langroid 0.36.0__tar.gz → 0.37.0__tar.gz

{langroid-0.36.0 → langroid-0.37.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langroid
-Version: 0.36.0
+Version: 0.37.0
 Summary: Harness LLMs with Multi-Agent Programming
 Author-email: Prasad Chalasani <pchalasani@gmail.com>
 License: MIT
@@ -12,6 +12,7 @@ Requires-Dist: async-generator<2.0,>=1.10
 Requires-Dist: bs4<1.0.0,>=0.0.1
 Requires-Dist: cerebras-cloud-sdk<2.0.0,>=1.1.0
 Requires-Dist: colorlog<7.0.0,>=6.7.0
+Requires-Dist: docling<3.0.0,>=2.16.0
 Requires-Dist: docstring-parser<1.0,>=0.16
 Requires-Dist: duckduckgo-search<7.0.0,>=6.0.0
 Requires-Dist: faker<19.0.0,>=18.9.0
@@ -32,9 +33,10 @@ Requires-Dist: onnxruntime<2.0.0,>=1.16.1
 Requires-Dist: openai<2.0.0,>=1.45.0
 Requires-Dist: pandas<3.0.0,>=2.0.3
 Requires-Dist: prettytable<4.0.0,>=3.8.0
-Requires-Dist: pydantic<2.10.2,>=1
+Requires-Dist: pydantic<3.0.0,>=1
 Requires-Dist: pygithub<2.0.0,>=1.58.1
 Requires-Dist: pygments<3.0.0,>=2.15.1
+Requires-Dist: pymupdf4llm<0.1.0,>=0.0.17
 Requires-Dist: pyparsing<4.0.0,>=3.0.9
 Requires-Dist: pytest-rerunfailures<16.0,>=15.0
 Requires-Dist: python-dotenv<2.0.0,>=1.0.0
@@ -55,14 +57,15 @@ Provides-Extra: all
 Requires-Dist: arango-datasets<2.0.0,>=1.2.2; extra == 'all'
 Requires-Dist: chainlit<3.0.0,>=2.0.1; extra == 'all'
 Requires-Dist: chromadb<=0.4.23,>=0.4.21; extra == 'all'
+Requires-Dist: docling<3.0.0,>=2.16.0; extra == 'all'
 Requires-Dist: fastembed<0.4.0,>=0.3.1; extra == 'all'
-Requires-Dist: huggingface-hub<0.22.0,>=0.21.2; extra == 'all'
+Requires-Dist: huggingface-hub<1.0.0,>=0.21.2; extra == 'all'
 Requires-Dist: litellm<2.0.0,>=1.30.1; extra == 'all'
 Requires-Dist: metaphor-python<0.2.0,>=0.1.23; extra == 'all'
 Requires-Dist: neo4j<6.0.0,>=5.14.1; extra == 'all'
 Requires-Dist: pdf2image<2.0.0,>=1.17.0; extra == 'all'
-Requires-Dist: pdfplumber<0.11.0,>=0.10.2; extra == 'all'
 Requires-Dist: psycopg2<3.0.0,>=2.9.7; extra == 'all'
+Requires-Dist: pymupdf4llm<0.1.0,>=0.0.17; extra == 'all'
 Requires-Dist: pymupdf<2.0.0,>=1.23.3; extra == 'all'
 Requires-Dist: pymysql<2.0.0,>=1.1.0; extra == 'all'
 Requires-Dist: pypdf>=5.1.0; extra == 'all'
@@ -74,7 +77,7 @@ Requires-Dist: sentence-transformers<3.0.0,>=2.2.2; extra == 'all'
 Requires-Dist: sqlalchemy<3.0.0,>=2.0.19; extra == 'all'
 Requires-Dist: torch<3.0.0,>=2.0.0; extra == 'all'
 Requires-Dist: transformers<5.0.0,>=4.40.1; extra == 'all'
-Requires-Dist: unstructured[docx,pdf,pptx]<0.10.18,>=0.10.16; extra == 'all'
+Requires-Dist: unstructured[docx,pdf,pptx]<1.0.0,>=0.16.15; extra == 'all'
 Requires-Dist: weaviate-client>=4.9.6; extra == 'all'
 Provides-Extra: arango
 Requires-Dist: arango-datasets<2.0.0,>=1.2.2; extra == 'arango'
@@ -89,13 +92,16 @@ Requires-Dist: psycopg2<3.0.0,>=2.9.7; extra == 'db'
 Requires-Dist: pymysql<2.0.0,>=1.1.0; extra == 'db'
 Requires-Dist: sqlalchemy<3.0.0,>=2.0.19; extra == 'db'
 Provides-Extra: doc-chat
+Requires-Dist: docling<3.0.0,>=2.16.0; extra == 'doc-chat'
 Requires-Dist: pdf2image<2.0.0,>=1.17.0; extra == 'doc-chat'
-Requires-Dist: pdfplumber<0.11.0,>=0.10.2; extra == 'doc-chat'
