chaskiwasi 0.2.0__tar.gz

chaskiwasi-0.2.0/.github/workflows/ci.yml

@@ -0,0 +1,31 @@
+name: CI (Tests)
+
+on:
+  push:
+    branches: [ "main", "master" ]
+  pull_request:
+    branches: [ "main", "master" ]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
+
+    steps:
+    - uses: actions/checkout@v4
+    
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v5
+      with:
+        python-version: "3.10"
+        
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install pytest
+        pip install .      
+    - name: Run tests
+      run: |
+        pytest

chaskiwasi-0.2.0/.github/workflows/publish.yml

@@ -0,0 +1,35 @@
+name: Publish to PyPI
+
+on:
+  release:
+    types: [published]
+  workflow_dispatch: # Permite activarlo manualmente desde la interfaz de GitHub
+
+permissions:
+  id-token: write # Requerido obligatoriamente para Trusted Publishers (OIDC)
+
+jobs:
+  pypi-publish:
+    name: Upload release to PyPI
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/chaskiwasi # Reemplaza con el nombre real de tu paquete en PyPI
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: "3.10"
+
+    - name: Install build dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install build
+
+    - name: Build package
+      run: python -m build
+
+    - name: Publish package to PyPI
+      uses: pypa/gh-action-pypi-publish@release/v1

chaskiwasi-0.2.0/.gitignore

@@ -0,0 +1,142 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff
+instance/
+.webassets-cache
+
+# Scrapy stuff
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+# For a library or package, you might want to ignore these files since the code is
+# intended to run in multiple environments; otherwise, let them be committed:
+.python-version
+
+# Pipenv
+#Rely on Pipenv to maintain the version lockfile, but commit locked dependencies:
+#Pipfile.lock
+
+# Poetry
+#Rely on Poetry to maintain the version lockfile, but commit locked dependencies:
+#poetry.lock
+
+# pdm
+#Rely on PDM to maintain the version lockfile, but commit locked dependencies:
+#pdm.lock
+
+# PEP 582; used by e.g. __pypackages__
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mypy
+.mypy_cache/
+.dmypy.json
+.dmypy.root
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# IDEs and editors
+.idea/
+.vscode/
+*.swp
+*.swo
+.DS_Store
+Thumbs.db

chaskiwasi-0.2.0/PKG-INFO

@@ -0,0 +1,31 @@
+Metadata-Version: 2.5
+Name: chaskiwasi
+Version: 0.2.0
+Summary: Sistema de consulta semántica para fuentes jurídicas.
+Project-URL: Homepage, https://github.com/Maorda/chaskywasi
+Project-URL: Issues, https://github.com/Maorda/chaskywasi/issues
+Author-email: Luis Maurtua <luis.maurtua@ejemplo.com>
+License: MIT
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Python: >=3.10
+Requires-Dist: chromadb>=0.5.0
+Requires-Dist: docling>=2.0.0
+Requires-Dist: google-genai>=0.1.0
+Requires-Dist: httpx>=0.27.0
+Requires-Dist: ijson>=3.3.0
+Requires-Dist: requests>=2.32.0
+Requires-Dist: streamlit>=1.30.0
+Requires-Dist: tiktoken>=0.7.0
+Provides-Extra: dev
+Requires-Dist: anyio>=4.0.0; extra == 'dev'
+Requires-Dist: faker>=20.0.0; extra == 'dev'
+Requires-Dist: pytest-asyncio>=0.23.0; extra == 'dev'
+Requires-Dist: pytest>=8.0.0; extra == 'dev'
+Description-Content-Type: text/markdown
+
+GOLA

chaskiwasi-0.2.0/README.md

@@ -0,0 +1 @@
+GOLA

chaskiwasi-0.2.0/pyproject.toml

@@ -0,0 +1,47 @@
+[build-system]
+requires = ["hatchling>=1.18.0"]
+build-backend = "hatchling.build"
+
+[project]
+name = "chaskiwasi"  # <-- Cambiado al nombre real de tu librería
+version = "0.2.0"
+description = "Sistema de consulta semántica para fuentes jurídicas."
+readme = "README.md"
+authors = [
+    { name = "Luis Maurtua", email = "luis.maurtua@ejemplo.com" }
+]
+license = { text = "MIT" }
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+]
+requires-python = ">=3.10"
+dependencies = [
+    "docling>=2.0.0",
+    "tiktoken>=0.7.0",
+    "chromadb>=0.5.0",
+    "ijson>=3.3.0",
+    "requests>=2.32.0",
+    "streamlit>=1.30.0",
+    "google-genai>=0.1.0",
+    "httpx>=0.27.0"
+]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=8.0.0",
+    "pytest-asyncio>=0.23.0",
+    "anyio>=4.0.0",
+    "faker>=20.0.0"
+]
+
+[project.urls]
+Homepage = "https://github.com/Maorda/chaskywasi"
+Issues = "https://github.com/Maorda/chaskywasi/issues"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/chaskiwasi"]

