core: basemessage.text()

[efriis]

No description provided.

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

1 Skipped Deployment

Name	Status	Preview	Comments	Updated (UTC)
langchain	⬜️ Ignored (Inspect)	Visit Preview		Feb 18, 2025 7:48pm

[ccurme]

Looks good. Nicer than having to use StrOutputParser. For streaming this will work with Anthropic but haven't looked at any other providers.

How it works for StrOutputParser:

langchain/libs/core/langchain_core/outputs/chat_generation.py

Lines 50 to 63 in c5bee0a

	if isinstance(self.message.content, str):
	text = self.message.content
	# HACK: Assumes text in content blocks in OpenAI format.
	# Uses first text block.
	elif isinstance(self.message.content, list):
	for block in self.message.content:
	if isinstance(block, str):
	text = block
	break
	elif isinstance(block, dict) and "text" in block:
	text = block["text"]
	break
	else:
	pass

[ccurme]

Would be good to document a streaming example, e.g.,

for chunk in llm.stream(...):
    print(chunk.text(), end="|")

Should target being able to replace this page: https://python.langchain.com/docs/how_to/output_parser_string/