+Requires-Dist: pymupdf4llm<0.1.0,>=0.0.17; extra == 'doc-chat'
 Requires-Dist: pymupdf<2.0.0,>=1.23.3; extra == 'doc-chat'
 Requires-Dist: pypdf>=5.1.0; extra == 'doc-chat'
 Requires-Dist: pytesseract<0.4.0,>=0.3.10; extra == 'doc-chat'
 Requires-Dist: python-docx<2.0.0,>=1.1.0; extra == 'doc-chat'
-Requires-Dist: unstructured[docx,pdf,pptx]<0.10.18,>=0.10.16; extra == 'doc-chat'
+Requires-Dist: unstructured[docx,pdf,pptx]<1.0.0,>=0.16.15; extra == 'doc-chat'
+Provides-Extra: docling
+Requires-Dist: docling<3.0.0,>=2.16.0; extra == 'docling'
 Provides-Extra: docx
 Requires-Dist: python-docx<2.0.0,>=1.1.0; extra == 'docx'
 Provides-Extra: fastembed
@@ -104,7 +110,7 @@ Provides-Extra: hf-embeddings
 Requires-Dist: sentence-transformers<3.0.0,>=2.2.2; extra == 'hf-embeddings'
 Requires-Dist: torch<3.0.0,>=2.0.0; extra == 'hf-embeddings'
 Provides-Extra: hf-transformers
-Requires-Dist: huggingface-hub<0.22.0,>=0.21.2; extra == 'hf-transformers'
+Requires-Dist: huggingface-hub<1.0.0,>=0.21.2; extra == 'hf-transformers'
 Requires-Dist: sentence-transformers<3.0.0,>=2.2.2; extra == 'hf-transformers'
 Requires-Dist: torch<3.0.0,>=2.0.0; extra == 'hf-transformers'
 Requires-Dist: transformers<5.0.0,>=4.40.1; extra == 'hf-transformers'
@@ -125,13 +131,16 @@ Requires-Dist: pymysql<2.0.0,>=1.1.0; extra == 'mysql'
 Provides-Extra: neo4j
 Requires-Dist: neo4j<6.0.0,>=5.14.1; extra == 'neo4j'
 Provides-Extra: pdf-parsers
+Requires-Dist: docling<3.0.0,>=2.16.0; extra == 'pdf-parsers'
 Requires-Dist: pdf2image<2.0.0,>=1.17.0; extra == 'pdf-parsers'
-Requires-Dist: pdfplumber<0.11.0,>=0.10.2; extra == 'pdf-parsers'
+Requires-Dist: pymupdf4llm<0.1.0,>=0.0.17; extra == 'pdf-parsers'
 Requires-Dist: pymupdf<2.0.0,>=1.23.3; extra == 'pdf-parsers'
 Requires-Dist: pypdf>=5.1.0; extra == 'pdf-parsers'
 Requires-Dist: pytesseract<0.4.0,>=0.3.10; extra == 'pdf-parsers'
 Provides-Extra: postgres
 Requires-Dist: psycopg2<3.0.0,>=2.9.7; extra == 'postgres'
+Provides-Extra: pymupdf4llm
+Requires-Dist: pymupdf4llm<0.1.0,>=0.0.17; extra == 'pymupdf4llm'
 Provides-Extra: scrapy
 Requires-Dist: scrapy<3.0.0,>=2.11.0; extra == 'scrapy'
 Provides-Extra: sql
@@ -139,11 +148,11 @@ Requires-Dist: psycopg2<3.0.0,>=2.9.7; extra == 'sql'
 Requires-Dist: pymysql<2.0.0,>=1.1.0; extra == 'sql'
 Requires-Dist: sqlalchemy<3.0.0,>=2.0.19; extra == 'sql'
 Provides-Extra: transformers
-Requires-Dist: huggingface-hub<0.22.0,>=0.21.2; extra == 'transformers'
+Requires-Dist: huggingface-hub<1.0.0,>=0.21.2; extra == 'transformers'
 Requires-Dist: torch<3.0.0,>=2.0.0; extra == 'transformers'
 Requires-Dist: transformers<5.0.0,>=4.40.1; extra == 'transformers'
 Provides-Extra: unstructured
-Requires-Dist: unstructured[docx,pdf,pptx]<0.10.18,>=0.10.16; extra == 'unstructured'
+Requires-Dist: unstructured[docx,pdf,pptx]<1.0.0,>=0.16.15; extra == 'unstructured'
 Provides-Extra: vecdbs
 Requires-Dist: chromadb<=0.4.23,>=0.4.21; extra == 'vecdbs'
 Requires-Dist: lancedb<0.9.0,>=0.8.2; extra == 'vecdbs'
@@ -292,20 +301,28 @@ teacher_task.run()
 <summary> <b>Click to expand</b></summary>
 
