lumis-ai 0.1.0a1__py3-none-any.whl → 0.1.1a2__py3-none-any.whl

lumis/core/django/__init__.py

@@ -1,5 +1,8 @@
 from __future__ import annotations
 
-from .models.chat_memory import ChatMemory
+try:
+    from .models.chat_memory import ChatMemory
 
-__all__ = ["ChatMemory"]
+    __all__ = ["ChatMemory"]
+except ImportError:
+    __all__ = []

lumis/nlp/information_extraction/FactExtractor.py

@@ -1,9 +1,6 @@
-from typing import Optional, Set, Tuple
+from __future__ import annotations
 
-import spacy
-from spacy.language import Language
-from spacy.matcher import Matcher
-from spacy.tokens import Doc, Span, Token
+from typing import Optional, Set, Tuple
 
 
 # TODO: Convert this to a spacy factory
@@ -13,9 +10,19 @@ class FactExtractor:
         Initialize the FactExtractor with a spaCy NLP pipeline.
         :param model: The name of the spaCy model to load.
         """
+        try:
+            import spacy
+            from spacy.matcher import Matcher
+            from spacy.tokens import Doc
+        except ImportError:
+            raise ImportError(
+                "spacy is required for FactExtractor. "
+                "Install it with: pip install lumis-ai[spacy]"
+            )
+
         # spaCy handles caching and singleton behavior automatically
-        self.nlp: Language = spacy.load(model)
-        self.matcher: Matcher = Matcher(self.nlp.vocab)
+        self.nlp = spacy.load(model)
+        self.matcher = Matcher(self.nlp.vocab)
         self._init_patterns()
 
         # Register custom extension if not already registered

lumis/nlp/ner/ner.py

@@ -1,13 +1,23 @@
+from __future__ import annotations
+
 from pathlib import Path
+from typing import TYPE_CHECKING
 
-from spacy.tokens import Doc
-from spacy_llm.util import assemble
+if TYPE_CHECKING:
+    from spacy.tokens import Doc
 
 relative_path = Path(__file__).parent / "config" / "entity_rel_extraction.cfg"
 
 
 class Ner:
     def __init__(self, config: str = str(relative_path)):
+        try:
+            from spacy_llm.util import assemble
+        except ImportError:
+            raise ImportError(
+                "spacy and spacy-llm are required for Ner. "
+                "Install them with: pip install lumis-ai[spacy]"
+            )
         self.nlp = assemble(config)
 
     def get_entities(self, text: str) -> list[tuple[str, str]]:

lumis/tools/scraper.py

@@ -278,16 +278,6 @@ class WebScrapper:
             raise e
 
         return None
-
-    # async def from_gotenberg(self, url: str, metadata: Optional[dict]) -> Optional[Document]:
-    #     # Same logic as before, but uses injected gb_client
-    #     try:
-    #         response = await self.gb_client.afrom_url(url)
-    #         return await self.parse_pdf(response.content, metadata)
-    #     except Exception as e:
-    #         logger.debug(f"Gotenberg failed for {url}: {e}")
-    #         return None
-
     def _url_with_protocol(self, url: str):
         if not url.startswith("http://") and not url.startswith("https://"):
             url = "https://" + url

lumis/tools/search/arxiv.py

@@ -7,12 +7,15 @@ import logging
 from typing import Any, Dict, Optional
 import xml.etree.ElementTree as ET
 
-import arxiv
-from arxiv import Client, SortCriterion, SortOrder
+from typing import TYPE_CHECKING
+
 import httpx
 from lumis.core.document import Document
 import pymupdf
 
+if TYPE_CHECKING:
+    from arxiv import SortCriterion, SortOrder
+
 logger = logging.getLogger(__name__)
 
