but many servers/clients do not expect them. We mark
-            ``POST``, ``PUT``, and ``PATCH`` as True.
-        """
-        return self in {HttpMethod.POST, HttpMethod.PUT, HttpMethod.PATCH}
-
-
 class OperatorName(CoercibleStrEnum):
     """Supported comparison operators with helpers."""
 

etlplus/file/enums.py

@@ -123,7 +123,7 @@ class FileFormat(CoercibleStrEnum):
     RDS = 'rds'  # R data file
     SAS7BDAT = 'sas7bdat'  # SAS data file
     SAV = 'sav'  # SPSS data file
-    SYLK = 'sylk'  # Symbolic Link (SYmbolic LinK)
+    SYLK = 'sylk'  # Symbolic Link
     XPT = 'xpt'  # SAS Transport file
     ZSAV = 'zsav'  # Compressed SPSS data file
 

etlplus/{validation → ops}/README.md

@@ -1,4 +1,4 @@
-# etlplus.validation subpackage
+# etlplus.ops subpackage
 
 Documentation for the `etlplus.validation` subpackage: data validation utilities and helpers.
 
@@ -8,7 +8,7 @@ Documentation for the `etlplus.validation` subpackage: data validation utilities
 
 Back to project overview: see the top-level [README](../../README.md).
 
-- [etlplus.validation subpackage](#etlplusvalidation-subpackage)
+- [etlplus.ops subpackage](#etlplusops-subpackage)
   - [Validation Features](#validation-features)
   - [Defining Validation Rules](#defining-validation-rules)
   - [Example: Validating Data](#example-validating-data)

etlplus/ops/__init__.py

@@ -0,0 +1,61 @@
+"""
+:mod:`etlplus.ops` package.
+
+Data operations helpers.
+
+Importing :mod:`etlplus.ops` exposes the coarse-grained helpers most users care
+about: ``extract``, ``transform``, ``load``, ``validate``, ``run``, and
+``run_pipeline``. Each helper delegates to the richer modules under
+``etlplus.ops.*`` while presenting a compact public API surface. Conditional
+validation orchestration is available via
+:func:`etlplus.ops.utils.maybe_validate`. The legacy compatibility module
+:mod:`etlplus.ops.__init__validation` is deprecated in favor of this package.
+
+Examples
+--------
+>>> from etlplus.ops import extract, transform
+>>> raw = extract('file', 'input.json')
+>>> curated = transform(raw, {'select': ['id', 'name']})
+
+>>> from etlplus.ops.utils import maybe_validate
+>>> payload = {'name': 'Alice'}
+>>> rules = {'required': ['name']}
+>>> def validator(data, config):
+...     missing = [field for field in config['required'] if field not in data]
+...     return {'valid': not missing, 'errors': missing, 'data': data}
+>>> maybe_validate(
+...     payload,
+...     when='both',
+...     enabled=True,
+...     rules=rules,
+...     phase='before_transform',
+...     severity='warn',
+...     validate_fn=validator,
+...     print_json_fn=lambda message: message,
+... )
+{'name': 'Alice'}
+
+See Also
+--------
+:mod:`etlplus.ops.run`
+:mod:`etlplus.ops.utils`
+"""
+
+from .extract import extract
+from .load import load
+from .run import run
+from .run import run_pipeline
+from .transform import transform
+from .validate import validate
+
+# SECTION: EXPORTS ========================================================== #
+
+
+__all__ = [
+    'extract',
+    'load',
+    'run',
+    'run_pipeline',
+    'transform',
+    'validate',
+]

etlplus/{extract.py → ops/extract.py}

@@ -1,5 +1,5 @@
 """
-:mod:`etlplus.extract` module.
+:mod:`etlplus.ops.extract` module.
 
