Tuesday, December 12, 2017

Haskell to MySQL via YeshQL (Part 2.)

In the previous blog post, we built the skeleton of a console app in Haskell that talks to MySQL via YeshQL. We used a Makefile to rebuild the database, compile the code and run the app. This is the commit point where we left it at the end of Part 1.

In this article we will add functionality to create clients records, we will extract the logic that acquires, commits and closes the connection, and finally, we will move the database related code into its own module.

Insert a Client record

We can query the clients table, however, there are no records in that table. Let's add one script to the SQL template.

Modify the YeshQL templates by adding the following code:

 [yesh|
     -- name:countClientSQL :: (Int)
     SELECT count(id) FROM clients;
     ;;;
     -- name:insertClientSQL
     -- :client_name :: String
     -- :subdomain :: String
     INSERT INTO clients (name, subdomain) VALUES (:client_name, :subdomain);
 |]

Three semicolons are used to separate SQL statements. The -- name is used to provide a function name that we can refer to in our code. The two lines are the input arguments to the generated function. Those arguments are the values in the INSERT statement on the last line.

Let's add the following Haskell function to insert a new Client record:

insertClient :: String -> String -> IO ()
insertClient name subdomain = do
    conn <- getConn
    clientId <- insertClientSQL name subdomain conn
    commit conn
    disconnect conn
    putStrLn $ "New client's id is " ++ show clientId

The generated insertClientSQL (note the name in the SQL template) is invoked with the two specified arguments plus the connection. The uncommitted connection will not write the record to the table, without that statement the countClientSQL function would return 0 records.
Disconnecting a connection is a good practice (if you can't pool those connections), free up resources when you don't need them.

Invoke the created insertClient function from main function like this:

main = do
    insertClient "TestClient" "testclient"
    countClient

When I try to build the app (use make build for it), I get the following errors:

/Haskell/hashmir/app/Main.hs:34:5: error:
    Variable not in scope: commit :: Connection -> IO a0

It turns out that Database.HDBC library is needed to invoke commit and disconnect.
Add the following import statement to the top of the file:

import qualified Database.HDBC as H

Update the function to use the qualified names for the commit and disconnect calls.

insertClient :: String -> String -> IO ()
insertClient name subdomain = do
    conn <- getConn
    clientId <- insertClientSQL name subdomain conn
    H.commit conn
    H.disconnect conn
    putStrLn $ "New client's id is " ++ show clientId

The project should successfully build, when you run the app (use make run to do it), you should see this output:

% make run
Dropping and rebuilding database hashmir_test
time ~/.local/bin/hashmir-exe
New client's id is 1
There are 1 records.

real 0m0.020s
user 0m0.009s
sys 0m0.006s

The database got rebuilt, one client record was inserted and the count function counted that record.

Commit point

Introducing: withConn

Committing and disconnecting a connection is generally a good practice. Let's match the countClient function with the insertClient function and call commit and disconnect the connection there as well.

countClient :: IO ()
countClient = do
    conn <- getConn
    Just (clientCount) <- countClientSQL conn
    H.commit conn -- added line
    H.disconnect conn -- added line
    putStrLn $ "There are " ++ show clientCount ++ " records."

Now both the countClient and the insertClient has the same duplicated logic:

    conn <- getConn
    ...
    H.commit conn -- added line
    H.disconnect conn -- added line

This reminds me of the use of withFile from the IO module. withFile accepts a lambda where the handle is passed to it and the code in the lambda can use the provided handle. We need the same thing here, withConn would accept an active connection. Consider this function:

withConn :: (Connection -> IO b) -> IO b
withConn f = do
    conn <- getConn
    result <- f conn
    H.commit conn
    H.disconnect conn
    return result

Our refactored insertClient function would look like this:

insertClient :: String -> String -> IO ()
insertClient name subdomain = do
    clientId <-
        withConn (\conn -> do
            insertClientSQL name subdomain conn
        )
    putStrLn $ "New client's id is " ++ show clientId

When you build the project and run it, it should work without errors.

Commit point

Thanks to Haskell's currying, this function can be further simplified. No need to provide the input argument in the lambda:

insertClient :: String -> String -> IO ()
insertClient name subdomain = do
    clientId <-
        withConn (do insertClientSQL name subdomain)
    putStrLn $ "New client's id is " ++ show clientId

This looks much better, but we can further simplify this code:

insertClient :: String -> String -> IO ()
insertClient name subdomain = do
    clientId <- withConn $ do insertClientSQL name subdomain
    putStrLn $ "New client's id is " ++ show clientId

Now that function is easy to read!

Commit point

Let's refactor the countClient function similarly.

countClient :: IO ()
countClient = do
    Just (clientCount) <- withConn countClientSQL
    putStrLn $ "There are " ++ show clientCount ++ " records."

Commit point

Extract Data Access Logic

Having code that prints information on the screen from the functions that talks with the database makes them mixed with responsibilities. They should only return primitives, and the caller main function should print the reports. Let's make those functions a bit more clean:

insertClient :: String -> String -> IO Integer
insertClient name subdomain = do
    withConn $ do insertClientSQL name subdomain

countClient :: IO (Maybe Int)
countClient = do withConn countClientSQL

I really like how Haskell makes the functions that uses IO impure or dirty: once they have been tainted, they are tainted. You should always try to isolate functions that are tainted from the pure ones as the pure ones are easier test - they have no side effects.

