Provision a project workspace in one idempotent pass.

A cross-domain "new project" setup that stitches together a document library with typed metadata columns, a task list with a custom view, and a few starter rows. It is safe to run repeatedly: lists and fields are ensured (get-or-create), views are created only when missing, and starter rows are seeded only while the task list is empty.

python provision_project_workspace.py
python provision_project_workspace.py --doc-library "Project Files" --task-list "Project Tasks"

Permissions

Requires Sites.ReadWrite.All or an owner-level account on the target site.

Reference

• https://learn.microsoft.com/en-us/sharepoint/dev/apis/rest-api

View source

from __future__ import annotations

import argparse

from office365.sharepoint.client_context import ClientContext
from office365.sharepoint.fields.creation_information import FieldCreationInformation
from office365.sharepoint.fields.type import FieldType
from office365.sharepoint.lists.templates.type import ListTemplateType
from tests.settings import cert_path, cert_thumbprint, client_id, site_url, tenant

#: Document library columns — declared as rich field specs.
DOC_LIBRARY_FIELDS = [
    FieldCreationInformation(
        Title="ProjectStage",
        FieldTypeKind=FieldType.Choice,
        Choices=["Discovery", "In Progress", "Review", "Done"],
    ),
    FieldCreationInformation(Title="Budget", FieldTypeKind=FieldType.Number),
    FieldCreationInformation(Title="ClientDueDate", FieldTypeKind=FieldType.DateTime),
]

#: Task list columns.
TASK_FIELDS = [
    FieldCreationInformation(
        Title="Status",
        FieldTypeKind=FieldType.Choice,
        Choices=["Not Started", "In Progress", "Blocked", "Done"],
    ),
    FieldCreationInformation(Title="Owner", FieldTypeKind=FieldType.Text),
    FieldCreationInformation(Title="DueDate", FieldTypeKind=FieldType.DateTime),
]

SEED_TASKS = [
    {"Title": "Kickoff and scope", "Status": "Done", "Owner": "PM"},
    {"Title": "Collect requirements", "Status": "In Progress", "Owner": "BA"},
    {"Title": "Draft solution design", "Status": "Not Started", "Owner": "Architect"},
]


def main() -> None:
    parser = argparse.ArgumentParser(description="Provision a project workspace (library, task list, views, seed rows)")
    parser.add_argument("--doc-library", default="Project Files", help="document library title")
    parser.add_argument("--task-list", default="Project Tasks", help="task list title")
    args = parser.parse_args()

    ctx = ClientContext(site_url).with_client_certificate(
        tenant, client_id=client_id, thumbprint=cert_thumbprint, cert_path=cert_path
    )
    web = ctx.web.get().execute_query()
    print(f"Site: {web.title} ({web.url})\n")

    # 1. Document library with typed metadata columns and a by-stage view.
    library = ctx.web.lists.ensure_list(
        args.doc_library,
        description="Project deliverables and working documents",
        template_type=ListTemplateType.DocumentLibrary,
        on_conflict="update",
    ).execute_query()
    doc_fields = library.ensure_fields(DOC_LIBRARY_FIELDS)
    ctx.execute_query()
    library.views.ensure_view("By stage", fields=["Title", "ProjectStage", "ClientDueDate", "Budget"]).execute_query()
    doc_columns = ", ".join(f.internal_name or "?" for f in doc_fields)
    print(f"Library  : {library.title}  (columns: {doc_columns}, view: By stage)")

    # 2. Task list with typed columns and a view.
    tasks = ctx.web.lists.ensure_list(
        args.task_list,
        description="Track project tasks",
        template_type=ListTemplateType.GenericList,
        on_conflict="update",
    ).execute_query()
    task_fields = tasks.ensure_fields(TASK_FIELDS)
    ctx.execute_query()
    tasks.views.ensure_view("Open tasks", fields=["Title", "Status", "Owner", "DueDate"]).execute_query()
    task_columns = ", ".join(f.internal_name or "?" for f in task_fields)
    print(f"Task list: {tasks.title}  (columns: {task_columns}, view: Open tasks)")

    # 3. Seed starter rows once, so re-runs stay idempotent.
    if not tasks.items.get().execute_query():
        for row in SEED_TASKS:
            tasks.add_item(row)
        ctx.execute_query()
        print(f"Seeded   : {len(SEED_TASKS)} starter task(s)")
    else:
        print("Seeded   : task list already has rows (nothing to do)")

    print(f"\nWorkspace ready at {web.url}/{args.doc_library}")


if __name__ == "__main__":
    main()

---

← Back to Recipes