Add 37 integration tests covering:

Manifest Parsing:
- Parse dbt metadata, models, columns, types
- Filter models by tag and name
- Extract model properties and dependencies

Feast Object Mapping:
- Create entities, data sources, feature views from dbt models
- Map dbt types to Feast types (int, float, string, etc.)
- Support BigQuery, Snowflake, and File data sources
- Handle timestamp fields and TTL configuration

Code Generation:
- Generate valid Python code for Feast objects
- Include proper imports and type definitions
- Support different data source types

Error Handling:
- Handle missing manifest files
- Validate JSON format
- Test unknown type fallbacks

Edge Cases:
- Parameterized types (VARCHAR(255), NUMBER(10,2))
- Array types, boolean types
- Custom timestamp fields
- Column exclusion

End-to-End Workflow:
- Complete dbt → Feast import flow
- Multiple models with different entities
- Tag-based filtering

Tests use dynamic values from dbt-generated manifests to ensure
real integration testing rather than static fixtures.

Signed-off-by: David Soria Parra <dsp@php.net>
Signed-off-by: yassinnouh21 <yassinnouh21@gmail.com>

Add CI workflow that:
1. Installs Python dependencies and Feast
2. Installs dbt-core and dbt-duckdb
3. Runs dbt build to generate manifest.json
4. Sets up Feast project with SQLite
5. Tests feast dbt import command
6. Verifies Feast objects were created
7. Runs pytest integration tests

Workflow runs on PRs that modify:
- dbt integration code (feast/dbt/**)
- dbt integration tests (tests/integration/dbt/**)
- The workflow file itself

Tests run on Python 3.11 and 3.12 to ensure compatibility.

Signed-off-by: David Soria Parra <dsp@php.net>
Signed-off-by: yassinnouh21 <yassinnouh21@gmail.com>