Skip to content

Instantly share code, notes, and snippets.

@jorinvo
Last active January 1, 2018 14:41
Show Gist options
  • Select an option

  • Save jorinvo/3d7f6a60fcede1863fa9f0788b8cc1b4 to your computer and use it in GitHub Desktop.

Select an option

Save jorinvo/3d7f6a60fcede1863fa9f0788b8cc1b4 to your computer and use it in GitHub Desktop.
Dirty Money: Code Challenge https://jorin.me/dirtymoney

Follow the dirty money

A shady Internet business has been discovered.

The website has been made public by a whistle blower. We have enough evidence about the dirty deals they did. But to charge them we need to get hands on precise numbers about the transactions that happened on their platform.

Unfortunately no record of the transactions could be seized so far. The only hint we have is this one transaction:

fd0d929f-966f-4d1a-89cd-feee5a1c5347.json

What we need is the total of all transactions in Dollar. Can you trace down all other transactions and get the total?

Be careful to count each transaction only once, even if it is linked multiple times. You can use whatever tool works best.

Please share the total and your solution below!

Cheers, Jorin

@jorinvo

(For 2015 challenge see https://jorin.me/csv-challenge)

@adamantmc

Copy link
Copy Markdown
import requests
import json
import re

regex =  re.compile(r"\$[0-9]*[,.][0-9]*")
visited_links = {}
queue = ["https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json"]
total = 0
transactions = 0

for link in queue:
    obj = json.loads(requests.get(link).text)
    if not link in visited_links:
        transactions += 1
        for new_link in obj["links"]:
            queue.append(new_link)
        total += float(regex.findall(obj["content"])[0][1:].replace(",","."))
        visited_links[link] = 1

print("Total: " + str(total))
print("Transactions: " + str(transactions))

Output:
Total: 9064.789999999999
Transactions: 50

@XANi

XANi commented Feb 6, 2017

Copy link
Copy Markdown

Ugly Perl version:

use JSON;
use LWP::Simple;
use List::Util qw(sum);

my $tr_list;
p('https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json');
my @tr_val = map { s{,}{.}; m/\$(\d+\.\d+)/; $1 } values %$tr_list;
print "Sum: " . sum(@tr_val) . " count: " . scalar @tr_val . "\n";

sub p {
    my $tr = decode_json( get(shift) );
    if ( defined( $tr_list->{ $tr->{'id'} } ) ) { return }
    for my $url ( @{ $tr->{'links'} } ) {
        p($url);
    }
    $tr_list->{ $tr->{'id'} } = $tr->{'content'};
}

Sum: 9064.79 count: 50

@fGeorjje

fGeorjje commented Feb 6, 2017

Copy link
Copy Markdown

Kotlin:

import com.google.gson.Gson
import java.math.BigDecimal
import java.net.URL
import java.util.*

data class Transaction(val id: String, val content: String, val links: List<String>)
fun main(args: Array<String>) {
  val q = LinkedList<String>(listOf("https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json"))
  generateSequence { q.poll() }
    .distinct()
    .map { Gson().fromJson(URL(it).readText(), Transaction::class.java) }
    .fold(BigDecimal.ZERO) { sum, t ->
      q.addAll(t.links)
      sum + Regex("\\$(\\d+\\.\\d+)").find(t.content.replace(',', '.'))!!.groupValues[1].let { BigDecimal(it) }
    }.let { println(it) }
}

9064.79

@kasunben

kasunben commented Feb 6, 2017

Copy link
Copy Markdown

a simple PHP version;


<?php
function processTransactions($reqUrl, &$queue) {
  $response = json_decode(file_get_contents($reqUrl));

  $id = $response->id;
  $content = $response->content;
  $queue[$id] = $content;

  $links = $response->links;
  foreach($links as $link) {
    processTransactions($link, $queue);
  }

  return $queue;
}

$initialReqUrl = "https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json";
$queue = array();
$queueFinal = processTransactions($initialReqUrl, $queue);
$totalRecords = count($queueFinal);
$queueUnique = array_unique($queueFinal); // remove duplicates

// Count Total
$totalTransactions = array(
  'records' => $totalRecords,
  'records_unique' => count($queueUnique),
  'amount' => 0
);

foreach($queueUnique as $key => $value) {
  preg_match_all('/\$[0-9]+[,.][0-9]*/', $value, $matchedArr);
  $amount = (float) str_replace(array('$',','), '', $matchedArr[0][0]);
  $totalTransactions['amount'] += $amount;
}

print "Result: `".$totalTransactions['records']." total records`, `".$totalTransactions['records_unique']." unique transaction` and `$".$totalTransactions['amount']."`";

// eof.