 
@@ -36,8 +39,16 @@ class ArxivResult:
 
 class ArxivSearcher:
     def __init__(self, max_results: int = 10, max_concurrent_pdfs: int = 5) -> None:
+        try:
+            import arxiv
+        except ImportError:
+            raise ImportError(
+                "arxiv is required for ArxivSearcher. "
+                "Install it with: pip install lumis-ai[search]"
+            )
+        self._arxiv = arxiv
         self.max_results: int = max_results
-        self.client = Client()
+        self.client = arxiv.Client()
         self._http_client = httpx.AsyncClient()
         self._result_cache: dict[str, ArxivResult] = {}  # Cache for results by arxiv_id
         self._pdf_semaphore = asyncio.Semaphore(max_concurrent_pdfs)
@@ -46,8 +57,8 @@ class ArxivSearcher:
         self,
         query: str,
         max_results: Optional[int] = None,
-        sort_by: SortCriterion = SortCriterion.SubmittedDate,
-        sort_order: SortOrder = SortOrder.Descending,
+        sort_by: SortCriterion | None = None,
+        sort_order: SortOrder | None = None,
         id_list: Optional[list[str]] = None,
         read_pdfs: bool = False,
     ) -> list[ArxivResult]:
@@ -65,8 +76,13 @@ class ArxivSearcher:
         Returns:
             list of ArxivResult objects
         """
+        if sort_by is None:
+            sort_by = self._arxiv.SortCriterion.SubmittedDate
+        if sort_order is None:
+            sort_order = self._arxiv.SortOrder.Descending
+
         # Create search object with all available options
-        search = arxiv.Search(query=query, max_results=max_results or self.max_results, sort_by=sort_by, sort_order=sort_order, id_list=id_list or [])
+        search = self._arxiv.Search(query=query, max_results=max_results or self.max_results, sort_by=sort_by, sort_order=sort_order, id_list=id_list or [])
 
         results: list[ArxivResult] = []
         pdf_tasks = []

lumis/tools/search/pytrends.py

@@ -3,7 +3,6 @@ from __future__ import annotations
 from typing import Any, Dict, List
 
 from asgiref.sync import sync_to_async
-from pytrends.request import TrendReq
 
 
 class PyTrends:
@@ -12,6 +11,13 @@ class PyTrends:
     """
 
     def __init__(self, tz: int = 0, retries: int = 3, backoff: float = 0.1):
+        try:
+            from pytrends.request import TrendReq
+        except ImportError:
+            raise ImportError(
+                "pytrends is required for PyTrends. "
+                "Install it with: pip install lumis-ai[search]"
+            )
         self.pytrends = TrendReq(retries=retries, backoff_factor=backoff, tz=tz)  # type: ignore
 
     @sync_to_async

lumis/tools/search/yahoofinance.py

@@ -7,7 +7,6 @@ from typing import Any, cast, Optional
 from asgiref.sync import sync_to_async
 import pandas as pd
 from pydantic import BaseModel, Field
-import yfinance as yf
 
 logger = logging.getLogger(__name__)
 
@@ -53,6 +52,16 @@ class YahooFinance:
     with async codebases.
     """
 
+    def __init__(self):
+        try:
+            import yfinance as yf
+        except ImportError:
+            raise ImportError(
+                "yfinance is required for YahooFinance. "
+                "Install it with: pip install lumis-ai[search]"
+            )
+        self._yf = yf
+
     @sync_to_async
     def get_quote(self, ticker: str) -> QuoteData:
         """
@@ -70,7 +79,7 @@ class YahooFinance:
         """
         try:
             logger.debug(f"Fetching quote data for ticker: {ticker}")
-            t = yf.Ticker(ticker)
+            t = self._yf.Ticker(ticker)
             info = t.info
 
             return QuoteData(
@@ -114,7 +123,7 @@ class YahooFinance:
         """
         try:
             logger.debug(f"Fetching historical data for {ticker} with period={period}, interval={interval}")
-            t = yf.Ticker(ticker)
+            t = self._yf.Ticker(ticker)
             hist = t.history(period=period, interval=interval, start=start, end=end)
 