chaskiwasi-0.2.0/src/chaskiwasi/chunking/chunk_tokenizer.py

@@ -0,0 +1,43 @@
+# dantesito/chasky/chunking/chunk_tokenizer.py
+
+from typing import List
+
+import tiktoken
+
+from chaskiwasi.config.settings import Settings
+
+
+class ChunkTokenizer:
+    """Divide textos en fragmentos mediante ventanas de tokens de tiktoken."""
+
+    def __init__(self) -> None:
+        self._encoding = tiktoken.get_encoding(Settings.TIKTOKEN_ENCODING)
+
+    def split_text(self, text: str) -> List[str]:
+        """Divide el texto en chunks determinados exclusivamente por tokens."""
+        if not isinstance(text, str) or not text:
+            return []
+
+        tokens: List[int] = self._encoding.encode(text)
+
+        if not tokens:
+            return []
+
+        chunk_size: int = Settings.CHUNK_SIZE
+        step: int = Settings.CHUNK_SIZE - Settings.CHUNK_OVERLAP
+
+        chunks: List[str] = []
+        start: int = 0
+        total_tokens: int = len(tokens)
+
+        while start < total_tokens:
+            end: int = min(start + chunk_size, total_tokens)
+            chunk_tokens: List[int] = tokens[start:end]
+            chunks.append(self._encoding.decode(chunk_tokens))
+
+            if end >= total_tokens:
+                break
+
+            start += step
+
+        return chunks

chaskiwasi-0.2.0/src/chaskiwasi/classification/cascade_factory.py

@@ -0,0 +1,68 @@
+import logging
+from typing import Any, List, Optional, Tuple
+
+from chaskiwasi.classification.strategies.base_strategy import BaseStrategy
+
+logger = logging.getLogger(__name__)
+
+class CascadeFactory:
+    """
+    Fábrica y orquestador de clasificación en cascada agnóstico para chaskywasi.
+    Aplica una arquitectura de enrutamiento por estrategias ordenadas jerárquicamente.
+    """
+
+    def __init__(
+        self,
+        strategies: Optional[List[BaseStrategy]] = None,
+        default_source: Optional[Any] = None,
+        default_section: Optional[Any] = None,
+    ) -> None:
+        self._strategies: List[BaseStrategy] = strategies or []
+        self.default_source = default_source
+        self.default_section = default_section
+
+    def process_chunk(
+        self,
+        text: Optional[str],
+        current_source: Optional[Any] = None,
+    ) -> Tuple[Optional[Any], Optional[Any]]:
+        """Procesa un fragmento individual de texto iterando sobre la lista de estrategias."""
+        effective_default_source = current_source or self.default_source
+
+        if not text or not text.strip():
+            return effective_default_source, self.default_section
+
+        for strategy in self._strategies:
+            try:
+                src_res, sec_res = strategy.classify(
+                    text, current_source=effective_default_source
+                )
+
+                if sec_res is not None:
+                    final_source = src_res or effective_default_source
+                    return final_source, sec_res
+            except Exception as e:
+                logger.warning(
+                    "Estrategia %s falló durante la clasificación: %s",
+                    strategy.__class__.__name__,
+                    str(e),
+                )
+
+        return effective_default_source, self.default_section
+
+    def process_chunks_batch(
+        self,
+        chunks: List[str],
+        current_source: Optional[Any] = None,
+    ) -> List[Tuple[Optional[Any], Optional[Any]]]:
+        """Procesa una lista de fragmentos secuencialmente manteniendo el contexto actual."""
+        results = []
+        active_source = current_source or self.default_source
+        
+        for chunk in chunks:
+            src, sec = self.process_chunk(chunk, current_source=active_source)
+            if src is not None:
+                active_source = src
+            results.append((src, sec))
+            
+        return results