The main function is now responsible for reporting the result:

main :: IO ()
main = do
    clientId <- insertClient "TestClient" "testclient"
    putStrLn $ "New client's id is " ++ show clientId
    Just clientCount <- countClient
    putStrLn $ "There are " ++ show clientCount ++ " records."

Commit point

Move Data Logic Into a Library

Our Main.hs file has all the application logic. It looks pretty solid, let's move it into a library module. Copy all database access related code from app/Main.hs into src/Hashmir/Data.hs file like this:

{-#LANGUAGE TemplateHaskell #-}
{-#LANGUAGE QuasiQuotes #-}

module Hashmir.Data where

import Database.YeshQL
import qualified Database.HDBC as H
import Database.HDBC.MySQL

[yesh|
    -- name:countClientSQL :: (Int)
    SELECT count(id) FROM clients;
    ;;;
    -- name:insertClientSQL
    -- :client_name :: String
    -- :subdomain :: String
    INSERT INTO clients (name, subdomain) VALUES (:client_name, :subdomain);
|]

getConn :: IO Connection
getConn = do
    connectMySQL defaultMySQLConnectInfo {
        mysqlHost     = "localhost",
        mysqlDatabase = "hashmir_test",
        mysqlUser     = "hashmir_user",
        mysqlPassword = "shei7AnganeihaeF",
        mysqlUnixSocket = "/tmp/mysql.sock"
    }

withConn :: (Connection -> IO b) -> IO b
withConn f = do
    conn <- getConn
    result <- f conn
    H.commit conn
    H.disconnect conn
    return result

insertClient :: String -> String -> IO Integer
insertClient name subdomain =
    withConn $ insertClientSQL name subdomain

countClient :: IO (Maybe Int)
countClient = withConn countClientSQL

This is the same logic we had in the app/Main.hs file, but now it is in the Hashmir.Data module.

The Main module becomes small once we remove all the code we just moved out of it:

module Main where

import qualified Hashmir.Data as D

main :: IO ()
main = do
    clientId <- D.insertClient "TestClient" "testclient"
    putStrLn $ "New client's id is " ++ show clientId
    Just clientCount <- D.countClient
    putStrLn $ "There are " ++ show clientCount ++ " records."

We also have to tell Cabal where it can find this code. Change the Lib directive to Hashmir.Data in the package.yaml:

library:
  source-dirs: src/
  exposed-modules:
    - Hashmir.Data

The project should build and when you run the app it should insert and return the number of clients records:

% make run
Dropping and rebuilding database hashmir_test
time ~/.local/bin/hashmir-exe
New client's id is 1
There are 1 records.

real 0m0.022s
user 0m0.010s
sys 0m0.007s

Commit point

This last change wraps up our Part 2 in this series. We can now create clients records, count them with a simple withConn function that properly opens, commits and closes the connection.

In the third post in this series, I will show you how we can insert two records in one transaction, how we can deal with errors and how this logic can be tested.

Thursday, November 30, 2017

Haskell to MySQL via YeshQL (Part 1.)

As I was looking for an easy way to talk to Postgres from Clojure, I discovered Yesql. I wanted to find something similar in Haskell, and I found YeshQL. It's a template parsing library on top of HDBC, exactly what I needed to keep SQL and Haskell code separate.

This blog post will show you how you can easily get YeshQL up and running and run queries against MySQL. I'll build a simple CRUD console app that you can follow along, I'll list commit points at the end of each section.

The Clojure tutorial I created for my blog posts is named Kashmir, I'll name this project Hashmir.

You will need stack and ghc installed, I have stack Version 1.5.1 x86_64 hpack-0.17.1 and ghc version 8.0.2. MySQL is also needed, I have version 5.7.20, but I won't use anything fancy as far as the database goes, if you have MySQL installed, I am sure that will do it.

Generate the project with stack

Generate the project with this command: stack new hashmir. Go inside the directory, build and deploy the app with stack build && stack install. Run the skeleton app:

% ~/.local/bin/hashmir-exe
someFunc

Commit point

Using hpack

Since I learned about hpack, I never touch a Haskell project's cabal file any more. This blog post assumes you are familiar with this tool, feel free to learn about hpack more before you proceed.

Add this package.yaml to the project's root directory:

name: hashmir
version: 0.1.0.0
author: Attila Domokos <adomokos@gmail.com>
maintainer: adomokos@gmail.com
copyright: 2017 Attila Domokos
category: Console App
homepage: https://github.com/adomokos/hashmir#readme

ghc-options: -Wall

dependencies:
  - base >= 4.7 && < 5

library:
  source-dirs: src/
  exposed-modules:
    - Lib

executables:
  hashmir-exe:
    source-dirs: app/
    main: Main.hs
    dependencies:
      hashmir
    ghc-options: -threaded -rtsopts -with-rtsopts=-N

Feel free to use your name and Github repo in this file.
Delete and regenerate the project's cabal file with this command: rm -f hashmir.cabal && stack build. stack install should produce the same executable file.

