sibi-flux 2026.1.7__tar.gz → 2026.1.9__tar.gz

{sibi_flux-2026.1.7 → sibi_flux-2026.1.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: sibi-flux
-Version: 2026.1.7
+Version: 2026.1.9
 Summary: Sibi Toolkit: A collection of tools for Data Analysis/Engineering.
 Author: Luis Valverde
 Author-email: Luis Valverde <lvalverdeb@gmail.com>

{sibi_flux-2026.1.7 → sibi_flux-2026.1.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sibi-flux"
-version = "2026.1.7"
+version = "2026.1.9"
 description = "Sibi Toolkit: A collection of tools for Data Analysis/Engineering."
 readme = "README.md"
 authors = [
@@ -97,6 +97,7 @@ members = [
     "test-prj",
     "test_prj",
     "test_prj_verify",
+    "clean-test-prj",
 ]
 
 [tool.pytest.ini_options]

{sibi_flux-2026.1.7 → sibi_flux-2026.1.9}/src/sibi_flux/datacube/cli.py

@@ -203,7 +203,7 @@ def set_context_defaults(
 
 
 def _get_db_url_callback(
-    registry: DatacubeRegistry, db_url_map: Optional[str]
+    registry: DatacubeRegistry, db_url_map: Optional[str], params: Optional[Dict[str, Any]] = None
 ) -> Callable[[str], str]:
     """Helper to create a callback that resolves DB URLs from CLI overrides or registry."""
     cli_urls = json.loads(db_url_map) if db_url_map else {}
@@ -212,8 +212,24 @@ def _get_db_url_callback(
         # 1. CLI Override
         if conf_name in cli_urls:
             return cli_urls[conf_name]
+        
         # 2. Dynamic Resolution
-        url = resolve_db_url(conf_name, registry.global_imports)
+        imports = registry.global_imports
+        
+        # Check specific import_spec from params
+        if params and "databases" in params:
+             for db in params.get("databases", []):
+                  ref = db.get("connection_ref") or db.get("connection_obj")
+                  if ref == conf_name:
+                       spec = db.get("import_spec")
+                       if spec and isinstance(spec, dict):
+                            # Prepend specific module
+                            imports = [spec.get("module")] + imports
+                       elif db.get("global_import"):
+                            imports = [db.get("global_import")] + imports
+                       break
+        
+        url = resolve_db_url(conf_name, imports)
         if url:
             return url
         raise ValueError(
@@ -283,9 +299,16 @@ def sync(
                 
                 flat_tables = {}
                 for grp, tbls in existing_reg_data.items():
-                    if isinstance(tbls, dict):
+                    if grp == "tables" and isinstance(tbls, dict):
+                        # Modern format: tables are in 'tables' dict
+                        for t, t_meta in tbls.items():
+                            flat_tables[t] = t_meta
+                        continue
+
+                    if isinstance(tbls, dict) and grp != "global_imports":
+                        # Legacy Scoped Format (grp is connection_obj)
                         for t, t_meta in tbls.items():
-                            # Inject the config object (group key) so DatacubeRegistry knows the connection
+                            # Inject/Overwrite connection_obj only if implied by scope
                             t_meta["connection_obj"] = grp
                             flat_tables[t] = t_meta
                 
@@ -295,6 +318,11 @@ def sync(
                 console.print(f"[yellow]Warning: Could not load existing registry: {e}[/yellow]")
 
     registry = DatacubeRegistry(config_data, params=context.params)
+    
+    # Fix 2 Re-applied: Ensure global_imports are populated from params if registry loaded from file didn't have them
+    # This is critical for Clean Project + Force flows where registry might be partial or JIT
+    if not registry.global_imports and context.params.get("global_imports"):
+        registry.global_imports = context.params.get("global_imports")
 
     # --- Aggregation Phase ---
     # --- Aggregation Phase ---
@@ -453,7 +481,7 @@ def sync(
     registry.valid_paths = list(valid_paths)
     registry.valid_fieldmap_paths = context.valid_fieldmap_paths
 
-    get_url = _get_db_url_callback(registry, db_url_map)
+    get_url = _get_db_url_callback(registry, db_url_map, params=context.params)
 
     # Group tables by target file
     file_groups = registry.group_tables_by_file()
@@ -770,8 +798,22 @@ def discover(
         if not wl_filename:
             # Default convention: discovery_whitelist_<db_name>.yaml
             wl_filename = f"discovery_whitelist_{conn_obj}.yaml"
+        
         whitelist_path = config_path.parent / wl_filename
 
