Back

langchain #34805

chore(text-splitters): overlap threshold introduced in `TextSplitter`

by keenborder786 · Feb 06, 2026 at 19:30 UTC · scan-1296b705ce312cbd

Low Risk (20%)

Get this automatically on every PR

Install the Axiomo GitHub App to get Signals as check runs and PR comments on every pull request.

Install App

Risk Assessment

Risk level: Low (20%)

Risk Drivers

  • api_surface_change: API surface changed in 1 file(s)

Intent

3/3 criteria met

Introduce overlap threshold in TextSplitter to apply overlapping for smaller chunks.

Acceptance Criteria

  • โœ“
    Ensure overlapping is applied even for chunks not exceeding size limit.

    base.py: New overlap_threshold parameter enables overlap for small chunks.

  • โœ“
    Maintain backward compatibility.

    base.py: overlap_threshold parameter is optional to ensure compatibility.

  • โœ“
    Tests validate the new functionality.

    test_text_splitters.py: Tests added for overlap_threshold validation.

Confidence: 95.0% Source: pr description AI: openai

Contributors

keenborder786 PR Author 12 commits ~ Established
Account Age: 2634 days
Prior PRs: 163
Merged: 89

Established contributor with 89 merged PRs. maintains 101 public repositories. unfamiliar with 2 files.

Evidence

Evidence Completeness: 80.0%
ci_passing Passing
tests_passing Passing
build_successful Passing
lint_passing Passing
Missing: security_scan_clean, coverage_maintained

Supply Chain

None Risk
Modifies dependencies
Modifies lockfile
Modifies CI config
Modifies build scripts

Focus Files

Review 1 high-priority file(s)

libs/text-splitters/tests/unit_tests/test_text_splitters.py +107

107 lines changed; Source code

high
libs/text-splitters/langchain_text_splitters/base.py +35

Source code

medium

Triage

23

minutes to review

medium

effort level

none

staleness risk

Standard review process

Recommendation

APPROVE 76.0% readiness

Good evidence, acceptable risk, ready for merge

Next Steps

Question

Why is security_scan_clean missing? Consider adding this check.

Question

Why is coverage_maintained missing? Consider adding this check.