Commit point

Setting up the Database

We will need a MySQL user and a database we can use in this project. Let's add a Makefile to script that out for us. You might need the master or root user in MySQL to create a role and grant access. If the Makefile target does not work, just log on to the MySQL console with your root account and add the role and grant access with the scripts you see in the Makefile.

This is the schema file we will work with. I won't include it in this post, but this script should drop and recreate the tables. Put it into the ./resources directory.

These targets in the Makefile will create the role and will rebuild the database:

create-db-user: ## Creates a DB user with the root MySQL user
    mysql -u root --host $(HOST) -e "CREATE USER '$(DBUSER)'@'$(HOST)' IDENTIFIED BY '$(DBPASSWD)';" > /dev/null 2>&1
    mysql -u root --host $(HOST) -e "GRANT ALL PRIVILEGES ON `$(DBNAME)`.* TO '$(DBUSER)'@'$(HOST)';" > /dev/null 2>&1

build-db: ## Builds the DB
    @echo "Dropping and rebuilding database $(DBNAME)"
    @mysql -u $(DBUSER) --password='$(DBPASSWD)' --host $(HOST) -e "DROP DATABASE IF EXISTS $(DBNAME);" > /dev/null 2>&1
    @mysql -u $(DBUSER) --password='$(DBPASSWD)' --host $(HOST) -e "CREATE DATABASE $(DBNAME);" > /dev/null 2>&1
    @mysql -u $(DBUSER) --password='$(DBPASSWD)' --host $(HOST) $(DBNAME) < resources/schema.sql > /dev/null 2>&1

You should be able to execute make build to rebuild the app, and make run to rebuild the DB and run the app:

% make run
Dropping and rebuilding database crud_yeshql_test
time ~/.local/bin/hashmir-exe
someFunc

real    0m0.011s
user    0m0.002s
sys 0m0.007s

Commit point

Writing the First Query

There are two parts of using YeshQL's code:

  1. The SQL templates
  2. Code that uses the generated functions from the template

Modify the app/Main.hs file like this:

{-#LANGUAGE TemplateHaskell #-}
{-#LANGUAGE QuasiQuotes #-}

module Main where

import Database.YeshQL
import Database.HDBC.MySQL

[yesh|
    -- name:countClientSQL :: (Int)
    SELECT count(id) FROM clients;
|]

getConn :: IO Connection
getConn = do
    connectMySQL defaultMySQLConnectInfo {
        mysqlHost     = "localhost",
        mysqlDatabase = "hashmir_test",
        mysqlUser     = "hashmir_user",
        mysqlPassword = "shei7AnganeihaeF",
        mysqlUnixSocket = "/tmp/mysql.sock"
    }

countClient :: IO ()
countClient = do
    conn <- getConn
    Just (clientCount) <- countClientSQL conn
    putStrLn $ "There are " ++ show clientCount ++ " records."

main :: IO ()
main = countClient

When you try to build the project (I conveniently use make build) one of these errors are displayed:

Failed to load interface for 'Database.YeshQL'

I started referencing Database.YeshQL, however, I did not add that library to the project. This is where hpack is helpful, we only have to add it to the package.yaml file, that will generate a cabal file with the correct references.

Let's modify the dependencies section of package.yaml file like this:

...

dependencies:
  - base >= 4.7 && < 5
  - yeshql
  - HDBC
  - HDBC-mysql

...

When I try to build the project, I get the following error:

Error: While constructing the build plan, the following exceptions were encountered:

In the dependencies for hashmir-0.1.0.0:
    yeshql must match -any, but the stack configuration has no specified version (latest applicable is 3.0.1.3)
needed since hashmir-0.1.0.0 is a build target.

Recommended action: try adding the following to your extra-deps in ~/hashmir/stack.yaml:
- yeshql-3.0.1.3

By modifying the extra-deps: [] array like this in stack.yaml

extra-deps: [
    yeshql-3.0.1.3
]

will download YeshQL for the project and build it successfully.

When I run the app (use make run to do it), this is what I see:

% make run
Dropping and rebuilding database hashmir_test
time ~/.local/bin/hashmir-exe
There are 0 records.

real    0m0.019s
user    0m0.009s
sys 0m0.006s

YeshQL generated a function on-the-fly, named countClientSQL. That's the function I invoked in the countClient function. Since there are no records in the table, 0 was returned from the function.

Commit point

We set up the project, ran the first query against MySQL from Haskell via YeshQL templates. This brings us to the end of part 1 of this series. In the next article, we'll start adding more SQL queries to insert and query various records.

Friday, August 25, 2017

String Calculator with Applicative Functors in Haskell

I like the simplicity of the String Calculator kata. It's a typical example of the map-reduce algorithm, where the string has to be split by a delimiter, mapped into a list of integers and then reduced to their sum. I have often used it as an example to quickly evaluate engineering candidates, try new languages and tools. This was the first challenge I tried to solve in Haskell about a year ago. I found the code the other day, I wanted to see how I could improve upon it with everything I've learned so far.