chaskiwasi-0.2.0/src/chaskiwasi/classification/strategies/base_strategy.py

@@ -0,0 +1,43 @@
+# dantesito/chasky/classification/strategies/base_strategy.py
+from abc import ABC, abstractmethod
+from typing import Tuple, Optional, Any
+
+# 🚀 CONTRATO ACTIVO: Importación obligatoria para control de tipos dinámicos
+from chaskiwasi.config.taxonomy_registry import TaxonomyRegistry
+
+
+class BaseStrategy(ABC):
+    """
+    Interfaz contractual agnóstica para todas las estrategias de inferencia semántica.
+    
+    Expone de forma segura las clases de Enums inyectadas por el cliente mediante
+    propiedades dinámicas de solo lectura, eliminando la fragilidad de constructores.
+    """
+
+    @property
+    def source_enum_class(self) -> Any:
+        """
+        Devuelve de forma dinámica la clase SourceEnum del cliente.
+        🛡️ CORTAFUEGOS EN CALIENTE: Lanza RuntimeError si no fue inyectada al arrancar.
+        """
+        return TaxonomyRegistry.get_source_enum()
+
+    @property
+    def section_enum_class(self) -> Any:
+        """
+        Devuelve de forma dinámica la clase SectionEnum del cliente.
+        🛡️ CORTAFUEGOS EN CALIENTE: Lanza RuntimeError si no fue inyectada al arrancar.
+        """
+        return TaxonomyRegistry.get_section_enum()
+
+    @abstractmethod
+    def classify(self, chunk: str, current_source: Optional[str] = None) -> Tuple[Optional[Any], Optional[Any]]:
+        """
+        Analiza un fragmento y devuelve una tupla conteniendo las instancias
+        de los Enums dinámicos del cliente (SourceEnum, SectionEnum).
+
+        :param chunk: Fragmento de texto Markdown a evaluar.
+        :param current_source: Contexto de origen acumulado o arrastrado.
+        :return: Tupla conteniendo (Instancia de SourceEnum, Instancia de SectionEnum)
+        """
+        pass

chaskiwasi-0.2.0/src/chaskiwasi/classification/strategies/context_overlap_strategy.py

@@ -0,0 +1,21 @@
+# dantesito/chasky/classification/strategies/context_overlap_strategy.py
+
+from typing import Optional, Tuple
+
+from chaskiwasi.classification.strategies.base_strategy import BaseStrategy
+from chaskiwasi.config.taxonomy_registry import SectionEnum, SourceEnum
+
+
+class ContextOverlapStrategy(BaseStrategy):
+    """Arrastra la fuente institucional desde el contexto del chunk anterior."""
+
+    def classify(
+        self,
+        chunk_text: str,
+        current_source: Optional[SourceEnum] = None,
+    ) -> Tuple[Optional[SourceEnum], Optional[SectionEnum]]:
+        """Mantiene la fuente previa sin inspeccionar el contenido del chunk."""
+        if current_source is not None:
+            return current_source, None
+
+        return None, None

chaskiwasi-0.2.0/src/chaskiwasi/classification/strategies/cpu_regex_strategy.py

@@ -0,0 +1,62 @@
+import re
+from dataclasses import dataclass, field
+from typing import Any, List, Optional, Pattern, Tuple
+from chaskiwasi.classification.strategies.base_strategy import BaseStrategy
+
+
+@dataclass
+class RegexRule:
+    """Estructura de datos para definir reglas de clasificación por Regex."""
+    source: Any
+    section: Any
+    patterns: List[str]
+    flags: int = re.IGNORECASE | re.DOTALL
+    compiled_patterns: List[Pattern[str]] = field(init=False, repr=False)
+
+    def __post_init__(self) -> None:
+        compiled: List[Pattern[str]] = []
+        for pattern in self.patterns:
+            try:
+                compiled.append(re.compile(pattern, self.flags))
+            except re.error as err:
+                raise ValueError(
+                    f"Patrón Regex inválido '{pattern}' para la regla ({self.source}, {self.section}): {err}"
+                ) from err
+        self.compiled_patterns = compiled
+
+
+class CPURegexStrategy(BaseStrategy):
+    """Estrategia genérica que ejecuta reglas Regex inyectadas."""
+
+    def __init__(self, rules: Optional[List[RegexRule]] = None) -> None:
+        self.rules: List[RegexRule] = list(rules) if rules else []
+
+    def add_rule(self, rule: RegexRule) -> None:
+        """Permite registrar reglas dinámicamente."""
+        self.rules.append(rule)
+
+    def classify(
+        self,
+        chunk_text: str,
+        current_source: Optional[Any] = None,
+    ) -> Tuple[Optional[Any], Optional[Any]]:
+        if not isinstance(chunk_text, str) or not chunk_text:
+            return None, None
+
+        # Evaluación con preferencia contextual por current_source si aplica
+        for rule in self.rules:
+            if current_source and rule.source != current_source:
+                continue
+                
+            if any(pattern.search(chunk_text) for pattern in rule.compiled_patterns):
+                return rule.source, rule.section
+
+        # Segunda pasada para reglas generales si no coincidió en el contexto actual
+        if current_source:
+            for rule in self.rules:
+                if rule.source == current_source:
+                    continue
+                if any(pattern.search(chunk_text) for pattern in rule.compiled_patterns):
+                    return rule.source, rule.section
+
+        return None, None

