github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2026-07-01 14:47:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs(core): document multimodal handling in get_buffer_string (#37994)

Browse Source 
Clarifies how `get_buffer_string` treats multimodal message content
across output formats. The docs now make the default prefix format's
text-only behavior explicit and point users to XML when they need
structured multimodal block representations.

This behavior may change in future iterations
This commit is contained in:
Mason Daugherty
2026-06-09 17:28:44 -04:00
committed by GitHub
parent 53f2ad18a0
commit ac18ef5871
1 changed files with 8 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
libs/core/langchain_core/messages/utils.py
View File

@@ -306,10 +306,15 @@ def get_buffer_string(
tool_prefix: The prefix to prepend to contents of `ToolMessage`s.
message_separator: The separator to use between messages.
format: The output format. `'prefix'` uses `Role: content` format (default).
For multimodal messages, only string content and `text` blocks are
included; non-text blocks such as images, audio, and video are omitted.
`'xml'` uses XML-style `<message type='role'>` format with proper character
escaping, which is useful when message content may contain role-like
prefixes that could cause ambiguity.
`'xml'` uses XML-style `<message type='role'>` format with proper
character escaping, which is useful when message content may
contain role-like prefixes that could cause ambiguity.
Use `'xml'` when you need a structured representation of supported
multimodal content blocks.
Returns:
A single string concatenation of all input messages.