 - **Jan 2025:**
-  - [0.33.0](https://github.com/langroid/langroid/releases/tag/0.33.3) Move from Poetry to uv!
+  - [0.36.0](https://github.com/langroid/langroid/releases/tag/0.36.0): Weaviate vector-db support (thanks @abab-dev).
+  - [0.35.0](https://github.com/langroid/langroid/releases/tag/0.35.0): Capture/Stream reasoning content from 
+    Reasoning LLMs (e.g. DeepSeek, OpenAI o1) in addition to final answer.
+  - [0.34.0](https://github.com/langroid/langroid/releases/tag/0.34.0): DocChatAgent 
+    chunk enrichment to improve retrieval. (collaboration with @dfm88). 
+  - [0.33.0](https://github.com/langroid/langroid/releases/tag/0.33.3) Move from Poetry to uv! (thanks @abab-dev).
   - [0.32.0](https://github.com/langroid/langroid/releases/tag/0.32.0) DeepSeek v3 support.
 - **Dec 2024:**
   - [0.31.0](https://github.com/langroid/langroid/releases/tag/0.31.0) Azure OpenAI Embeddings
-  - [0.30.0](https://github.com/langroid/langroid/releases/tag/0.30.0) Llama-cpp embeddings.
-  - [0.29.0](https://github.com/langroid/langroid/releases/tag/0.29.0) Custom Azure OpenAI Client 
+  - [0.30.0](https://github.com/langroid/langroid/releases/tag/0.30.0) Llama-cpp embeddings (thanks @Kwigg).
+  - [0.29.0](https://github.com/langroid/langroid/releases/tag/0.29.0) Custom Azure OpenAI Client (thanks 
+    @johannestang).
   - [0.28.0](https://github.com/langroid/langroid/releases/tag/0.28.0) `ToolMessage`: `_handler` field to override 
-default handler method name in `request` field. 
+default handler method name in `request` field (thanks @alexagr).
   - [0.27.0](https://github.com/langroid/langroid/releases/tag/0.27.0) OpenRouter Support.
   - [0.26.0](https://github.com/langroid/langroid/releases/tag/0.26.0) Update to latest Chainlit.
-  - [0.25.0](https://github.com/langroid/langroid/releases/tag/0.25.0) True Async Methods for agent and user-response.
+  - [0.25.0](https://github.com/langroid/langroid/releases/tag/0.25.0) True Async Methods for agent and 
+    user-response (thanks @alexagr).
 - **Nov 2024:**
   - **[0.24.0](https://langroid.github.io/langroid/notes/structured-output/)**: 
      Enables support for `Agent`s with strict JSON schema output format on compatible LLMs and strict mode for the OpenAI tools API.
+    (thanks @nilspalumbo).
   - **[0.23.0](https://langroid.github.io/langroid/tutorials/local-llm-setup/#local-llms-hosted-on-glhfchat)**: 
       support for LLMs (e.g. `Qwen2.5-Coder-32b-Instruct`) hosted on glhf.chat 
   - **[0.22.0](https://langroid.github.io/langroid/notes/large-tool-results/)**: 

