feat: 🚀 Add initial setup for Datalia

- Updated `.gitignore` to exclude dbt and data files.
- Revised README for improved clarity.
- Added `data` and `datalia` folder structure with essential initialization files.
- Imported essential packages in `assets.py` and defined assets for IPC data retrieval.
- Created `dbt` configurations including models for IPC data and CFFI setup.
- Included Python dependencies and configurations in `pyproject.toml`.
- Locked dependencies in `uv.lock`.

.gitignore

@@ -160,3 +160,12 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+# dbt
+dbt_packages/
+logs/
+dbt/.user.yml
+
+# data
+data/*
+!data/.gitkeep

README.md

@@ -1,2 +1,3 @@
-# datalia
-💃 Plataforma de datos abiertos a nivel de España para unificar y armonizar información proveniente de diferentes fuentes.
+# Datalia 💃
+
+Plataforma de datos abiertos a nivel de España con el objetivo de unificar y armonizar información proveniente de diferentes fuentes.

datalia/assets.py

@@ -0,0 +1,25 @@
+import polars as pl
+from dagster import AssetExecutionContext, RetryPolicy, asset
+from dagster_dbt import DbtCliResource, dbt_assets
+
+from .dbt import dbt_project
+
+
+@dbt_assets(manifest=dbt_project.manifest_path)
+def dbt(context: AssetExecutionContext, dbt: DbtCliResource):
+    yield from dbt.cli(["build"], context=context).stream()
+
+
+@asset(
+    retry_policy=RetryPolicy(max_retries=5),
+)
+def raw_spain_ipc() -> pl.DataFrame:
+    """
+    Raw IPC data from INE.
+    """
+
+    df = pl.read_csv(
+        "https://www.ine.es/jaxiT3/files/t/csv_bdsc/50904.csv", separator=";"
+    )
+
+    return df

datalia/dbt.py

@@ -0,0 +1,12 @@
+from pathlib import Path
+
+from dagster_dbt import DbtProject
+
+RELATIVE_PATH_TO_MY_DBT_PROJECT = "../dbt"
+
+dbt_project = DbtProject(
+    project_dir=Path(__file__)
+    .joinpath("..", RELATIVE_PATH_TO_MY_DBT_PROJECT)
+    .resolve(),
+)
+dbt_project.prepare_if_dev()

datalia/definitions.py

@@ -0,0 +1,18 @@
+import os
+
+from dagster import Definitions, load_assets_from_modules
+from dagster_dbt import DbtCliResource
+from dagster_duckdb_polars import DuckDBPolarsIOManager
+
+from . import assets, dbt
+
+DATABASE_PATH = os.getenv("DATABASE_PATH", "./data/database.duckdb")
+
+all_assets = load_assets_from_modules([assets])
+
+resources = {
+    "io_manager": DuckDBPolarsIOManager(database=DATABASE_PATH),
+    "dbt": DbtCliResource(project_dir=dbt.dbt_project),
+}
+
+defs = Definitions(assets=all_assets, resources=resources)

dbt/dbt_project.yml

@@ -0,0 +1,12 @@
+name: "datalia"
+version: "1.0.0"
+
+profile: "default"
+
+clean-targets:
+  - "target"
+  - "dbt_packages"
+
+models:
+  datalia:
+    +materialized: table

dbt/models/schema.yml

@@ -0,0 +1,5 @@
+version: 2
+models:
+  - name: spain_ipc
+    description: Índice de Precios al Consumo (IPC) en España.
+    columns: []

dbt/models/sources.yml

@@ -0,0 +1,9 @@
+version: 2
+
+sources:
+  - name: public
+    tables:
+      - name: raw_spain_ipc
+        meta:
+          dagster:
+            asset_key: ["raw_spain_ipc"]

dbt/models/spain_ipc.sql

@@ -0,0 +1,40 @@
+with source as (
+    select * from {{ source('public', 'raw_spain_ipc') }}
+),
+
+renamed as (
+    select
+        {{ adapter.quote("Clases") }} as class,
+        {{ adapter.quote("Tipo de dato") }} as type,
+        {{ adapter.quote("Periodo") }} as date,
+        {{ adapter.quote("Total") }} as value
+    from source
+),
+
+parsed as (
+    select
+        cast(strptime(REPLACE(date, 'M', '-'), '%Y-%m') as date) AS date,
+        class,
+        type,
+        try_cast(replace(value, ',', '.') AS FLOAT) AS value,
+    from renamed
+),
+
+cleaned as (
+    select
+        date,
+        case
+            when class != 'Índice general' then split_part(class, ' ', 1)
+            else '0000'
+        end as class_id,
+        case
+            when class != 'Índice general' then substring(class from position(' ' in class) + 1)
+            else 'Índice general'
+        end as class_name,
+        value
+    from parsed
+    where type = 'Índice'
+    order by date desc
+)
+
+select * from cleaned order by date desc

dbt/profiles.yml

@@ -0,0 +1,8 @@
+default:
+  outputs:
+    dev:
+      type: duckdb
+      path: "../data/database.duckdb"
+      threads: 1
+
+  target: dev

pyproject.toml

@@ -0,0 +1,24 @@
+[project]
+name = "datalia"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.12"
+dependencies = [
+    "dagster>=1.8.6",
+    "duckdb>=1.1.0",
+    "dagster-dbt>=0.24.6",
+    "dagster-duckdb>=0.24.6",
+    "dbt-duckdb>=1.8.3",
+    "ipykernel>=6.29.5",
+    "pyarrow>=17.0.0",
+    "dagster-duckdb-polars>=0.24.6",
+    "python-slugify>=8.0.4",
+]
+
+[tool.dagster]
+module_name = "datalia.definitions"
+code_location_name = "datalia"
+
+[tool.uv]
+dev-dependencies = ["dagster-webserver>=1.8.6"]