⚡ Bolt: Optimize htmlToText parsing to zero-allocation manual loop - #180
⚡ Bolt: Optimize htmlToText parsing to zero-allocation manual loop#180MEKXH wants to merge 1 commit into
Conversation
Replace multi-pass `regexp.MustCompile` and `ReplaceAllString` calls with a single-pass `strings.Builder` byte loop and `strings.IndexByte`. This eliminates three expensive regular expression initializations and drastically reduces memory allocations and parsing time per webpage fetched. Co-authored-by: MEKXH <59291264+MEKXH@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
💡 What:
Replaced the
htmlToTextfunction's underlying logic. Instead of chaining three separateregexp.MustCompileandReplaceAllStringoperations to strip out<script>,<style>, and general HTML tags, it now uses a single-passstrings.Builderbyte loop. It checks for specific tags using exact slice matching (e.g.,s[i:i+7] == "<script") and skips over tag bodies usingstrings.IndexByteand a custom zero-allocationindexIgnoreCasehelper function. The unused regex definitions were completely removed.🎯 Why:
The previous implementation performed three complete passes over the input HTML string. Every regex replacement required intermediate string allocations, resulting in high memory usage and GC pressure during web searching and fetching, particularly for large HTML documents.
📊 Impact:
The
BenchmarkHtmlToTextbenchmark shows a massive performance improvement. Execution time for the test HTML blob dropped from ~15025 ns/op to ~2444 ns/op (a ~6.14x speedup). Additionally, the manual string builder drops intermediate string allocations caused byReplaceAllStringto almost zero.🔬 Measurement:
The performance can be verified by running the ad-hoc benchmark provided during implementation or simply measuring execution times for
htmlToTexton large HTML payloads. All unit and integration tests continue to pass identically (make test && make lint).PR created automatically by Jules for task 18221297299795523237 started by @MEKXH