From 5080f81770a593aff4451625bd025777fb7fe7d8 Mon Sep 17 00:00:00 2001
From: Shizuo Fujita <fujita@clear-code.com>
Date: Wed, 8 Jan 2025 14:29:18 +0900
Subject: [PATCH] in_tail: fix slice length to reduce memory usage (#4760)

**Which issue(s) this PR fixes**:
Fixes #

**What this PR does / why we need it**:

`String#slice` will allocate redundant heap area internally if an
incorrect length is specified.
This PR might reduce the memory usage with `in_tail` plugin.


* verify code
```ruby
require 'bundler/inline'
gemfile do
  source 'https://rubygems.org'
  gem 'benchmark-memory'
end

str = "a" * 10_000_000
split_point = 9_999_900

Benchmark.memory do |x|
  x.report("1") {
    str.slice(split_point, str.length)
  }

  x.report("2") {
    str.slice(split_point, str.length - split_point)
  }

  x.report("3") {
    str.slice(split_point..str.length)
  }
end
```

* result
```
Calculating -------------------------------------
                   1    10.000M memsize (    10.000M retained)
                         2.000  objects (     1.000  retained)
                         2.000  strings (     1.000  retained)
                   2    40.000  memsize (     0.000  retained)
                         1.000  objects (     0.000  retained)
                         1.000  strings (     0.000  retained)
                   3    80.000  memsize (     0.000  retained)
                         2.000  objects (     0.000  retained)
                         1.000  strings (     0.000  retained)
```

The all cases will generate same String object. However, `1` case will
use 10 MB heap area with above code.

**Docs Changes**:

**Release Note**:

Signed-off-by: Shizuo Fujita <fujita@clear-code.com>
---
 lib/fluent/plugin/in_tail.rb | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/lib/fluent/plugin/in_tail.rb b/lib/fluent/plugin/in_tail.rb
index 852f2723bd..703d9d01c1 100644
--- a/lib/fluent/plugin/in_tail.rb
+++ b/lib/fluent/plugin/in_tail.rb
@@ -1056,8 +1056,9 @@ def read_lines(lines)
             # Using freeze and slice is faster than slice!
             # See https://github.com/fluent/fluentd/pull/2527
             @buffer.freeze
-            rbuf = @buffer.slice(0, idx + 1)
-            @buffer = @buffer.slice(idx + 1, @buffer.size)
+            slice_position = idx + 1
+            rbuf = @buffer.slice(0, slice_position)
+            @buffer = @buffer.slice(slice_position, @buffer.size - slice_position)
             idx = @buffer.index(@eol)
 
             is_long_line = @max_line_size && (