<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Chenyu Hu</title>
    <description>The latest articles on DEV Community by Chenyu Hu (@huchenyu).</description>
    <link>https://dev.to/huchenyu</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4248919%2F85a47baf-5215-4291-8c90-3d90f1b0998d.png</url>
      <title>DEV Community: Chenyu Hu</title>
      <link>https://dev.to/huchenyu</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/huchenyu"/>
    <language>en</language>
    <item>
      <title>Looking for reproducers: a local log analyzer where every finding carries a line number and a sha256</title>
      <dc:creator>Chenyu Hu</dc:creator>
      <pubDate>Tue, 17 Nov 2026 05:31:48 +0000</pubDate>
      <link>https://dev.to/huchenyu/looking-for-reproducers-a-local-log-analyzer-where-every-finding-carries-a-line-number-and-a-sha256-348b</link>
      <guid>https://dev.to/huchenyu/looking-for-reproducers-a-local-log-analyzer-where-every-finding-carries-a-line-number-and-a-sha256-348b</guid>
      <description>&lt;p&gt;Most log-analysis tools answer one question: "what broke?" They print a sentence. You have no way to check whether that sentence is right, short of re-reading the raw log yourself.&lt;/p&gt;

&lt;p&gt;&lt;code&gt;log-ai-compressor&lt;/code&gt; does the same job, but every finding it prints is attached to a &lt;strong&gt;line range in your original file plus a sha256 of that range&lt;/strong&gt;. You can go back to those lines, and you can run the built-in &lt;code&gt;verify&lt;/code&gt; to check that a finding was not altered after it was produced. It runs entirely locally: no log upload, no API calls, no LLM keys.&lt;/p&gt;

&lt;h2&gt;
  
  
  Two commands, about 30 seconds, no dataset download
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;pip install -U --index-url https://pypi.org/simple &lt;span class="s2"&gt;"log-ai-compressor==2.16.1"&lt;/span&gt;
log-ai-compressor run examples/app.log -f summary
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;examples/app.log&lt;/code&gt; is 24 lines and ships with the package. On my machine the summary is about 234 tokens (a 38% reduction against the raw file), with 8 error lines grouped into 4 signatures, and a fingerprint &lt;code&gt;48e5781a...c2886&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I can claim, and what I cannot
&lt;/h2&gt;

&lt;p&gt;Full run over all ten Loghub-2.0 datasets, 1,940,760 lines, 1,621 template clusters:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;PA 85.05%, GA 88.68%&lt;/li&gt;
&lt;li&gt;Same data, Drain3: PA 58.06%, GA 71.58%&lt;/li&gt;
&lt;li&gt;CPU only, zero API spend&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Three limits I will not paper over:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Reproducible is not the same as state of the art. There is no third-party leaderboard behind these numbers.&lt;/li&gt;
&lt;li&gt;The LLM comparison arm is a 2,000-line sample per dataset, not the full set.&lt;/li&gt;
&lt;li&gt;The HDFS block-level causal chain cannot be re-run here: the original 1.5 GB log is not on my disk, so that 37.17% recall number is documented, not re-executable.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Why I am posting this
&lt;/h2&gt;

&lt;p&gt;The repo currently has 0 stars, 0 forks and 1 issue. The only external verification so far is another AI agent that followed the README on a different machine and matched the fingerprint. That is one run, not a public benchmark.&lt;/p&gt;

&lt;p&gt;So the ask is small: &lt;strong&gt;run the two commands on your machine and reply either way&lt;/strong&gt;. Failures are more useful than passes. Three things are enough:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;OS and Python version&lt;/li&gt;
&lt;li&gt;the exit code of the second command&lt;/li&gt;
&lt;li&gt;if it failed, the last 20 lines of output&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Links
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Repo: &lt;a href="https://gitee.com/hu-chenyu-dev/log-ai-compressor" rel="noopener noreferrer"&gt;https://gitee.com/hu-chenyu-dev/log-ai-compressor&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Release v2.16.1: &lt;a href="https://gitee.com/hu-chenyu-dev/log-ai-compressor/releases/tag/v2.16.1" rel="noopener noreferrer"&gt;https://gitee.com/hu-chenyu-dev/log-ai-compressor/releases/tag/v2.16.1&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Reproduction notes and the reply thread: &lt;a href="https://gitee.com/hu-chenyu-dev/log-ai-compressor/issues/IKVUSO" rel="noopener noreferrer"&gt;https://gitee.com/hu-chenyu-dev/log-ai-compressor/issues/IKVUSO&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Package: &lt;a href="https://pypi.org/project/log-ai-compressor/" rel="noopener noreferrer"&gt;https://pypi.org/project/log-ai-compressor/&lt;/a&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Note: the repo is hosted on Gitee, which is usable but not fast from outside China. PyPI is the reliable entry point for installing it.&lt;/p&gt;

</description>
      <category>python</category>
      <category>discuss</category>
      <category>opensource</category>
    </item>
  </channel>
</rss>
