github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-08-23 03:22:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

content_blocks: add missing blocks to aliases and formatting

Browse Source
This commit is contained in:
Mason Daugherty 2025-07-31 17:08:10 -04:00
parent 2cb48b685f
commit e1b8ae2e6c
No known key found for this signature in database
1 changed files with 14 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
libs/core/langchain_core/messages/content_blocks.py
View File

@ -5,9 +5,8 @@
change in future releases. change in future releases.
This module provides a standardized data structure for representing inputs to and This module provides a standardized data structure for representing inputs to and
outputs from Large Language Models. The core abstraction is the **Content Block**, a outputs from LLMs. The core abstraction is the **Content Block**, a ``TypedDict`` that
``TypedDict`` that can represent a piece of text, an image, a tool call, or other can represent a piece of text, an image, a tool call, or other structured data.
structured data.
Data **not yet mapped** to a standard block may be represented using the Data **not yet mapped** to a standard block may be represented using the
``NonStandardContentBlock``, which allows for provider-specific data to be included ``NonStandardContentBlock``, which allows for provider-specific data to be included
@ -58,8 +57,8 @@ that additional provider-specific fields are allowed.
Different LLM providers use distinct and incompatible API schemas. This module Different LLM providers use distinct and incompatible API schemas. This module
introduces a unified, provider-agnostic format to standardize these interactions. A introduces a unified, provider-agnostic format to standardize these interactions. A
message to or from a model is simply a `list` of `ContentBlock` objects, allowing for message to or from a model is simply a ``list`` of ``ContentBlock`` objects, allowing
the natural interleaving of text, images, and other content in a single, ordered for the natural interleaving of text, images, and other content in a single, ordered
sequence. sequence.
An adapter for a specific provider is responsible for translating this standard list of An adapter for a specific provider is responsible for translating this standard list of
@ -129,7 +128,7 @@ class Citation(TypedDict):
"""Annotation for citing data from a document. """Annotation for citing data from a document.
.. note:: .. note::
``start/end`` indices refer to the **response text**, ``start``/``end`` indices refer to the **response text**,
not the source text. This means that the indices are relative to the model's not the source text. This means that the indices are relative to the model's
response, not the original document (as specified in the ``url``). response, not the original document (as specified in the ``url``).
@ -157,7 +156,7 @@ class Citation(TypedDict):
# For future consideration, if needed: # For future consideration, if needed:
# provenance: NotRequired[str] # provenance: NotRequired[str]
# """Provenance of the document, e.g., "Wikipedia", "arXiv", etc. # """Provenance of the document, e.g., ``'Wikipedia'``, ``'arXiv'``, etc.
# Included for future compatibility; not currently implemented. # Included for future compatibility; not currently implemented.
# """ # """
@ -299,7 +298,7 @@ class ToolCall(TypedDict):
class ToolCallChunk(TypedDict): class ToolCallChunk(TypedDict):
"""A chunk of a tool call (e.g., as part of a stream). """A chunk of a tool call (e.g., as part of a stream).
When merging ToolCallChunks (e.g., via ``AIMessageChunk.__add__``), When merging ``ToolCallChunks`` (e.g., via ``AIMessageChunk.__add__``),
all string attributes are concatenated. Chunks are only merged if their all string attributes are concatenated. Chunks are only merged if their
values of ``index`` are equal and not ``None``. values of ``index`` are equal and not ``None``.
@ -793,7 +792,7 @@ class NonStandardContentBlock(TypedDict):
The purpose of this block should be to simply hold a provider-specific payload. The purpose of this block should be to simply hold a provider-specific payload.
If a provider's non-standard output includes reasoning and tool calls, it should be If a provider's non-standard output includes reasoning and tool calls, it should be
the adapter's job to parse that payload and emit the corresponding standard the adapter's job to parse that payload and emit the corresponding standard
ReasoningContentBlock and ToolCallContentBlocks. ``ReasoningContentBlock`` and ``ToolCallContentBlocks``.
.. note:: .. note::
``create_non_standard_block`` may also be used as a factory to create a ``create_non_standard_block`` may also be used as a factory to create a
@ -822,6 +821,7 @@ class NonStandardContentBlock(TypedDict):
# --- Aliases --- # --- Aliases ---
DataContentBlock = Union[ DataContentBlock = Union[
ImageContentBlock, ImageContentBlock,
VideoContentBlock, VideoContentBlock,
@ -832,6 +832,7 @@ DataContentBlock = Union[
ToolContentBlock = Union[ ToolContentBlock = Union[
ToolCall, ToolCall,
ToolCallChunk,
CodeInterpreterCall, CodeInterpreterCall,
CodeInterpreterOutput, CodeInterpreterOutput,
CodeInterpreterResult, CodeInterpreterResult,
@ -842,6 +843,9 @@ ToolContentBlock = Union[
ContentBlock = Union[ ContentBlock = Union[
TextContentBlock, TextContentBlock,
ToolCall, ToolCall,
ToolCallChunk,
Citation,
NonStandardAnnotation,
InvalidToolCall, InvalidToolCall,
ReasoningContentBlock, ReasoningContentBlock,
NonStandardContentBlock, NonStandardContentBlock,
@ -851,7 +855,7 @@ ContentBlock = Union[
def _extract_typedict_type_values(union_type: Any) -> set[str]: def _extract_typedict_type_values(union_type: Any) -> set[str]:
"""Extract the values of the 'type' field from a TypedDict union type.""" """Extract the values of the ``'type'`` field from a TypedDict union type."""
result: set[str] = set() result: set[str] = set()
for value in get_args(union_type): for value in get_args(union_type):
annotation = value.__annotations__["type"] annotation = value.__annotations__["type"]