This is what I found from a year before:

import Test.Hspec
import Data.List.Split (splitOn)

calculator :: String -> Integer
calculator x = sum $ coerceString x ","

coerceString :: String -> String -> [Integer]
coerceString x c = map (coerce) (splitOn c x)
    where
        coerce "" = 0
        coerce a = read a :: Integer

main :: IO ()
main = hspec $ do
    describe "String Calculator" $ do
        it "returns 0 for empty string" $ do
            calculator "" `shouldBe` 0
        it "returns 1 for '1'" $ do
            calculator "1" `shouldBe` 1
        it "returns 3 for '1,2,3'" $ do
            calculator "1,2,3" `shouldBe` 6

This works for simple cases, but what should I do when there is a non-numeric string in the input arguments?

        it "returns 0 for '1,2,!'" $ do
            calculator "1,2,!" `shouldBe` 0

The algorithm crashes as I suspected:

String Calculator
  returns 0 for empty string
  returns 1 for '1'
  returns 3 for '1,2,3'
  returns 0 for '1,2,!' FAILED [1]

Failures:

  string_calculator_01.hs:22:
  1) String Calculator returns 0 for '1,2,!'
       uncaught exception: ErrorCall (Prelude.read: no parse)

Randomized with seed 1943023196

Finished in 0.0023 seconds
4 examples, 1 failure

I could easily wrap the entire logic and return zero when an exception occurs, however, Haskell can do better. Much better.

I can return a Maybe Int from the operation that parses the string to an integer: if the result is Nothing here, the overall result will be Nothing.
There is a string parser in the Text.Read module that does just that, it's called readMaybe.

Here is how it works:

 % ghci
GHCi, version 8.0.2: http://www.haskell.org/ghc/  :? for help
Loaded GHCi configuration from /Users/adomokos/.ghci
λ> import Text.Read
λ> (readMaybe "12") :: Maybe Int
Just 12
λ> (readMaybe "!") :: Maybe Int
Nothing
λ>

I need to parse the list of strings, which is achieved by mapping over the values with readMaybe:

GHCi, version 8.0.2: http://www.haskell.org/ghc/  :? for help
Loaded GHCi configuration from /Users/adomokos/.ghci
λ> import Text.Read (readMaybe)
λ> import Data.List.Split (splitOn)
λ> let xs = "1,2,3"
λ> splitOn (",") xs
["1","2","3"]
λ> map (\x -> readMaybe x :: Maybe Int) $ splitOn (",") xs
[Just 1,Just 2,Just 3]

I now have a list of Maybe Int values that can be reduced into a single Maybe Int value. Reducing an array of numbers would be super easy (foldl (+) 0 [1,2,3] or foldl1 (+) [1,2,3] or just simply sum [1,2,3]). It's obvious that I will need something similar.

Adding a number to a Maybe Int can be achieved with a Functor:

λ> fmap (+10) (Just 4)
Just 14
λ> (+10) `fmap` (Just 4)
Just 14
λ> (+10) <$> Just 4
Just 14

All three expressions mean the same thing. The first is using the fmap as a conventional function name and arguments style, the second one uses the infix version of fmap and the third one is using a symbol.

This works for adding a number to a Maybe Int, however, I need to use an Applicative Functor to calculate the sum of two Maybe Ints.

λ> (+) <$> Just 10 <*> Just 4
Just 14

Using Applicative Functors, folding the list of Maybe Ints happens like this:

λ> let xs = [Just 1, Just 2, Just 3]
λ> foldl (\acc x -> (+) <$> x <*> acc) (Just 0) xs
Just 6
λ> foldl1 (\acc x -> (+) <$> x <*> acc) xs
Just 6

The solution now works, although a bit hard to read:

import Test.Hspec
import Text.Read (readMaybe)
import Data.List.Split (splitOn)

calculator :: String -> Maybe Int
calculator input =
    foldr (\x acc -> (+) <$> x <*> acc) (Just 0) $
        map (\x -> readMaybe x) $ splitOn "," input

main :: IO ()
main = hspec $ do
    describe "String Calculator" $ do
        it "returns Nothing for empty string" $ do
            calculator "" `shouldBe` Nothing
        it "returns Just 1 for '1'" $ do
            calculator "1" `shouldBe` Just 1
        it "returns Just 3 for '1,2,3'" $ do
            calculator "1,2,3" `shouldBe` Just 6
        it "returns Nothing for '1,2,3,!'" $ do
            calculator "1,2,3,!" `shouldBe` Nothing

It's more meaningful after I've refactored it into chunks:

calculator :: String -> Maybe Int
calculator input =
    let foldingFn acc x =  (+) <$> acc <*> x
        parsedInts = map (\x -> readMaybe x) . splitOn (",")
    in foldr1 (foldingFn) (parsedInts input)

The parsedInts function can be further simplified:

calculator :: String -> Maybe Int
calculator input =
    let foldingFn acc x =  (+) <$> acc <*> x
        parsedInts = map (readMaybe) . splitOn (",")
    in foldr1 (foldingFn) (parsedInts input)

And finally, the sum of list of Maybe values can be calculated by using the sequence function like this:

calculator :: String -> Maybe Int
calculator input =
    let parsedInts = map (readMaybe) . splitOn (",")
    in fmap sum . sequence $ parsedInts input

I find this form to be the most readable, but I liked the journey of getting there through the Applicative Functors.