             return [
@@ -149,8 +158,8 @@ class YahooFinance:
         """
         try:
             logger.debug(f"Searching for tickers matching: {query}")
-            # Using yf.Tickers() as get_tickers_by_search is deprecated
-            tickers = yf.Tickers(query)
+            # Using self._yf.Tickers() as get_tickers_by_search is deprecated
+            tickers = self._yf.Tickers(query)
             return [{"symbol": symbol, "name": ticker.info.get("shortName", symbol)} for symbol, ticker in tickers.tickers.items()]
         except Exception as e:
             logger.error(f"Error searching tickers for {query}: {str(e)}")
@@ -172,7 +181,7 @@ class YahooFinance:
         """
         try:
             logger.debug(f"Fetching recommendations for {ticker}")
-            t = yf.Ticker(ticker)
+            t = self._yf.Ticker(ticker)
             if t.recommendations is None:
                 return []
             df = cast(pd.DataFrame, t.recommendations)
@@ -197,7 +206,7 @@ class YahooFinance:
         """
         try:
             logger.debug(f"Fetching major holders for {ticker}")
-            t = yf.Ticker(ticker)
+            t = self._yf.Ticker(ticker)
             if t.major_holders is None:
                 return {}
             df = cast(pd.DataFrame, t.major_holders)

{lumis_ai-0.1.0a1.dist-info → lumis_ai-0.1.1a2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lumis-ai
-Version: 0.1.0a1
+Version: 0.1.1a2
 Summary: An AI agent framework for building LLM-powered applications
 Project-URL: Homepage, https://github.com/tareksanger/lumis
 Project-URL: Repository, https://github.com/tareksanger/lumis
@@ -20,15 +20,9 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.11
-Requires-Dist: arxiv>=0.5.0
 Requires-Dist: asgiref>=3.8.1
 Requires-Dist: beautifulsoup4>=4.12.0
-Requires-Dist: coreferee-model-en>=1.0.0
-Requires-Dist: coreferee>=1.3.0
 Requires-Dist: deepmerge>=2.0.0
-Requires-Dist: django>=4.2
-Requires-Dist: en-core-web-lg<3.8.0,>=3.7.0
-Requires-Dist: en-core-web-trf<3.8.0,>=3.7.0
 Requires-Dist: faiss-cpu>=1.7.0
 Requires-Dist: googlesearch-python>=1.2.0
 Requires-Dist: httpx>=0.27.0
@@ -47,15 +41,12 @@ Requires-Dist: pydantic>=2.9.2
 Requires-Dist: pymupdf>=1.22.0
 Requires-Dist: pypdf2>=3.0.0
 Requires-Dist: python-dateutil>=2.8.0
-Requires-Dist: pytrends>=5.1.0
 Requires-Dist: pyvis>=0.3.0
 Requires-Dist: readability-lxml>=3.3.0
 Requires-Dist: redis>=5.0.0
 Requires-Dist: requests>=2.28.0
 Requires-Dist: scikit-learn>=1.0.0
 Requires-Dist: sentence-transformers>=2.2.2
-Requires-Dist: spacy-llm>=0.7.2
-Requires-Dist: spacy<3.8.0,>=3.7.0
 Requires-Dist: tavily-python>=0.3.0
 Requires-Dist: tenacity>=9.0.0
 Requires-Dist: textstat>=0.7.0
@@ -64,7 +55,19 @@ Requires-Dist: torch>=2.5.0
 Requires-Dist: transformers>=4.0.0
 Requires-Dist: typing-extensions>=4.0.0
 Requires-Dist: wikipedia>=1.4.0
-Requires-Dist: yfinance>=0.2.0
+Provides-Extra: django
+Requires-Dist: django>=4.2; extra == 'django'
+Provides-Extra: search
+Requires-Dist: arxiv>=0.5.0; extra == 'search'
+Requires-Dist: pytrends>=5.1.0; extra == 'search'
+Requires-Dist: yfinance>=0.2.0; extra == 'search'
+Provides-Extra: spacy
+Requires-Dist: coreferee-model-en>=1.0.0; extra == 'spacy'
+Requires-Dist: coreferee>=1.3.0; extra == 'spacy'
+Requires-Dist: en-core-web-lg<3.8.0,>=3.7.0; extra == 'spacy'
+Requires-Dist: en-core-web-trf<3.8.0,>=3.7.0; extra == 'spacy'
+Requires-Dist: spacy-llm>=0.7.2; extra == 'spacy'
+Requires-Dist: spacy<3.8.0,>=3.7.0; extra == 'spacy'
 Description-Content-Type: text/markdown
 
