fix: update langchain imports for v1.0+ compatibility

[majiayu000]

Summary

• Update ResponseSchema and StructuredOutputParser imports to use langchain_classic.output_parsers
• Update create_extraction_chain import to use langchain_classic.chains
• Add langchain-classic>=1.0.0 as explicit dependency
• Relax async-timeout requirement to >=4.0.0 for compatibility
• Fix variable name mismatch in generate_answer_node_k_level.py: use content instead of context

Fixes #1017
Fixes #995

Changes

File	Change
scrapegraphai/nodes/generate_code_node.py	Import from langchain_classic.output_parsers
scrapegraphai/builders/graph_builder.py	Import from langchain_classic.chains
scrapegraphai/nodes/generate_answer_node_k_level.py	Use content variable instead of context
pyproject.toml	Add langchain-classic dependency, relax async-timeout

Test plan

• Verified imports work correctly with Python 3.14
• Ran ruff, black, isort checks
• All modified modules import successfully

[Copilot]

Pull request overview

This PR updates the codebase to be compatible with LangChain v1.0+ by migrating deprecated imports to langchain-classic and fixing a variable name mismatch that was causing runtime errors.

Key Changes:

• Migrated deprecated LangChain imports (ResponseSchema, StructuredOutputParser, create_extraction_chain) to use langchain-classic package
• Fixed variable name inconsistency in generate_answer_node_k_level.py where template expects content but code was passing context
• Added langchain-classic>=1.0.0 as explicit dependency and relaxed async-timeout version constraint for better compatibility

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
scrapegraphai/nodes/generate_code_node.py	Updated imports to use langchain_classic.output_parsers for ResponseSchema and StructuredOutputParser
scrapegraphai/builders/graph_builder.py	Updated import to use langchain_classic.chains for create_extraction_chain; minor formatting improvement
scrapegraphai/nodes/generate_answer_node_k_level.py	Fixed variable name from context to content in three locations to match template expectations
pyproject.toml	Added langchain-classic>=1.0.0 dependency and relaxed async-timeout requirement from >=5.0.1 to >=4.0.0

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[VinciGit00]

Review Summary

Great work on this PR! The changes correctly address LangChain v1.0+ compatibility issues. Here's my assessment:

✅ Approved Changes

1. Import Updates: All import changes from langchain.chains and langchain_core.output_parsers to langchain_classic equivalents are correct and necessary for LangChain v1.0+ compatibility.

2. Bug Fix: The variable name fix in generate_answer_node_k_level.py (changing context to content) is correct. I verified that all templates use {content}, not {context}, so this was indeed a bug that needed fixing.

3. Dependency Management: Adding langchain-classic>=1.0.0 is appropriate. The relaxation of async-timeout version seems reasonable, though worth verifying it doesn't break anything.

4. Code Quality: The formatting improvements are welcome.

📝 Testing Notes

I attempted to test the imports but encountered dependency conflicts in the local environment (expected in a mixed environment). However, the changes are syntactically correct and follow the official LangChain migration path for v1.0+.

✅ Recommendation: Approve

This PR correctly addresses the compatibility issues mentioned in #1017 and #995. The changes are minimal, focused, and follow best practices for LangChain v1.0+ migration.

Ready to merge! 🚀

[github-actions]

🎉 This PR is included in version 1.71.0 🎉

The release is available on:

• v1.71.0
• GitHub release

Your semantic-release bot 📦🚀