refactor: remove dependency on wtforms from the job endpoint

examples/scripts/client.py

@@ -170,18 +170,18 @@ def submit_job(
         queue: str = "tensorflow_cpu",
         timeout: str = "24h",
     ) -> dict[str, Any]:
-        job_form = {
-            "experiment_name": experiment_name,
+        job_form: dict[str, Any] = {
+            "experimentName": experiment_name,
             "queue": queue,
             "timeout": timeout,
-            "entry_point": entry_point,
+            "entryPoint": entry_point,
         }
 
         if entry_point_kwargs is not None:
-            job_form["entry_point_kwargs"] = entry_point_kwargs
+            job_form["entryPointKwargs"] = entry_point_kwargs
 
         if depends_on is not None:
-            job_form["depends_on"] = depends_on
+            job_form["dependsOn"] = depends_on
 
         workflows_file = Path(workflows_file)
 

src/dioptra/client/_client.py

@@ -667,18 +667,18 @@ def submit_job(
             See https://pages.nist.gov/dioptra/user-guide/api-reference-restapi.html
             for more information on Dioptra's REST api.
         """
-        job_form = {
-            "experiment_name": experiment_name,
+        job_form: dict[str, Any] = {
+            "experimentName": experiment_name,
             "queue": queue,
             "timeout": timeout,
-            "entry_point": entry_point,
+            "entryPoint": entry_point,
         }
 
         if entry_point_kwargs is not None:
-            job_form["entry_point_kwargs"] = entry_point_kwargs
+            job_form["entryPointKwargs"] = entry_point_kwargs
 
         if depends_on is not None:
-            job_form["depends_on"] = depends_on
+            job_form["dependsOn"] = depends_on
 
         workflows_file = Path(workflows_file)
 

src/dioptra/restapi/job/controller.py

@@ -22,17 +22,18 @@
 
 import flask
 import structlog
+from flask import request
 from flask_accepts import accepts, responds
 from flask_login import login_required
 from flask_restx import Namespace, Resource
 from injector import inject
 from structlog.stdlib import BoundLogger
 
-from dioptra.restapi.utils import as_api_parser
+from dioptra.restapi.utils import as_api_parser, as_parameters_schema_list
 
-from .errors import JobDoesNotExistError, JobSubmissionError
-from .model import Job, JobForm, JobFormData
-from .schema import JobSchema, TaskEngineSubmission, job_submit_form_schema
+from .errors import JobDoesNotExistError
+from .model import Job
+from .schema import JobSchema, TaskEngineSubmission
 from .service import JobService
 
 LOGGER: BoundLogger = structlog.stdlib.get_logger()
@@ -68,28 +69,31 @@ def get(self) -> List[Job]:
         return self._job_service.get_all(log=log)
 
     @login_required
-    @api.expect(as_api_parser(api, job_submit_form_schema))
-    @accepts(job_submit_form_schema, api=api)
+    @api.expect(
+        as_api_parser(
+            api,
+            as_parameters_schema_list(JobSchema, operation="load", location="form"),
+        )
+    )
+    @accepts(form_schema=JobSchema, api=api)
     @responds(schema=JobSchema, api=api)
     def post(self) -> Job:
         """Creates a new job via a job submission form with an attached file."""
         log: BoundLogger = LOGGER.new(
             request_id=str(uuid.uuid4()), resource="job", request_type="POST"
         )  # noqa: F841
-        job_form: JobForm = JobForm()
-
+        parsed_obj = request.parsed_form  # type: ignore
         log.info("Request received")
-
-        if not job_form.validate_on_submit():
-            log.error("Form validation failed")
-            raise JobSubmissionError
-
-        log.info("Form validation successful")
-        job_form_data: JobFormData = self._job_service.extract_data_from_form(
-            job_form=job_form,
+        return self._job_service.submit(
+            queue_name=parsed_obj["queue"],
+            experiment_name=parsed_obj["experiment_name"],
+            timeout=parsed_obj["timeout"],
+            entry_point=parsed_obj["entry_point"],
+            entry_point_kwargs=parsed_obj["entry_point_kwargs"],
+            depends_on=parsed_obj["depends_on"],
+            workflow=request.files["workflow"],
             log=log,
         )
-        return self._job_service.submit(job_form_data=job_form_data, log=log)
 
 
 @api.route("/<string:jobId>")

src/dioptra/restapi/job/dependencies.py

@@ -29,14 +29,6 @@
 from dioptra.restapi.shared.request_scope import request
 from dioptra.restapi.shared.rq.service import RQService
 
-from .schema import JobFormSchema
-
-
-class JobFormSchemaModule(Module):
-    @provider
-    def provide_job_form_schema_module(self) -> JobFormSchema:
-        return JobFormSchema()
-
 
 @dataclass
 class RQServiceConfiguration(object):
@@ -97,5 +89,4 @@ def register_providers(modules: List[Callable[..., Any]]) -> None:
         modules: A list of callables used for configuring the dependency injection
             environment.
     """
-    modules.append(JobFormSchemaModule)
     modules.append(RQServiceModule)

src/dioptra/restapi/job/model.py

@@ -18,21 +18,9 @@
 from __future__ import annotations
 
 import datetime
-from typing import Optional
-
-from flask_wtf import FlaskForm
-from flask_wtf.file import FileAllowed, FileField, FileRequired
-from typing_extensions import TypedDict
-from werkzeug.datastructures import FileStorage
-from wtforms.fields import StringField
-from wtforms.validators import UUID, InputRequired
-from wtforms.validators import Optional as OptionalField
-from wtforms.validators import Regexp, ValidationError
+from typing import Any
 
 from dioptra.restapi.app import db
-from dioptra.restapi.utils import slugify
-
-from .interface import JobUpdateInterface
 
 job_statuses = db.Table(
     "job_statuses", db.Column("status", db.String(255), primary_key=True)
@@ -89,7 +77,7 @@ class Job(db.Model):
     experiment = db.relationship("Experiment", back_populates="jobs")
     queue = db.relationship("Queue", back_populates="jobs")
 
-    def update(self, changes: JobUpdateInterface):
+    def update(self, changes: dict[str, Any]):
         """Updates the record.
 