{langroid-0.36.0 → langroid-0.37.0}/README.md

@@ -136,20 +136,28 @@ teacher_task.run()
 <summary> <b>Click to expand</b></summary>
 
 - **Jan 2025:**
-  - [0.33.0](https://github.com/langroid/langroid/releases/tag/0.33.3) Move from Poetry to uv!
+  - [0.36.0](https://github.com/langroid/langroid/releases/tag/0.36.0): Weaviate vector-db support (thanks @abab-dev).
+  - [0.35.0](https://github.com/langroid/langroid/releases/tag/0.35.0): Capture/Stream reasoning content from 
+    Reasoning LLMs (e.g. DeepSeek, OpenAI o1) in addition to final answer.
+  - [0.34.0](https://github.com/langroid/langroid/releases/tag/0.34.0): DocChatAgent 
+    chunk enrichment to improve retrieval. (collaboration with @dfm88). 
+  - [0.33.0](https://github.com/langroid/langroid/releases/tag/0.33.3) Move from Poetry to uv! (thanks @abab-dev).
   - [0.32.0](https://github.com/langroid/langroid/releases/tag/0.32.0) DeepSeek v3 support.
 - **Dec 2024:**
   - [0.31.0](https://github.com/langroid/langroid/releases/tag/0.31.0) Azure OpenAI Embeddings
-  - [0.30.0](https://github.com/langroid/langroid/releases/tag/0.30.0) Llama-cpp embeddings.
-  - [0.29.0](https://github.com/langroid/langroid/releases/tag/0.29.0) Custom Azure OpenAI Client 
+  - [0.30.0](https://github.com/langroid/langroid/releases/tag/0.30.0) Llama-cpp embeddings (thanks @Kwigg).
+  - [0.29.0](https://github.com/langroid/langroid/releases/tag/0.29.0) Custom Azure OpenAI Client (thanks 
+    @johannestang).
   - [0.28.0](https://github.com/langroid/langroid/releases/tag/0.28.0) `ToolMessage`: `_handler` field to override 
-default handler method name in `request` field. 
+default handler method name in `request` field (thanks @alexagr).
   - [0.27.0](https://github.com/langroid/langroid/releases/tag/0.27.0) OpenRouter Support.
   - [0.26.0](https://github.com/langroid/langroid/releases/tag/0.26.0) Update to latest Chainlit.
-  - [0.25.0](https://github.com/langroid/langroid/releases/tag/0.25.0) True Async Methods for agent and user-response.
+  - [0.25.0](https://github.com/langroid/langroid/releases/tag/0.25.0) True Async Methods for agent and 
+    user-response (thanks @alexagr).
 - **Nov 2024:**
   - **[0.24.0](https://langroid.github.io/langroid/notes/structured-output/)**: 
      Enables support for `Agent`s with strict JSON schema output format on compatible LLMs and strict mode for the OpenAI tools API.
+    (thanks @nilspalumbo).
   - **[0.23.0](https://langroid.github.io/langroid/tutorials/local-llm-setup/#local-llms-hosted-on-glhfchat)**: 
       support for LLMs (e.g. `Qwen2.5-Coder-32b-Instruct`) hosted on glhf.chat 
   - **[0.22.0](https://langroid.github.io/langroid/notes/large-tool-results/)**: 

{langroid-0.36.0 → langroid-0.37.0}/langroid/agent/special/doc_chat_agent.py

@@ -15,6 +15,7 @@ pip install "langroid[hf-embeddings]"
 """
 
 import logging
+import textwrap
 from collections import OrderedDict
 from functools import cache
 from typing import Any, Callable, Dict, List, Optional, Set, Tuple, no_type_check
@@ -81,7 +82,7 @@ You will be given various passages from these documents, and asked to answer que
 about them, or summarize them into coherent answers.
 """
 
-CHUNK_ENRICHMENT_DELIMITER = "<##-##-##>"
+CHUNK_ENRICHMENT_DELIMITER = "\n<##-##-##>"
 
 has_sentence_transformers = False
 try:
@@ -99,7 +100,7 @@ hf_embed_config = SentenceTransformerEmbeddingsConfig(
 
 oai_embed_config = OpenAIEmbeddingsConfig(
     model_type="openai",
-    model_name="text-embedding-ada-002",
+    model_name="text-embedding-3-small",
     dims=1536,
 )
 