 Helpers to extract data from files, databases, and REST APIs.
 """
@@ -10,56 +10,81 @@ from pathlib import Path
 from typing import Any
 from typing import cast
 
-import requests  # type: ignore[import]
-
-from .enums import DataConnectorType
-from .enums import HttpMethod
-from .file import File
-from .file import FileFormat
-from .types import JSONData
-from .types import JSONDict
-from .types import JSONList
-from .types import StrPath
+from ..api import HttpMethod
+from ..api.utils import resolve_request
+from ..enums import DataConnectorType
+from ..file import File
+from ..file import FileFormat
+from ..types import JSONData
+from ..types import JSONDict
+from ..types import JSONList
+from ..types import StrPath
 
 # SECTION: FUNCTIONS ======================================================== #
 
 
-# -- File Extraction -- #
-
-
-def extract_from_file(
-    file_path: StrPath,
-    file_format: FileFormat | str | None = FileFormat.JSON,
+def extract_from_api(
+    url: str,
+    method: HttpMethod | str = HttpMethod.GET,
+    **kwargs: Any,
 ) -> JSONData:
     """
-    Extract (semi-)structured data from a local file.
+    Extract data from a REST API.
 
     Parameters
     ----------
-    file_path : StrPath
-        Source file path.
-    file_format : FileFormat | str | None, optional
-        File format to parse. If ``None``, infer from the filename
-        extension. Defaults to `'json'` for backward compatibility when
-        explicitly provided.
+    url : str
+        API endpoint URL.
+    method : HttpMethod | str, optional
+        HTTP method to use. Defaults to ``GET``.
+    **kwargs : Any
+        Extra arguments forwarded to the underlying ``requests`` call
+        (for example, ``timeout``). To use a pre-configured
+        :class:`requests.Session`, provide it via ``session``.
+        When omitted, ``timeout`` defaults to 10 seconds.
 
     Returns
     -------
     JSONData
-        Parsed data as a mapping or a list of mappings.
-    """
-    path = Path(file_path)
-
-    # If no explicit format is provided, let File infer from extension.
-    if file_format is None:
-        return File(path, None).read()
-    fmt = FileFormat.coerce(file_format)
+        Parsed JSON payload, or a fallback object with raw text.
 
-    # Let file module perform existence and format validation.
-    return File(path, fmt).read()
+    Raises
+    ------
+    TypeError
+        If a provided ``session`` does not expose the required HTTP
+        method (for example, ``get``).
+    """
+    timeout = kwargs.pop('timeout', None)
+    session = kwargs.pop('session', None)
+    request_callable, timeout, _ = resolve_request(
+        method,
+        session=session,
+        timeout=timeout,
+    )
+    response = request_callable(url, timeout=timeout, **kwargs)
+    response.raise_for_status()
 
+    content_type = response.headers.get('content-type', '').lower()
+    if 'application/json' in content_type:
+        try:
+            payload: Any = response.json()
+        except ValueError:
+            # Malformed JSON despite content-type; fall back to text
+            return {
+                'content': response.text,
+                'content_type': content_type,
+            }
+        if isinstance(payload, dict):
+            return cast(JSONDict, payload)
+        if isinstance(payload, list):
+            if all(isinstance(x, dict) for x in payload):
+                return cast(JSONList, payload)
+            # Coerce non-dict array items into objects for consistency
+            return [{'value': x} for x in payload]
+        # Fallback: wrap scalar JSON
+        return {'value': payload}
 
-# -- Database Extraction (Placeholder) -- #
+    return {'content': response.text, 'content_type': content_type}
 
 
 def extract_from_database(
@@ -94,77 +119,36 @@ def extract_from_database(
     ]
 
 
-# -- REST API Extraction -- #
-
-
-def extract_from_api(
-    url: str,
-    method: HttpMethod | str = HttpMethod.GET,
-    **kwargs: Any,
+def extract_from_file(
+    file_path: StrPath,
+    file_format: FileFormat | str | None = FileFormat.JSON,
 ) -> JSONData:
     """
-    Extract data from a REST API.
+    Extract (semi-)structured data from a local file.
 
     Parameters
     ----------
-    url : str
-        API endpoint URL.
-    method : HttpMethod | str, optional
-        HTTP method to use. Defaults to ``GET``.
-    **kwargs : Any
-        Extra arguments forwarded to the underlying ``requests`` call
-        (for example, ``timeout``). To use a pre-configured
-        :class:`requests.Session`, provide it via ``session``.
+    file_path : StrPath
+        Source file path.
+    file_format : FileFormat | str | None, optional
+        File format to parse. If ``None``, infer from the filename
+        extension. Defaults to `'json'` for backward compatibility when
+        explicitly provided.
 