Wednesday, June 28, 2017

Scenario Outline in RSpec

Some business logic is best described in a table format. The rows and columns can better detail all the different permutations, than words and sentences. Gherkin handles this well with its scenario outlines, however, RSpec does not have such feature.

As I wanted to test all the different permutation of a business logic, I ended up with a ~200 lines spec file. It was hard to review what use cases were covered as I had to scroll up and down to review the file. Had RSpec have a table format, this logic could have been described in 40 lines, which can easily fit into a screen.

It was obvious a simple DSL was needed, but what would that look like? This is what I came up with first:

RSpec.describe 'eating apples' do
  def method_under_test(start, eat, left)
    start - eat == left
  end

  def example_runner(examples)
    lines = examples.split(/\n/)
    lines.delete_at(0) # Don't need the title
    lines.each do |line|
      example = line.split('|')
      start, eat, left = example.reject(&:blank?).map(&:to_i)
      expect(method_under_test(start, eat, left)).to eq(true)
    end
  end

  it 'can tell how many are remaining' do
    examples =
      <<~TABLE
        | start | eat | left |
        | 12    | 5   | 7    |
        | 20    | 5   | 15   |
      TABLE

    example_runner(examples)
  end
end

I am using the example_runner method to parse the examples table and run the specs one by one. The spec itself is clear, the examples TABLE is easy to visualize.

The spec documentation, however, does not reveal the used table. This is what it looks like:

eating apples
  can tell how many are remaining

Let’s say, one of the examples will trigger an error. When I change the second example’s “left” column’s 15 value to 16, this is what I see when I run the spec.

F

Failures:

  1) eating apples can tell how many are remaining
     Failure/Error: expect(method_under_test(start, eat, left)).to eq(true)

       expected: true
            got: false

       (compared using ==)

This is not very useful. I don’t know which example failed. Fortunately RSpec lets us provide more context for an expectation. When I change the line of the assertion with an added context like this,

expect(method_under_test(start, eat, left)).to eq(true), line

the example that failed is properly reported by RSpec:

F

Failures:

  1) eating apples can tell how many are remaining
     Failure/Error: expect(method_under_test(start, eat, left)).to eq(true), line
       | 20    | 5   | 16   |

This could be a decent solution, however, not printing the examples for the documentation output has major drawbacks. Every row in the table should be a separate example, it should be expressed as such. With some clever formatting and using include_example a similar table layout can be achieved. Consider this example:

RSpec.describe 'eating apples' do
  def method_under_test(start, eat, left)
    start - eat == left
  end

  RSpec.shared_examples 'eating_apples' do |example|
    it "calculates the remaining apple for - #{example}" do
      parsed_example = example.split('|')
      start, eat, left = parsed_example.reject(&:blank?).map(&:to_i)

      result = method_under_test(start, eat, left)
      expect(result).to eq(true)
    end
  end

                                     # start | eat | left
                                     # ______|_____|_____
  include_examples 'eating_apples', '  12    | 5   | 7'
  include_examples 'eating_apples', '  20    | 5   | 15'
end

The table is shifted to the right, but it’s clearly visible, and the commented out header provides a different color in my IDE that separates it from the data.

include_examples

The shared_example is also a lot simpler, I don’t need to deal with the header information, as it’s not passed to it, examples are being ran line by line.

By making the test description dynamic, the reporter prints out the tested values:

eating apples
  calculates the remaining apple for -   12    | 5   | 7
  calculates the remaining apple for -   20    | 5   | 15

The triggered error is properly reported by both the line number and the example used for that particular use case.

.F

Failures:

  1) eating apples calculates the remaining apple for -   20    | 5   | 16
     Failure/Error: expect(result).to eq(true)

       expected: true
            got: false

       (compared using ==)

I am not a big fan of adjusted white space for positioning equal signs under one another, but am happy to use white space to format test code to better describe the logic it validates.

RSpec does not have scenario outlines, but hopefully, with a shared_example and a bit of clever formatting the same can be accomplished.

Wednesday, May 31, 2017

Bit Shifting for a Shard ID in Ruby

As our database grew, we had to take a serious look at how can we could split it up by our clients, as most of them wanted to have their own data separate from the others anyway. A few months ago I found a great article from Pinterest, that describes how they sharded their MySQL database.

A sharded entity needs a UUID to uniquely identify the record across all shards. Most of the programming languages can generate a UUID easily, however, what was amazing to me, was that Pinterest generated its own unique ids by encapsulating three distinct numbers in one. Wait, what??! Read that article, it’s definitely worth your time.

While Pinterest encapsulated three numbers into one, we only needed two, a client_id and an entity_id. Our client_id would be a much smaller number than our entity_id, we wanted to reserve more bits for the latter.