+        # Fallback 1: Auto-generated per-db (from scan/map fallback logic)
+        if not whitelist_path.exists():
+            fb = config_path.parent / f"discovery_whitelist_{conn_obj}.yaml"
+            if fb.exists():
+                whitelist_path = fb
+        
+        # Fallback 2: Global whitelist.yaml (Generated by 'whitelist' command)
+        if not whitelist_path.exists():
+            global_wl = config_path.parent / "whitelist.yaml"
+            if global_wl.exists():
+                console.print(f"[dim]Fallback: Using global {global_wl.name} for {db_name}[/dim]")
+                whitelist_path = global_wl
+
         # Determine rules path
         rules_filename = db_config.get("rules_file")
         if not rules_filename:
@@ -1014,8 +1056,21 @@ def scan(
             raise typer.Exit(code=1)
 
     # Resolve global output file
-    global_tables_file = params.get("all_tables_file") or "all_tables.yaml"
-    global_tables_path = config_path.parent / global_tables_file
+    # First check nested discovery block (standard), then root (fallback)
+    discovery_conf = params.get("discovery", {})
+    global_tables_file = (
+        discovery_conf.get("all_tables_file") 
+        or params.get("all_tables_file") 
+        or "all_tables.yaml"
+    )
+    
+    # If using absolute path (resolved from context), use it directly
+    # Otherwise treat as relative to config file location
+    path_obj = Path(global_tables_file)
+    if path_obj.is_absolute():
+        global_tables_path = path_obj
+    else:
+        global_tables_path = config_path.parent / global_tables_file
 
     # Load existing data to preserve config for DBs not being scanned
     all_tables_data = {}
@@ -1089,7 +1144,7 @@ def drift(
     config_data = _load_and_resolve_config(config_path)
 
     registry = DatacubeRegistry(config_data)
-    get_url = _get_db_url_callback(registry, db_url_map)
+    get_url = _get_db_url_callback(registry, db_url_map, params=config_data)
     cli_urls = json.loads(db_url_map) if db_url_map else {}
 
     drift_table = Table(title="Schema Drift Analysis")
@@ -1493,6 +1548,116 @@ def map(
     return
 
 
+@app.command()
+def workflow(
+    config_file: Optional[Path] = typer.Option(None, "--config"),
+    db_url_map: Optional[str] = typer.Option(None, "--db-urls"),
+    env_file: Optional[Path] = typer.Option(None, "--env-file", "-e"),
+    db_name: Optional[str] = typer.Option(
+        None, "--db", help="Target specific database (filtering map/sync)"
+    ),
+    app_name: Optional[str] = typer.Option(
+        None, "--app-name", help="Run create_cubes for the specified application"
+    ),
+    force: bool = typer.Option(False, "--force", "-f", help="Force rebuild/overwrite"),
+    dry_run: bool = typer.Option(
+        False, "--dry-run", help="Preview changes without executing"
+    ),
+) -> None:
+    """
+    Executes the full Datacube workflow: Scan -> Propose -> Discover -> Map -> Sync -> (Optional) Create Cubes.
+    """
+    console.rule("[bold magenta]Starting Datacube Workflow[/]")
+
+    # Resolve env_file: CLI > Params > Default (Consistent with discover/sync)
+    if env_file:
+        env_path = env_file
+    elif context.params and context.params.get("defaults", {}).get("env_file"):
+        env_path = Path(context.params.get("defaults", {})["env_file"])
+    else:
+        env_path = Path(".env.linux")
+    load_environment(env_path, logger=console.print)
+
+    # Ensure context is configured with params from the config file
+    config_path = config_file or context.default_config
+    if config_path and config_path.exists():
+        # Load and resolve config to ensure context.params is populated
+        # This is critical for 'discover' to find 'databases' list
+        try:
+            config_data = _load_and_resolve_config(config_path)
+            context.params = config_data
+            context.default_config = config_path
+            # We don't have easy access to other set_context_defaults args here, 
+            # but params is what 'discover' needs most.
+            console.print(f"[dim]Loaded workflow context from {config_path}[/dim]")
+        except Exception as e:
+            console.print(f"[yellow]Warning: Failed to load context from {config_path}: {e}[/yellow]")
+
+    # Step 0: Scan & Propose Rules (Critical for new projects/drift)
+    console.print(f"\n[bold cyan]Step 0a: Scan (Introspection)[/]")
+    scan(
+        config_file=config_file,
+        db_url_map=db_url_map,
+        env_file=env_file,
+        db_name=db_name,
+    )
+
+    console.print(f"\n[bold cyan]Step 0b: Propose Rules[/]")
+    propose_rules(
+        config_file=config_file,
+        dry_run=dry_run,  # Preview rules if dry-running
+    )
+
+    console.print(f"\n[bold cyan]Step 0c: Generate Whitelist[/]")
+    whitelist(
+        config_file=config_file,
+        db_name=db_name,
+        db_url_map=db_url_map,
+        env_file=env_file,
+        force=force,
+    )
+
+    # Step 1: Discover
+    # Note: Discover currently runs for all DBs in config.
+    console.print(f"\n[bold cyan]Step 1: Discovery[/]")
+    discover(
+        config_file=config_file,
+        db_conf="replica_db_conf", # Pass explicit string to avoid Typer OptionInfo default
+        db_url_map=db_url_map,
+        env_file=env_file,
+        update=True,  # Always update registry to enable chaining
+        prune=False,  # Safer default? Or should we expose prune?
+        run_sync=False, # We handle sync explicitly later
+        dry_run=dry_run,
+        generate_fields=False, # We use map command for this
+        force=False, # Discovery force is about fields usually?
+    )
+
+    # Step 2: Map (Skipped - Handled by Sync)
+    # console.print(f"\n[bold cyan]Step 2: Map Generation (Integrated into Sync)[/]")
+    
+    # Step 3: Sync -> Generates Code + Field Maps
+    console.print(f"\n[bold cyan]Step 3: Sync (Code Generation)[/]")
+    sync(
+        config_file=config_file,
+        db_url_map=db_url_map,
+        env_file=env_file, # sync command has env_file
+        force=force,
+        dry_run=dry_run,
+    )
+
+    # Step 4: Create Cubes (Optional)
+    if app_name:
+        console.print(f"\n[bold cyan]Step 4: Application Cubes ({app_name})[/]")
+        if dry_run:
+            console.print("[yellow]Skipping create_cubes in dry-run mode.[/yellow]")
+        else:
+            from sibi_flux.init.cube_extender import create_cubes
+            create_cubes(app_name)
+
+    console.rule("[bold magenta]Workflow Completed[/]")
+
+
 @app.command()
 def init(
     config_file: Optional[Path] = typer.Option(None, "--config"),

