heap-profiler 0.3.0 → 0.4.0

data/lib/heap_profiler/analyzer.rb

@@ -11,9 +11,6 @@ module HeapProfiler
 
       def process(_index, object)
         @objects += 1
-        unless object[:memsize]
-          p object
-        end
         @memory += object[:memsize]
       end
 

data/lib/heap_profiler/cli.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+require 'optparse'
 
 module HeapProfiler
   class CLI
@@ -7,13 +8,21 @@ module HeapProfiler
     end
 
     def run
-      if @argv.size == 1
-        print_report(@argv.first)
-        0
-      else
-        print_usage
-        1
+      parser.parse!(@argv)
+
+      begin
+        if @argv.size == 1
+          print_report(@argv.first)
+          return 0
+        end
+      rescue CapacityError => error
+        STDERR.puts(error.message)
+        STDERR.puts("Current size: #{Parser.batch_size}B")
+        STDERR.puts("Try increasing it with --batch-size")
+        STDERR.puts
       end
+      print_usage
+      1
     end
 
     def print_report(path)
@@ -28,5 +37,51 @@ module HeapProfiler
     def print_usage
       puts "Usage: #{$PROGRAM_NAME} directory_or_heap_dump"
     end
+
+    SIZE_UNITS = {
+      'B' => 1,
+      'K' => 1_000,
+      'M' => 1_000_000,
+      'G' => 1_000_000_000,
+    }
+    def parse_byte_size(size_string)
+      if (match = size_string.match(/\A(\d+)(\w)?B?\z/i))
+        digits = Float(match[1])
+        base = 1
+        unit = match[2]&.upcase
+        if unit
+          base = SIZE_UNITS.fetch(unit) { raise ArgumentError, "Unknown size unit: #{unit}" }
+        end
+        size = (digits * base).to_i
+        if size > 4_000_000_000
+          raise ArgumentError, "Batch size can't be bigger than 4G"
+        end
+        size
+      else
+        raise ArgumentError, "#{size_string} is not a valid size"
+      end
+    end
+
+    private
+
+    def parser
+      @parser ||= OptionParser.new do |opts|
+        opts.banner = "Usage: heap-profiler [ARGS]"
+        opts.separator ""
+        opts.separator "GLOBAL OPTIONS"
+        opts.separator ""
+
+        help = <<~EOS
+          Sets the simdjson parser batch size. It must be larger than the largest JSON document in the
+          heap dump, and defaults to 10MB.
+        EOS
+        opts.on('--batch-size SIZE', help.strip) do |size_string|
+          HeapProfiler::Parser.batch_size = parse_byte_size(size_string)
+        rescue ArgumentError => error
+          STDERR.puts "Invalid batch-size: #{error.message}"
+          exit 1
+        end
+      end
+    end
   end
 end

data/lib/heap_profiler/dump.rb

@@ -62,7 +62,7 @@ module HeapProfiler
     end
 
     def each_object(since: nil, &block)
-      Parser.load_many(path, since: since, batch_size: 10_000_000, &block)
+      Parser.load_many(path, since: since, &block)
     end
 
     def stats

data/lib/heap_profiler/parser.rb

@@ -4,6 +4,11 @@ module HeapProfiler
   module Parser
     CLASS_DEFAULT_PROC = ->(_hash, key) { "<Class#0x#{key.to_s(16)}>" }
 
+    class << self
+      attr_accessor :batch_size
+    end
+    self.batch_size = 10_000_000 # 10MB
+
     class Ruby
       def build_index(path)
         require 'json'
@@ -42,15 +47,13 @@ module HeapProfiler
     end
 
     class Native
-      DEFAULT_BATCH_SIZE = 10_000_000 # 10MB
-
-      def build_index(path, batch_size: DEFAULT_BATCH_SIZE)
+      def build_index(path, batch_size: Parser.batch_size)
         indexes = _build_index(path, batch_size)
         indexes.first.default_proc = CLASS_DEFAULT_PROC
         indexes
       end
 
-      def load_many(path, since: nil, batch_size: DEFAULT_BATCH_SIZE, &block)
+      def load_many(path, since: nil, batch_size: Parser.batch_size, &block)
         _load_many(path, since, batch_size, &block)
       end
     end

data/lib/heap_profiler/runtime.rb

@@ -8,6 +8,7 @@ require "heap_profiler/reporter"
 
 module HeapProfiler
   Error = Class.new(StandardError)
+  CapacityError = Class.new(Error)
 
   class << self
     attr_accessor :current_reporter

data/lib/heap_profiler/version.rb

@@ -1,4 +1,4 @@
 # frozen_string_literal: true
 module HeapProfiler
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: heap-profiler
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors:
 - Jean Boussier
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2021-01-26 00:00:00.000000000 Z
+date: 2021-01-30 00:00:00.000000000 Z
 dependencies: []
 description: Make several heap dumps and summarize allocated, retained memory
 email:
@@ -79,7 +79,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.4
+rubygems_version: 3.2.5
 signing_key:
 specification_version: 4
 summary: Ruby heap profiling tool