It turns out, Ruby has many friendlier tools to deal with binary operations. Let's look at them!

What is the binary representation of the integer number 123?

123-in-binary

Out of the 7 bits, you'll see that the 3rd one is turned off, all the others are turned on, giving us 64 + 32 + 16 + 8 + 2 +1 = 123. How can we get this binary representation in Ruby? It's super easy, just use the to_s(2) method to do it.

pry(main)> 123.to_s(2)
=>"1111011"

This is the exact same string representation as the one in the image above, where the third bit is turned off and represented with a zero.

I'd like to keep the client_id on the left side, but I'd like to reserve bits on the right side. For the sake of simplicity, I will keep this as a small number. Let's add 5 bits to the right-hand side of these bits by using the bitwise left shift operator.

pry(main)> (123 << 5).to_s(2)
=> "111101100000"

The original number, 123, is still represented on the left side, but 5 "0"s were added to the right-hand side. You get the numeric representation of this by leaving out the to_s(2) call at the end:

pry(main)> 123 << 5
=> 3936

This number can be converted back to binary:

pry(main)> 3936.to_s(2)
=> "111101100000"

On the left side I have the binary representation of 123, but how about those bits on the right side? What are those representing? Right now, those bits are all turned off, they will give you 0 ("00000".to_i(2) => 0). How can I store the number 3 on the right side? The bits should look like this:

3-on-right-side

The binary "|" will turn the two rightmost bits on:

pry(main)> (123 MM 5 | 3).to_s(2)
=> "111101100011"

Again, leaving out the to_s(2) will provide the number representation:

pry(main)> (123 << 5 | 3)
=> 3939

The storing part will work, but how can we get our two numbers back from this one combined number? Well, we have to split the bits and convert the binary representation to an integer.

Five bits were used on the right side to store our second number. We need to chop those off to get the number stored on the left side. The bitwise right shift (>>) will do just that:

pry(main)> (3939 >> 5).to_s(2)
=> "1111011"

The string "1111011" is our original 123 in a binary string format. We can convert that to an integer by using the to_i(2) String method:

pry(main)> (3939 >> 5).to_s(2).to_i(2)
=> 123

I right shifted the original number, 3939, converted it to a binary string and converted that to an Integer.

There are more efficient ways to do this by using a binary "&" (3939 >> 5) & 0b1111111 => 123 with the max value the bits can represent. That's what the Pinterest article had, but I found using the Ruby conversion methods a bit more friendly to those of us who are not dealing with binary data on a daily basis.

We have the number on the left side, but what about the number on the right side? When we convert the number representation (3939) to a binary string, we know that the five characters on the right side will represent the bits of our other number. Ruby String’s last(x) will do just that:

pry(main)> (3939 >> 0).to_s(2).last(5)
=> "00011"

Converting this binary String to Integer should be similar to what we've done before:

pry(main)> (3939 >> 0).to_s(2).last(5).to_i(2)
=> 3

Using the binary "&" with the max number the bits can store will do this conversion in one step: (3939 >> 0) & 0b11111 => 3. As a side note, the binary number can be represented as a hexadecimal value: (3939 >> 0) & 0x1F => 3. This is a lot shorter than a series of ones and zeros.

There is a limit of how large the numbers can be as you have a limit of bits to store those. The max number can be determined by flipping the available bits on. For an 7 bit number it's 64 + 32 + 16 + 8 + 4 + 2 + 1 = 127 or 2**x-1, where x is the number of bits. In our case it is 2**7-1 = 127.

We ended up using a 64-bit Integer for our shard_id, which is a BIGINT in MySQL. We store client_id in 22 bits giving us the maximum of 2**22-1 = 4_194_304 and 40 bits for the entity_id with 2**40-1 = 1_099_511_627_775 max value. The remaining two bits are "worth in gold, we can expand the number or store a third (albeit small) number in it.

Tuesday, April 25, 2017

Fireside Chat

When I saw a retweet from Jason Fried about available tickets to a fireside chat with him at Basecamp, I jumped on it. I figured if I can kill two birds with one stone, - meeting him in person and seeing their offices - it's a no-brainer. Company Culture was the topic of the conversation led by Aimee Groth, who visited Chicago to publicize her new book, Kingdom of Happiness about Zappos' culture.

Basecamp HQ

Basecamp HQ is as cool as you think it is. Very few desks, a couple of meeting rooms. It reminded me more of a train terminal with its large windows and limited furnishing than a real office. The office is centered around an auditorium, which is an effective PR and educational platform for the company.

I enjoyed looking at the walls covered with postcards from employees all over the world, but I especially liked David's H-1B approval notice from the USCIS from 2005. I laughed out loud when I noticed it, as I had to go through similar hassle myself, but mine is safely guarded with my documents at home.

Basecamp works in six weeks work schedule. Whatever the team can get down in six weeks, they will deliver it. The scope can change, but the six weeks schedule is hard set. This timeframe helps them delivering functionality, and since the company is working remotely, it worked out well for them.