 # lumis
@@ -88,7 +91,17 @@ An AI agent framework for building LLM-powered applications with multi-provider
 pip install lumis-ai
 ```
 
-After installing, download the required spaCy language models:
+### Optional extras
+
+Some integrations are opt-in to keep the base install lighter:
+
+| Extra      | What it adds                                  | Install                           |
+|------------|-----------------------------------------------|-----------------------------------|
+| `spacy`    | NER, fact extraction, coreference resolution  | `pip install lumis-ai[spacy]`     |
+| `search`   | arXiv, Google Trends, Yahoo Finance           | `pip install lumis-ai[search]`    |
+| `django`   | Django ORM memory backend                     | `pip install lumis-ai[django]`    |
+
+The `spacy` extra requires language models. After installing, download them:
 
 ```bash
 python -m spacy download en_core_web_lg

{lumis_ai-0.1.0a1.dist-info → lumis_ai-0.1.1a2.dist-info}/RECORD

@@ -24,7 +24,7 @@ lumis/core/document.py,sha256=bat9TjbHigFf5TLVy9r0Ft67zjIMPJMwmQrsE4gP-IA,1080
 lumis/core/event_emitter.py,sha256=2LwAo0gLIIKMqKocjrZ3g86BRjrtnplTQth0S8iDw_U,2094
 lumis/core/common/coloured_logger.py,sha256=Pc_WcpPXfMIGHagrnJQ2MjY6x8MI1U5ZbTt5xtb8mgg,943
 lumis/core/common/logger_mixin.py,sha256=TmswwCfgEi8BeatoWsa3gMk3qhCEMEmZyLK4rkZWLUU,1013
-lumis/core/django/__init__.py,sha256=1eLq5GHCa7iCrlb3eg85Ifx9TtnnsYgMeZ6rakyzkWY,105
+lumis/core/django/__init__.py,sha256=yOr6iSaA7VSlIpSwryrwtL5zIHuv5HcY-9Fa9czoyRI,155
 lumis/core/django/models/chat_memory.py,sha256=j5E36hza2nbIKjVlv0j0WjttmRvOLjvTnoMFgNyUVVQ,6612
 lumis/core/utils/__init__.py,sha256=2hfqwepX9pCGlsPHZ3P1Zv_7jW0BCuW5r9P0aCu-xvU,246
 lumis/core/utils/coroutine.py,sha256=B_KQfwa6_wWI7a3hVPWv8GLAyuDPQywPgBYDd0iI4cI,665
@@ -65,8 +65,8 @@ lumis/nlp/semantic_parser.py,sha256=ydJjGy5h-gzaG0AJkBZOwj8JagoPxefm8RZx9D9FnsM,
 lumis/nlp/spacy_summarizer.py,sha256=6_ZMOYJ0gs7bcsef-x6GEhzsssitlpf4Ovw58SfEllw,2762
 lumis/nlp/summarizer.py,sha256=R-oTjtS86TnySTXjVhNf_NaTNJNbrfX8p964V5YfEIA,6920
 lumis/nlp/vector_similarity_retriever.py,sha256=mR2FjISAkAgS5bzSBwZIdnSr3q9s_r0GhCgrt5PlMAA,1714
-lumis/nlp/information_extraction/FactExtractor.py,sha256=3gA2LbAO9xOdcguoZQVssQYfDCOT6O9glWqBCUWxu_o,9355
-lumis/nlp/ner/ner.py,sha256=Dfxz2fuDedrd5kwLTHnpQQ5KnFOsmHEe1IfmvgQMYcI,504
+lumis/nlp/information_extraction/FactExtractor.py,sha256=Chu-fMoyBJe0axM-aH6T_LWcWCVyh_lzZOSRzePszJY,9564