     Returns
     -------
     JSONData
-        Parsed JSON payload, or a fallback object with raw text.
-
-    Raises
-    ------
-    TypeError
-        If a provided ``session`` does not expose the required HTTP
-        method (for example, ``get``).
+        Parsed data as a mapping or a list of mappings.
     """
-    http_method = HttpMethod.coerce(method)
-
-    # Apply a conservative timeout to guard against hanging requests.
-    timeout = kwargs.pop('timeout', 10.0)
-    session = kwargs.pop('session', None)
-    requester = session or requests
-
-    request_callable = getattr(requester, http_method.value, None)
-    if not callable(request_callable):
-        raise TypeError(
-            'Session object must supply a callable'
-            f'"{http_method.value}" method',
-        )
-
-    response = request_callable(url, timeout=timeout, **kwargs)
-    response.raise_for_status()
+    path = Path(file_path)
 
-    content_type = response.headers.get('content-type', '').lower()
-    if 'application/json' in content_type:
-        try:
-            payload: Any = response.json()
-        except ValueError:
-            # Malformed JSON despite content-type; fall back to text
-            return {
-                'content': response.text,
-                'content_type': content_type,
-            }
-        if isinstance(payload, dict):
-            return cast(JSONDict, payload)
-        if isinstance(payload, list):
-            if all(isinstance(x, dict) for x in payload):
-                return cast(JSONList, payload)
-            # Coerce non-dict array items into objects for consistency
-            return [{'value': x} for x in payload]
-        # Fallback: wrap scalar JSON
-        return {'value': payload}
+    # If no explicit format is provided, let File infer from extension.
+    if file_format is None:
+        return File(path, None).read()
+    fmt = FileFormat.coerce(file_format)
 
-    return {'content': response.text, 'content_type': content_type}
+    # Let file module perform existence and format validation.
+    return File(path, fmt).read()
 
 
 # -- Orchestration -- #

etlplus/{load.py → ops/load.py}

@@ -1,5 +1,5 @@
 """
-:mod:`etlplus.load` module.
+:mod:`etlplus.ops.load` module.
 
 Helpers to load data into files, databases, and REST APIs.
 """
@@ -12,17 +12,16 @@ from pathlib import Path
 from typing import Any
 from typing import cast
 
-import requests  # type: ignore[import]
-
-from .enums import DataConnectorType
-from .enums import HttpMethod
-from .file import File
-from .file import FileFormat
-from .types import JSONData
-from .types import JSONDict
-from .types import JSONList
-from .types import StrPath
-from .utils import count_records
+from ..api import HttpMethod
+from ..api.utils import resolve_request
+from ..enums import DataConnectorType
+from ..file import File
+from ..file import FileFormat
+from ..types import JSONData
+from ..types import JSONDict
+from ..types import JSONList
+from ..types import StrPath
+from ..utils import count_records
 
 # SECTION: INTERNAL FUNCTIONS ============================================== #
 
@@ -69,7 +68,7 @@ def _parse_json_string(
 # SECTION: FUNCTIONS ======================================================== #
 
 
-# -- Data Loading -- #
+# -- Helpers -- #
 
 
 def load_data(
@@ -119,58 +118,59 @@ def load_data(
     )
 
 
-# -- File Loading -- #
-
-
-def load_to_file(
+def load_to_api(
     data: JSONData,
-    file_path: StrPath,
-    file_format: FileFormat | str | None = None,
+    url: str,
+    method: HttpMethod | str,
+    **kwargs: Any,
 ) -> JSONDict:
     """
-    Persist data to a local file.
+    Load data to a REST API.
 
     Parameters
     ----------
     data : JSONData
-        Data to write.
-    file_path : StrPath
-        Target file path.
-    file_format : FileFormat | str | None, optional
-        Output format. If omitted (None), the format is inferred from the
-        filename extension.
+        Data to send as JSON.
+    url : str
+        API endpoint URL.
+    method : HttpMethod | str
+        HTTP method to use.
+    **kwargs : Any
+        Extra arguments forwarded to ``requests`` (e.g., ``timeout``).
+        When omitted, ``timeout`` defaults to 10 seconds.
 
