Skip to content

v0.31.0 — forest memory containment and Python parity

Choose a tag to compare

View all tags
@odvcencio odvcencio released this 13 Jul 07:31
v0.31.0
7b31007
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Highlights

v0.31.0 tightens memory containment for large ambiguous parses, restores
Python real-corpus parity, and strengthens authenticated fleet reporting.

  • Failed forest attempts now honor the parser's runtime heap and system-memory
    guard. On the default-budget JavaScript Poppler witness, fallback elapsed
    time fell from 5.313 s to 2.610 s, allocation from 2.450 GB to 1.289 GB, and
    maximum RSS from 1,961,948 KiB to 1,012,296 KiB with identical stopped-tree
    hashes.
  • Poppler still does not complete under the ordinary 512 MiB production
    policy. This release bounds the failed forest attempt; it does not claim
    ordinary-budget completion.
  • Discarded pooled forest GSS batches no longer retain hidden slab tails. The
    Poppler witness's one-GC live heap fell from 1,475,142,360 to 608,167,688
    bytes while preserving identical output.
  • Removing a misfiring Python compatibility fold restores real-corpus
    S-expression and deep parity from 20/25 to 25/25.
  • Fleet reducers can publish authenticated PASS or FAIL boards while the
    certification mode remains blocking on valid failure evidence.
  • The tier-scan guide now accurately marks full-corpus tier publication as
    staged; stale tier artifacts remain explicitly unreleased.

Successful-forest timing remained neutral (+1.4%, p=0.142), with a small
measured allocation cost (+0.28% B/op and +0.01% allocs/op).

Install with:

go get github.com/odvcencio/gotreesitter@v0.31.0

See the v0.31.0 changelog for the complete change list.

Assets 2
Loading