They don't have managers who only manage people or projects, the teams are led by team leads. These team leads are developers as well, shipping code on a daily basis. Jason and his team realized that managers who do not code, grow away from the work. According to him, "professional (full time) managers forget to do the work".
At one point they've tried rotating team leads, but that did not work out, as the continuity was lost. I could see that: "I see this problem, but I won't deal with it, I'll leave it for the next person, who will take over." Basecamp is looking for people who are self-managed, however, Jason emphasized multiple times: "people like to be led". It's important to "rally the folks by clear goals and purpose".

Jason also talked about the Jeff Bezos investment in the company, which meant a small ownership stake in Basecamp. They did not need the money to survive, David and Jason felt having a person like Mr. Bezos is mutually beneficial to both parties. "Who would not like to have Jeff Bezos as an advisor in his or her company?!" They have not talked to Jeff Bezos for a while, but if they wanted to, they could just reach out to his secretary, set up a meeting, and Jeff would fly to Chicago for a meeting or dinner with them.

The best advice from Bezos - and according to Jason, this was worth the entire dividend they have paid for his investment - was: "invest in the things in your business, that won't change". Don't chase the shiny new things, stick to what will not change. For them, it's Basecamp. The company had 4-5 products that they sold a couple of years ago to focus on their main product, which is Basecamp.

Jason went into details why other products were divested (like HighRise, Backpack, Campfire). Maintaining the web, Android and iOS versions of their products resulted in 15 different projects. That led to insufficient focus for each platform for each product with the employees they had at the time. They could - of course - have hired other developers, but they intentionally wanted to stay small. They did not want to get richer, be the next billionaire, they were just as happy with what they had. This sounds strange, almost naive in the era of bloated startups that are bleeding money chasing to be the next Facebook.

I enjoyed the Q&A at the very end. Some interesting questions came up about the startup community in Chicago, about VCs in general. Jason kindly offered to stay as long as everybody's questions were answered. Really a courteous offer, considering it was after 8 pm on a Friday night.

Oh, yes, and one more thing: Basecamp has 130,000 paying customers. It's a remarkable achievement by a company that has never taken VC money, was profitable from the get-go, and created an exciting app in the "not-so-exciting" domain of project management.

Tuesday, March 28, 2017

Containers

As I was exploring how to make Golang even faster on AWS Lambda, I found a project that promised sub-millisecond execution time compared to my (already pretty good) ~60 millisecond. It used Python execution that ran the Go code in-process in contrast to my attempt, where I had to spawn a new process and execute the lambda code there. Very clever, no wonder that solution did not have the 60-millisecond penalty for running that code. However, in order to build the sample code for this AWS Lambda I had to use Docker.

I've heard about Docker years ago, understood what it's used for at a very high level, however, I have never really given it a serious try. I figured it was time. Boy, I was in for some pleasant surprise!

The project AWS Lambda Go from Eawsy used Docker to containerize their build environment on my laptop. What does that mean? Imagine having a build server running on your computer in seconds, where the version of the Go compiler, the Python environment is set by the author of the Dockerfile. I'd use a little build engine that takes in my code, runs its magic and a zip file comes out that I can run on Lambda. What?!

I wrote all these different tutorials about running MRI Ruby on AWS Lambda or interacting with a Postgres DB with Clojure and I had to set up all the prereqs in plain text: "you have to have Postgres running, and Clojure, and MRI Ruby". I provided all the different Makefile scripts to follow the examples. However, with Docker, I'd just provide a Dockerfile that sets up the environment in the future.

I believe containers are big and will be even bigger very soon.

I see more and more applications where the code describes the behavior and the container descriptor describes the environment.


They live side by side, clearly stating what virtual components the software needs to execute. Engineers can run the software with those containers locally, and the software can be deployed to the cloud with those images pre-built, with tight control over its execution context.

There are many resources to learn Docker. I started with reading the Docker in Action book and went further by reading the Docker in Practice book.

I created a Docker templates repository, where I collected ideas for different recipes. Do I need a Ruby worker with Redis and Postgres backend? I'll just run docker compose up with this docker_compose.yml file and I have an environment, where everything from the OS to the version of Redis and Postgres is predefined. If it works on my machine, it will work on yours, too.

There are many things I like about Docker as compared to Vagrant or other virtual machine solutions. The biggest thing for me is the low power Docker containers would need. Vagrant images would reserve 2 of your 4 cores and 8GB memory when Docker will only take from the host as much as it needs. If it's 32MB, that's it, if it's 1GB, it will take that much.

Docker is the future, and you will see more and more code repos with a Dockerfile in it.

Wednesday, February 8, 2017

Golang

The first time I heard about Golang was a few years back, when the great guys at Brad's Deals, our next door office neighbor organized and hosted the local Go meetup there. Then IO.js and Node.js war broke out and TJ Holowaychuck shifted from Node.js to Golang announcing the move in an open letter to the community.
I did not think much of the language, as its reputation was far from the beauty of a real functional language.

Fast forward a couple of years and I am giving Ruby a serious try on AWS Lambda. Ruby works there, however, it needs enough memory and 3000 ms (3 seconds) to do anything. We have to invoke some of them millions of times in a month and when we calculated the cost for it, the bill gets fairly large quickly.

