Why is this Elixir code so slow? -

i'm interested in learning elixir , typical approach of learning new language build simple programs it.

so decided write (very simple) grep-like program (or module) shown here:

defmodule linematch   def file(s, path)     case file.open(path, [:read])       {:ok, fd} -> match s, fd       {_, error} -> io.puts "#{:file.format_error error}"     end   end   defp match(s, fd)     case io.read fd, :line       {:error, error} -> io.puts("oh noez! error: #{error}")       line -> match(s, line, fd)     end   end   defp match(s, line, fd) when line !== :eof     if string.contains?(line, s)       io.write line     end     match(s, io.read(fd, :line), fd)   end   defp match(_, _line, _) when _line === :eof   end end 

this not elegant way , i'm quite new functional programming, didn't expect super fast. not not fast, super slow. slow did super obvious mistake.

can tell me, , how make better?

i typically test code seperate .exs file like

case system.argv   [searchterm, path] -> linematch.file(searchterm, path)   _ -> io.puts "usage: linematch searchterm path" end 

rather reading in whole file in lad2025's answer, can performance doing 2 things. first, use io.binstream build stream of file's lines, raw binary (for performance). using io.stream reads utf-8, , such incurs cost conversion reading file. if need utf-8 conversion, it's going slow. additionally, applying filtering , mapping operations using stream.filter/2 , stream.map/2 functions prevents iterating on lines multiple times.

defmodule grepper   def grep(path, str)     case file.open(path)       {:error, reason} -> io.puts "error grepping #{path}: #{reason}"       {:ok, file} ->         io.binstream(file, :line)         |> stream.filter(&(string.contains?(&1, str)))         |> stream.map(&(io.puts(io.ansi.green <> &1 <> io.ansi.reset)))         |> stream.run     end   end end 

i suspect greatest issue code utf-8 conversion, may "pulling" file line line calling io.read, rather having lines "pushed" filtering/printing operations using io.stream|binstream, incurring cost there. i'd have @ elixir's source know sure, above code quite performant on machine (i searching 143kb file olson timezone database, not sure how perform on files of large size don't have sample file handy).