Add find_by_name() to QuerysetEndpoint for content item lookup

[jacalata]

Background

PR #1802 introduces ContentItem as a Protocol that marks the primary content types (workbooks, datasources, views, flows). These are also exactly the endpoints that support server-side name filtering via QuerySet.filter(name=...).

QuerysetEndpoint already has .filter(**kwargs) returning a lazy QuerySet. A find_by_name convenience method belongs on QuerysetEndpoint as a thin, consistent wrapper — visible wherever a consumer would naturally look for it.

Proposed API

# On QuerysetEndpoint[T]:
def find_by_name(self, name: str) -> list[T]:
    return list(self.filter(name=name))

Usage:

matches = server.workbooks.find_by_name("Sales Dashboard")
matches = server.datasources.find_by_name("Orders")
matches = server.users.find_by_name("jsmith")
matches = server.projects.find_by_name("Marketing")

Returns a list (may be empty, may have multiple matches for same-name items in different projects). The caller chooses how to handle ambiguity — for example, filtering by owner_id or project on the result. A separate projects.find_by_path("Marketing/Reports/Monthly") would handle the hierarchical case (separate issue).

Why QuerysetEndpoint and not individual endpoints

All endpoints that inherit QuerysetEndpoint already support server-side name filtering. Putting find_by_name on the base class means:

• One implementation, consistent across all content types
• No endpoint inherits it unless it already supports name filtering (non-QuerysetEndpoint endpoints like jobs simply don't have it)
• Pairs naturally with the ContentItem Protocol from Add structural Protocol types for TSC item classes #1802 — find_by_name is meaningful precisely for the types that satisfy ContentItem

Context

tabcmd (tabcmd/commands/server.py) has a hand-rolled get_items_by_name() that does manual pagination + RequestOptions name filter. This is the pattern every TSC client ends up writing from scratch. Moving it to QuerysetEndpoint lets tabcmd and other clients delete their own copies.

Related: #1809 (CSV user import cleanup), #1802 (ContentItem Protocol).

[jorwoods]

Cool. I think I would propose that it still return a QuerySet instead of a list in case other filter or sort operations are desired to be chained on.