{sibi_flux-2026.1.7 → sibi_flux-2026.1.9}/src/sibi_flux/datacube/generator.py

@@ -778,6 +778,32 @@ def generate_datacube_module_code(
         else:
             imports.add(base_imp)
 
+        # Config Object Import
+        # We need to resolve import for conf_obj (e.g. clickhouse_conf)
+        # Using registry.params which contains databases list with import_spec
+        params = registry.params
+        dbs = params.get("databases", [])
+        for db in dbs:
+            ref = db.get("connection_ref") or db.get("connection_obj")
+            if ref == conf_obj:
+               spec = db.get("import_spec")
+               if spec and isinstance(spec, dict):
+                   mod = spec.get("module")
+                   sym = spec.get("symbol")
+                   if mod and sym:
+                       imports.add(f"from {mod} import {sym}")
+               elif db.get("global_import"):
+                   # Legacy or simple string import
+                   imp_str = db.get("global_import")
+                   if imp_str and sym and sym in imp_str:
+                        imports.add(imp_str)
+                   elif imp_str: # Fallback to add specific object if simple import
+                       # Try to parse or just add "from ... import conf_obj" logic?
+                       # If global_import is "from conf.credentials import *" -> hard to know
+                       # If it is "from conf.credentials import replica_conf" -> easy
+                       imports.add(imp_str)
+               break
+
         details = registry.get_table_details(table_name)
         field_map_str = details.get("field_map")
         sticky_filters = details.get("sticky_filters")

{sibi_flux-2026.1.7 → sibi_flux-2026.1.9}/src/sibi_flux/init/cube_proposer.py

@@ -56,6 +56,8 @@ def propose_cubes(db_domain: str, app_name: str) -> None:
     
     # Registry Structure: {conf_obj: {table_name: {class_name: ..., path: ...}}}
     for conf_obj, tables in global_data.items():
+        if not isinstance(tables, dict):
+             continue
         for table_name, meta in tables.items():
             path_str = meta.get("path", "")
             class_name = meta.get("class_name")
@@ -100,14 +102,16 @@ def propose_cubes(db_domain: str, app_name: str) -> None:
     
     # 4. Update App Registry
     
-    # Load existing
     current_config = {"cubes": []}
     if registry_path.exists():
         try:
             with open(registry_path, "r") as f:
                 loaded = yaml.safe_load(f)
-                if loaded and "cubes" in loaded:
-                    current_config = loaded
+                if loaded:
+                     current_config = loaded
+                     if "cubes" not in current_config or current_config["cubes"] is None:
+                         current_config["cubes"] = []
+    
         except Exception:
             pass