@@ -188,8 +189,8 @@ class DocChatAgentConfig(ChatAgentConfig):
             # NOTE: PDF parsing is extremely challenging, and each library
             # has its own strengths and weaknesses.
             # Try one that works for your use case.
-            # or "unstructured", "pdfplumber", "fitz", "pypdf"
-            library="pdfplumber",
+            # or "unstructured", "fitz", "pymupdf4llm", "pypdf"
+            library="pymupdf4llm",
         ),
     )
 
@@ -810,9 +811,11 @@ class DocChatAgent(ChatAgent):
         return "\n".join(
             [
                 f"""
-                [{i+1}]
+                -----[EXTRACT #{i+1}]----------
                 {content}
                 {source}
+                -----END OF EXTRACT------------
+                
                 """
                 for i, (content, source) in enumerate(zip(contents, sources))
             ]
@@ -949,12 +952,13 @@ class DocChatAgent(ChatAgent):
                     continue
 
                 # Combine original content with questions in a structured way
-                combined_content = f"""
-                {doc.content}
-                
+                combined_content = textwrap.dedent(
+                    f"""\
+                {doc.content}                
                 {enrichment_config.delimiter}
                 {enrichment}
-                """.strip()
+                """
+                )
 
                 new_doc = doc.copy(
                     update={
@@ -1440,7 +1444,7 @@ class DocChatAgent(ChatAgent):
         delimiter = self.config.chunk_enrichment_config.delimiter
         return [
             (
-                doc.copy(update={"content": doc.content.split(delimiter)[0].strip()})
+                doc.copy(update={"content": doc.content.split(delimiter)[0]})
                 if doc.content and getattr(doc.metadata, "has_enrichment", False)
                 else doc
             )

{langroid-0.36.0 → langroid-0.37.0}/langroid/embedding_models/models.py

@@ -18,7 +18,7 @@ AzureADTokenProvider = Callable[[], str]
 
 class OpenAIEmbeddingsConfig(EmbeddingModelsConfig):
     model_type: str = "openai"
-    model_name: str = "text-embedding-ada-002"
+    model_name: str = "text-embedding-3-large"
     api_key: str = ""
     api_base: Optional[str] = None
     organization: str = ""
@@ -28,7 +28,7 @@ class OpenAIEmbeddingsConfig(EmbeddingModelsConfig):
 
 class AzureOpenAIEmbeddingsConfig(EmbeddingModelsConfig):
     model_type: str = "azure-openai"
-    model_name: str = "text-embedding-ada-002"
+    model_name: str = "text-embedding-3-large"
     api_key: str = ""
     api_base: str = ""
     deployment_name: Optional[str] = None

{langroid-0.36.0 → langroid-0.37.0}/langroid/exceptions.py

@@ -1,4 +1,4 @@
-from typing import Optional
+from typing import List, Optional
 
 
 class XMLException(Exception):
@@ -15,7 +15,7 @@ class LangroidImportError(ImportError):
     def __init__(
         self,
         package: Optional[str] = None,
-        extra: Optional[str] = None,
+        extra: Optional[str | List[str]] = None,
         error: str = "",
         *args: object,
     ) -> None:
@@ -33,9 +33,21 @@ class LangroidImportError(ImportError):
             error = f"{package} is not installed by default with Langroid.\n"
 
         if extra:
+            if isinstance(extra, list):
+                help_preamble = f"""
+                If you want to use it, please install langroid with one of these 
+                extras: {', '.join(extra)}. The examples below use the first one, 
+                i.e. {extra[0]}.
+                """
+                extra = extra[0]
+            else:
+                help_preamble = f"""
+                If you want to use it, please install langroid with the
+                `{extra}` extra.
+                """
+
             install_help = f"""
-                If you want to use it, please install langroid 
-                with the `{extra}` extra, for example:
+                {help_preamble}
                 
                 If you are using pip:
                 pip install "langroid[{extra}]"

{langroid-0.36.0 → langroid-0.37.0}/langroid/parsing/code_parser.py

@@ -65,7 +65,7 @@ class CodeParsingConfig(BaseSettings):
         "bash",
     ]
     chunk_size: int = 500  # tokens
-    token_encoding_model: str = "text-embedding-ada-002"
+    token_encoding_model: str = "text-embedding-3-small"
     n_similar_docs: int = 4