chaskiwasi-0.2.0/src/chaskiwasi/classification/strategies/gemini_streamer.py

@@ -0,0 +1,99 @@
+import logging
+import os
+from typing import Any, Dict, Optional, Tuple
+
+from google import genai
+from google.genai import types
+
+from chaskiwasi.classification.strategies.base_strategy import BaseStrategy
+
+logging.getLogger("google.genai").setLevel(logging.ERROR)
+logging.getLogger("google.genai._api_client").setLevel(logging.ERROR)
+
+logger = logging.getLogger(__name__)
+
+
+class GeminiStreamer(BaseStrategy):
+    """Estrategia de inferencia remota en la nube totalmente agnóstica al dominio."""
+
+    def __init__(
+        self,
+        system_instruction: str,
+        label_mapping: Dict[str, Any],
+    ) -> None:
+        self.system_instruction = system_instruction
+        self.label_mapping = label_mapping
+        
+        api_key = ""
+        env_key = os.environ.get("GEMINI_API_KEY", "").strip()
+        if env_key:
+            api_key = env_key
+
+        if not api_key:
+            env_path = ".env"
+            if os.path.exists(env_path):
+                with open(env_path, "r", encoding="utf-8") as f:
+                    for line in f:
+                        clean_line = line.strip()
+                        if clean_line.startswith("GEMINI_API_KEY"):
+                            try:
+                                parsed_tokens = clean_line.split("=", 1)
+                                if len(parsed_tokens) == 2:
+                                    parsed_key = parsed_tokens[1].strip().strip('"').strip("'")
+                                    if parsed_key:
+                                        api_key = parsed_key
+                                        break
+                            except Exception:
+                                pass
+
+        if not api_key or not (api_key.startswith("AIzaSy") or api_key.startswith("AQ.")):
+            raise ValueError(
+                "🚨 ERROR CRÍTICO DE CONFIGURACIÓN: La variable 'GEMINI_API_KEY' "
+                "no está definida en el entorno ni en el archivo .env raíz."
+            )
+
+        self.client = genai.Client(api_key=api_key)
+        self.model_name = "gemini-3.5-flash-lite"
+
+    def classify(
+        self, chunk_text: str, current_source: Optional[Any] = None
+    ) -> Tuple[Optional[Any], Optional[Any]]:
+        if not chunk_text or not chunk_text.strip():
+            return None, None
+
+        try:
+            response = self.client.models.generate_content(
+                model=self.model_name,
+                contents=chunk_text,
+                config=types.GenerateContentConfig(
+                    system_instruction=self.system_instruction,
+                    temperature=0.1,
+                ),
+            )
+
+            if not response or not hasattr(response, "text") or not response.text:
+                logger.warning("Respuesta vacía o no válida recibida de la API de Gemini.")
+                return None, None
+
+            label = response.text.strip().lower()
+
+            if label == "desconocido":
+                return None, None
+
+            section_enum = self.label_mapping.get(label)
+            if section_enum is None:
+                logger.warning(
+                    "Etiqueta desconocida o fuera de taxonomía devuelta por Gemini: '%s'", label
+                )
+                return None, None
+
+            # Asumimos que Gemini por ahora solo devuelve la sección, manteniendo la firma original
+            return None, section_enum
+
+        except Exception as e:
+            logger.error(
+                "Error durante la inferencia remota con el cliente Google GenAI: %s",
+                str(e),
+                exc_info=True,
+            )
+            return None, None