I created a simple AWS Lambda with Ruby just to print the words "Hello, World!" with 128 MB memory. It took 5339 ms to execute it.

Ruby Hello World on AWS Lambda

Then one night I wrote a tiny Go program:

package main

import "fmt"

func main() {
  fmt.Println("Hello, World!")
}

I cross compiled (since I am working on OSX) with the command GOOS=linux GOARCH=amd64 go build github.com/adomokos/hello to Linux, packaged it up with a Node.JS executor and ran it. I couldn't believe my eyes, it took only 68 ms to get the string "Hello, World!" back. 68 ms! And it was on a 128 MB memory instance. It was beautiful!

Go Hello World on AWS Lambda

Ruby would need four times the memory and it would still execute ~10 times slower than Go. That was the moment when I got hooked.

Go is a simple language. I am not saying it's easy to learn, it's subjective: it depends on your background, your experience. But it's far from the beauty of Haskell or Clojure. However, the team I am working with would have no trouble switching between Go and Ruby multiple times a day.

What kind of a language today does not have map or reduce functions?! Especially when functions are first-class citizens in the language. It turns out, I can write my own map function if I need to:

package collections

import (
  "github.com/stretchr/testify/assert"
  "strconv"
  "testing"
)

func fmap(f func(int) string, numbers []int) []string {
  items := make([]string, len(numbers))

  for i, item := range numbers {
    items[i] = f(item)
  }

  return items
}

func TestFMap(t *testing.T) {
  numbers := []int{1, 2, 3}
  result := fmap(func(item int) string { return strconv.Itoa(item) }, numbers)
  assert.Equal(t, []string{"1", "2", "3"}, result)
}

Writing map with recursion would be more elegant, but it's not as performant as using a slice with defined length that does not have to grow during the operation.

History

Go was created by some very smart people at Google, I wanted to understand their decision to keep a language this pure.
Google has a large amount of code in C and C++, however, those languages are far from modern concepts, like parallel execution and web programming to name a few. Those languages were created in the 60s and 80s, well before the era of multi-core processors and the Internet. Compiling a massive codebase in C++ can easily take hour(s), and while they were waiting for compilation, the idea of a fast compiling, simple, modern language idea was born. Go does not aim to be shiny and nice, no, its designers kept it:

  • to be simple and easy to learn
  • to compile fast
  • to run fast
  • to make parallel processing easy

Google hires massive number of fresh CS graduates each year with some C++ and Java programming experience, these engineers can feel right at home with Go, where the syntax and concept is similar to those languages.

Tooling

Go comes with many built-in tools, like code formatting and benchmarking to name the few. In fact I set up Vim Go that leverages many of those tools for me. I can run, test code with only a couple of keystrokes.

Let's see how performant the procedure I wrote above is. But before I do that I'll introduce another function where the slice's length is not pre-determined at the beginning of the operation, this way it has to auto-scale internally.

func fmapAutoScale(f func(int) string, numbers []int) []string {
  // Initialize a slice with default length, it will auto-scale
  var items []string

  for _, item := range numbers {
    items = append(items, f(item))
  }

  return items
}

The function is doing the same as fmap, similar test should verify the logic.

I added two benchmark tests to cover these functions:

// Run benchmark with this command
// go test -v fmap_test.go -run="none" -benchtime="3s" -bench="BenchmarkFmap"
// -benchmem
func BenchmarkFmap(b *testing.B) {
  b.ResetTimer()

  numbers := []int{1, 2, 3, 4, 5, 6, 7, 8, 9, 10}
  for i := 0; i < b.N; i++ {
    fmap(func(item int) string { return strconv.Itoa(item)  }, numbers)
  }
}

func BenchmarkFmapAutoScale(b *testing.B) {
  b.ResetTimer()

  numbers := []int{1, 2, 3, 4, 5, 6, 7, 8, 9, 10}
  for i := 0; i < b.N; i++ {
    fmapAutoScale(func(item int) string { return strconv.Itoa(item)  },
    numbers)
  }
}

When I ran the benchmark tests, this is the result I received:

 % go test -v fmap_test.go -run="none" -benchtime="3s" -bench="BenchmarkFmap"
    -benchmem
 BenchmarkFmap-4   ‡ 10000000 | 485 ns/op | 172 B/op | 11 allocs/op
 BenchmarkFmapAS-4 ‡ 5000000  | 851 ns/op | 508 B/op | 15 allocs/op
 PASS
 ok   command-line-arguments  10.476s

The first function, where I set the slice size to the exact size is more performant than the second one, where I just initialize the slice and let it autoscale. The ns/op displays the execution length per operation in nanoseconds. The B/op output describes the bytes it uses per operation. The last column describes how many memory allocations it uses per operation. The difference is insignificant, but you can see how this can become very useful as you try writing performant code.

Popularity

Go is getting popular. In fact, very popular. It was TIOBE's "Language of the Year" gaining 2.16% in one year. I am sure you'll be seeing articles about Go more and more. Check it out if you haven't done it yet, as the chance of finding a project or job that uses Go is increasing every day.