         Args:
@@ -102,141 +90,3 @@ def update(self, changes: JobUpdateInterface):
             setattr(self, key, val)
 
         return self
-
-
-class JobForm(FlaskForm):
-    """The job submission form.
-
-    Attributes:
-        experiment_name: The name of a registered experiment.
-        queue: The name of an active queue.
-        timeout: The maximum alloted time for a job before it times out and is stopped.
-            If omitted, the job timeout will default to 24 hours.
-        entry_point: The name of the entry point in the MLproject file to run.
-        entry_point_kwargs: A list of entry point parameter values to use for the job.
-            The list is a string with the following format: `-P param1=value1
-            -P param2=value2`. If omitted, the default values in the MLproject file will
-            be used.
-        depends_on: A job UUID to set as a dependency for this new job. The new job will
-            not run until this job completes successfully. If omitted, then the new job
-            will start as soon as computing resources are available.
-        workflow: A tarball archive or zip file containing, at a minimum, a MLproject
-            file and its associated entry point scripts.
-    """
-
-    experiment_name = StringField(
-        "Name of Experiment",
-        validators=[InputRequired()],
-        description="The name of a registered experiment.",
-    )
-    queue = StringField(
-        "Queue", validators=[InputRequired()], description="The name of an active queue"
-    )
-    timeout = StringField(
-        "Job Timeout",
-        validators=[OptionalField(), Regexp(r"\d+?[dhms]")],
-        description="The maximum alloted time for a job before it times out and "
-        "is stopped. If omitted, the job timeout will default to 24 hours.",
-    )
-    entry_point = StringField(
-        "MLproject Entry Point",
-        validators=[InputRequired()],
-        description="The name of the entry point in the MLproject file to run.",
-    )
-    entry_point_kwargs = StringField(
-        "MLproject Parameter Overrides",
-        validators=[OptionalField()],
-        description="A list of entry point parameter values to use for the job. The "
-        'list is a string with the following format: "-P param1=value1 '
-        '-P param2=value2". If omitted, the default values in the MLproject file will '
-        "be used.",
-    )
-    depends_on = StringField(
-        "Job Dependency",
-        validators=[OptionalField(), UUID()],
-        description="A job UUID to set as a dependency for this new job. The new job "
-        "will not run until this job completes successfully. If omitted, then the new "
-        "job will start as soon as computing resources are available.",
-    )
-    workflow = FileField(
-        validators=[
-            FileRequired(),
-            FileAllowed(["tar", "tgz", "bz2", "gz", "xz", "zip"]),
-        ],
-        description="A tarball archive or zip file containing, at a minimum, a "
-        "MLproject file and its associated entry point scripts.",
-    )
-
-    def validate_experiment_name(self, field: StringField) -> None:
-        """Validates that the experiment is registered and not deleted.
-
-        Args:
-            field: The form field for `experiment_name`.
-        """
-        from dioptra.restapi.models import Experiment
-
-        standardized_name: str = slugify(field.data)
-
-        if (
-            Experiment.query.filter_by(name=standardized_name, is_deleted=False).first()
-            is None
-        ):
-            raise ValidationError(
-                f"Bad Request - The experiment {standardized_name} does not exist. "
-                "Please check spelling and resubmit."
-            )
-
-    def validate_queue(self, field: StringField) -> None:
-        """Validates that the queue is registered, active and not deleted.
-
-        Args:
-            field: The form field for `queue`.
-        """
-        from dioptra.restapi.models import Queue, QueueLock
-
-        standardized_name: str = slugify(field.data)
-        queue: Optional[Queue] = (
-            Queue.query.outerjoin(QueueLock, Queue.queue_id == QueueLock.queue_id)
-            .filter(
-                Queue.name == standardized_name,
-                QueueLock.queue_id == None,  # noqa: E711
-                Queue.is_deleted == False,  # noqa: E712
-            )
-            .first()
-        )
-
-        if queue is None:
-            raise ValidationError(
-                f"Bad Request - The queue {standardized_name} is not valid. "
-                "Please check spelling and resubmit."
-            )
-
-
-class JobFormData(TypedDict, total=False):
-    """The data extracted from the job submission form.
-
-    Attributes:
-        experiment_id: An integer identifying the registered experiment.
-        experiment_name: The name of the registered experiment.
-        queue_id: An integer identifying a registered queue.
-        queue: The name of an active queue.
-        timeout: The maximum alloted time for a job before it times out and is stopped.
-        entry_point: The name of the entry point in the MLproject file to run.
-        entry_point_kwargs: A list of entry point parameter values to use for the job.
-            The list is a string with the following format: `-P param1=value1
-            -P param2=value2`.
-        depends_on: A job UUID to set as a dependency for this new job. The new job will
-            not run until this job completes successfully.
-        workflow: A tarball archive or zip file containing, at a minimum, a MLproject
-            file and its associated entry point scripts.
-    """
-
-    experiment_id: int
-    experiment_name: str
-    queue_id: int
-    queue: str
-    timeout: Optional[str]
-    entry_point: str
-    entry_point_kwargs: Optional[str]
-    depends_on: Optional[str]
-    workflow: FileStorage