Fix Microsoft Fabric compatibility in core artifact models

[pradeepsrikakolapu]

Overview

Extend the package's existing SQL Server compatibility branches to Microsoft Fabric
Warehouse. This fixes package-level SQL and configuration that cannot be overridden
through adapter dispatch alone.

Update type - breaking / non-breaking

• Minor bug fix
• Documentation improvements
• Quality of Life improvements
• New features (non-breaking change)
• New features (breaking change)
• Other (non-breaking change)
• Other (breaking change)
• Release preparation

What does this solve?

Fabric uses target.type == "fabric", so SQL Server-only conditions currently
exclude it even where it needs the same SQL:

• Quote reserved column names such as "database" and "schema" in upload column
  lists and source, staging, dimension, and fact models.
• Use explicit bit predicates in dim_dbt__current_models full-refresh calculations.
• Disable relation and column documentation persistence for Fabric, where those
  operations are not supported by the adapter.

The changes add fabric to existing SQL Server branches. Other adapter paths are
unchanged, no dataset columns or column ordering change, and Snowflake-only
analytics models remain disabled.

Outstanding questions

This is the package-side companion to Fabric-specific upload macros being prepared
in microsoft/dbt-fabric. Those adapter
changes are not yet published. This PR alone does not provide standalone, complete
Fabric support with an older adapter.

The consuming project needs dispatch configured as follows, in addition to the
package's normal upload hook:

dispatch:
  - macro_namespace: dbt_artifacts
    search_order: ["dbt", "dbt_artifacts"]

No automated Fabric CI target is introduced here. The live regression currently
lives in the companion adapter worktree; upstream SQLFluff lint and other
warehouse integration suites were not run locally.

What databases have you tested with?

• Snowflake
• Google BigQuery
• Databricks
• Spark
• Microsoft Fabric Warehouse
• N/A

Live validation used dbt Core 1.12.5, dbt-adapters 1.24.5, and the in-progress
dbt-fabric 1.12.0 adapter changes. The scenario passed against dbt_artifacts
2.11.1 with this exact patch applied:

1. Initialize all 34 enabled core package models.
2. Execute seed, model, snapshot, and model data-test commands with upload hooks.
3. Rerun the core package models and verify all 11 metadata/execution datasets.
4. Check affected-row counts, apostrophe-containing JSON metadata, and test results.

The patch also applies cleanly to current upstream main at 16662c0.