+lumis/nlp/ner/ner.py,sha256=CVTft8EKgl3kvy9gzM0I2ANnLNmAGnYz0SdXXdf34II,819
 lumis/nlp/ner/config/entity_rel_extraction.cfg,sha256=rQKm-Agl_qse0EwI9XgLy0kjU4hdLe4B15bvspny5qY,923
 lumis/nlp/ner/config/fill-config.cfg,sha256=iMhF-QfZs15UtDHTBC3OPAhfaqyHMpnvqRAcWVAMU6U,4130
 lumis/reader/__init__.py,sha256=707frfd3cLTb7LTjhsWyFhb-u21YqaifWcyiCpVgYvg,52
@@ -76,15 +76,15 @@ lumis/storage/base_vector_db.py,sha256=ab4FyHzD6vYfyjuxPX19Vj10FGncndnRYWAzFiafR
 lumis/storage/chroma_vector_db.py,sha256=NPh5sYC_cdWWqWyACS4cLIrOjmJjKpDo7D2rS8WTmjw,5862
 lumis/storage/faiss_vector_db.py,sha256=Fkh3cFGcl6CFJ_Mja_ou9B2isXGvONRUx44qhco-YeY,12790
 lumis/tools/__init__.py,sha256=pzYG4pj-lNA1HtZ8ySDHhl1wjeDWJXWJyQje3zUTpPk,183
-lumis/tools/scraper.py,sha256=2aYyx5L0V3sKD_HHwIeiOq_ZlrRdcCKd7gyszOltwnY,14098
+lumis/tools/scraper.py,sha256=0iNQS2jptT093ZkOSKamvNDOQuJRBaAwM2_ztmS1w6U,13670
 lumis/tools/search/__init__.py,sha256=_MGFJAeizjcX0oI9Cr0B9efTfolRbL_3QQo2ZcNUklE,283
-lumis/tools/search/arxiv.py,sha256=wjgTKZSefC98Zt7cwiFNR3eRmFYrQoaU8Fa4ABlhvqs,9783
-lumis/tools/search/pytrends.py,sha256=yY5c4hlflqgvanWRAO4C1PtzBMx1rVPZfbqMaEukpm4,1764
+lumis/tools/search/arxiv.py,sha256=ldRHo_SSyhKm7CwjCuOjxZlIhkutwu85_UD8YcXCAtI,10245
+lumis/tools/search/pytrends.py,sha256=_tTmGMNNqSUbN3s7Scrx771KzX9nLHuYQrxFPC_nXIo,1980
 lumis/tools/search/search_engine_client.py,sha256=5Q0nb7Lztj0mjaN3EwZQfpRMNUbxcHyhp-wGbmOZdiA,11766
 lumis/tools/search/vector_search_retrieval_engine.py,sha256=RGZyFah2F-ojFzH9xiApILK5oQWWZEuihV2EN3Ae0cc,7889
 lumis/tools/search/wiki.py,sha256=gGm6I273KOLvDN0E7boAezjPq6cywRhbZEBqB_Pem4U,7864
-lumis/tools/search/yahoofinance.py,sha256=xQTTMJU7IDfH2WkzIWPjWwwLao0unfMHGrbhVEJJSJg,7548
-lumis_ai-0.1.0a1.dist-info/METADATA,sha256=bVQmNhCCxjia24iGCDc-brqlxqfrp2Poy74ubstINpU,3783
-lumis_ai-0.1.0a1.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-lumis_ai-0.1.0a1.dist-info/licenses/LICENSE,sha256=cKZn_mblbydX1fOrVMvUIrelUE9AElKLAR_7-VYRDow,1068
-lumis_ai-0.1.0a1.dist-info/RECORD,,
+lumis/tools/search/yahoofinance.py,sha256=0Oz7kfdWI2ehllo9NM2wKS0WSAUeGnPyuVnfobyBXHA,7851
+lumis_ai-0.1.1a2.dist-info/METADATA,sha256=Z0Ajv1lr3HYpO6izksSyOIU-79wK1KGzMgG5l0-6K4Y,4630
+lumis_ai-0.1.1a2.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+lumis_ai-0.1.1a2.dist-info/licenses/LICENSE,sha256=cKZn_mblbydX1fOrVMvUIrelUE9AElKLAR_7-VYRDow,1068
+lumis_ai-0.1.1a2.dist-info/RECORD,,