     Returns
     -------
     JSONDict
-        Result dictionary with status and record count.
+        Result dictionary including response payload or text.
     """
-    path = Path(file_path)
-    path.parent.mkdir(parents=True, exist_ok=True)
+    # Apply a conservative timeout to guard against hanging requests.
+    timeout = kwargs.pop('timeout', 10.0)
+    session = kwargs.pop('session', None)
+    request_callable, timeout, http_method = resolve_request(
+        method,
+        session=session,
+        timeout=timeout,
+    )
+    response = request_callable(url, json=data, timeout=timeout, **kwargs)
+    response.raise_for_status()
 
-    # If no explicit format is provided, let File infer from extension.
-    if file_format is None:
-        records = File(path).write(data)
-        ext = path.suffix.lstrip('.').lower()
-        fmt = FileFormat.coerce(ext) if ext else FileFormat.JSON
-    else:
-        fmt = FileFormat.coerce(file_format)
-        records = File(path, fmt).write(data)
-    if fmt is FileFormat.CSV and records == 0:
-        message = 'No data to write'
-    else:
-        message = f'Data loaded to {path}'
+    # Try JSON first, fall back to text.
+    try:
+        payload: Any = response.json()
+    except ValueError:
+        payload = response.text
 
     return {
         'status': 'success',
-        'message': message,
-        'records': records,
+        'status_code': response.status_code,
+        'message': f'Data loaded to {url}',
+        'response': payload,
+        'records': count_records(data),
+        'method': http_method.value.upper(),
     }
 
 
-# -- Database Loading (Placeholder) -- #
-
-
 def load_to_database(
     data: JSONData,
     connection_string: str,
@@ -206,69 +206,49 @@ def load_to_database(
     }
 
 
-# -- REST API Loading -- #
-
-
-def load_to_api(
+def load_to_file(
     data: JSONData,
-    url: str,
-    method: HttpMethod | str,
-    **kwargs: Any,
+    file_path: StrPath,
+    file_format: FileFormat | str | None = None,
 ) -> JSONDict:
     """
-    Load data to a REST API.
+    Persist data to a local file.
 
     Parameters
     ----------
     data : JSONData
-        Data to send as JSON.
-    url : str
-        API endpoint URL.
-    method : HttpMethod | str
-        HTTP method to use.
-    **kwargs : Any
-        Extra arguments forwarded to ``requests`` (e.g., ``timeout``).
+        Data to write.
+    file_path : StrPath
+        Target file path.
+    file_format : FileFormat | str | None, optional
+        Output format. If omitted (None), the format is inferred from the
+        filename extension.
 
     Returns
     -------
     JSONDict
-        Result dictionary including response payload or text.
-
-    Raises
-    ------
-    TypeError
-        If the session object is not valid.
+        Result dictionary with status and record count.
     """
-    http_method = HttpMethod.coerce(method)
-
-    # Apply a conservative timeout to guard against hanging requests.
-    timeout = kwargs.pop('timeout', 10.0)
-    session = kwargs.pop('session', None)
-    requester = session or requests
-
-    request_callable = getattr(requester, http_method.value, None)
-    if not callable(request_callable):
-        raise TypeError(
-            'Session object must supply a '
-            f'callable "{http_method.value}" method',
-        )
-
-    response = request_callable(url, json=data, timeout=timeout, **kwargs)
-    response.raise_for_status()
+    path = Path(file_path)
+    path.parent.mkdir(parents=True, exist_ok=True)
 
-    # Try JSON first, fall back to text.
-    try:
-        payload: Any = response.json()
-    except ValueError:
-        payload = response.text
+    # If no explicit format is provided, let File infer from extension.
+    if file_format is None:
+        records = File(path).write(data)
+        ext = path.suffix.lstrip('.').lower()
+        fmt = FileFormat.coerce(ext) if ext else FileFormat.JSON
+    else:
+        fmt = FileFormat.coerce(file_format)
+        records = File(path, fmt).write(data)
+    if fmt is FileFormat.CSV and records == 0:
+        message = 'No data to write'
+    else:
+        message = f'Data loaded to {path}'
 
     return {
         'status': 'success',
-        'status_code': response.status_code,
-        'message': f'Data loaded to {url}',
-        'response': payload,
-        'records': count_records(data),
-        'method': http_method.value.upper(),
+        'message': message,
+        'records': records,
     }