Results: 50 total records, 35 unique transaction and $519238.86

Note:
As my understanding, $9064.79 is incorrect.
Because;

  1. $149,97 != $149.97
  2. To remove duplicate transactions, we must check for transaction contents instead of links. It may include same transaction details in multiple pages with different links.

@chocolateboy

chocolateboy commented Feb 6, 2017

Copy link
Copy Markdown

@evowise

evowise commented Feb 6, 2017

Copy link
Copy Markdown
import json
import requests
import re
p = re.compile(r'\$\d+[\,,\.]\d+')

def get_sum(t):
 t = p.search(j["content"]).group(0)
 if ',' in t:
  return float(t.replace(',','.')[1:])
 return float(t[1:])

def get_node_data(url):
 return json.loads(requests.get(url).text)

j = get_node_data("https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json")
s = {j["id"]}
sum = get_sum(j["content"])

links = j.get("links")
while links:
 x = []
 for l in links:
  j = get_node_data(l)
  if j["id"] not in s:
   s.add(j["id"])
   print j["content"]
   sum += get_sum(j["content"])
   x.extend(j["links"])
 links = x

print sum

It works! :-) 9064.79

@jorinvo

jorinvo commented Feb 6, 2017

Copy link
Copy Markdown
Author

Thanks @chocolateboy! Next time :)
But looks like the discontinue their service? http://www.git.to/

@chocolateboy

chocolateboy commented Feb 6, 2017

Copy link
Copy Markdown

@jorinvo No, it's still up (it's GitHub's own link shortener). It's git.io rather than git.to, though :-)

@matthew-hallsworth

matthew-hallsworth commented Feb 7, 2017

Copy link
Copy Markdown

Working PHP solution

<?php

$start = 'https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json';
$original = fetch_and_decode($start);
$keyed_array = [];

$keyed_array = process_links($original, $keyed_array);

echo "transactions: " . count($keyed_array) . ", total: " . array_sum($keyed_array) . "\n";

//--

function process_links($payload, $keyed_array) {

  if (!array_key_exists($payload->id, $keyed_array)) {
    $keyed_array[$payload->id] = find_transaction($payload);
  }

  if (count($payload->links) == 0) return $keyed_array;

  foreach($payload->links as $link) {
    $link_fetched = fetch_and_decode($link);
    $keyed_array = process_links($link_fetched, $keyed_array);
  }

  return $keyed_array;
}


function fetch_and_decode($url) {
  return json_decode(file_get_contents($url));
}

function find_transaction($decoded_transaction) {
  preg_match('/\$(\d*(\.|,)\d*)/', $decoded_transaction->content, $matches);
  return(str_replace(',', '.', $matches[1]));
}

tested on php 5.6

transactions: 50, total: 9064.79

@jorinvo

jorinvo commented Feb 7, 2017

Copy link
Copy Markdown
Author

@chocolateboy, I see, cool :) But I won't change the links for this post anymore. Next time!

@aptinio

aptinio commented Feb 14, 2017

Copy link
Copy Markdown

In Elixir (not concurrent yet):

defmodule DirtyMoney do
  def total(link) do
    {:ok, agent} = Agent.start_link(fn -> %{} end)

    follow(link, agent)

    Agent.get agent, fn transactions ->
      transactions
      |> Map.values
      |> Enum.sum
      |> Kernel./(100)
    end
  end

  def follow(link, agent) do
    %{"id" => id, "content" => content, "links" => links} =
      link
      |> HTTPoison.get!([], follow_redirect: true)
      |> Map.fetch!(:body)
      |> Poison.decode!

    count(agent, id, content)

    Enum.each(links, fn link ->
      follow(link, agent)
    end)
  end

  def count(agent, id, content) do
    Agent.update agent, fn transactions ->
      Map.put_new_lazy transactions, id, fn ->
        parse(content)
      end
    end
  end

  def parse(content) do
    [_, whole, part] = Regex.run(~r/\$(\d+)[,.](\d+)/, content)

    cents =
      part
      |> String.pad_trailing(2, "0")
      |> String.to_integer

    cents + String.to_integer(whole) * 100
  end
end

DirtyMoney.total("https://git.io/vDCxb")

@jorinvo

jorinvo commented Nov 9, 2017

Copy link
Copy Markdown
Author

There are more amazing answers over at dev.to!

@assafmo

assafmo commented Jan 1, 2018

Copy link
Copy Markdown

bash, jq, wget, grep, sed, awk:

#!/bin/bash
mkdir xyz && cd xyz
echo -e '0\n1' > /tmp/count
wget https://gist.githubusercontent.com/jorinvo/6f68380dd07e5db3cf5fd48b2465bb04/raw/c02b1e0b45ecb2e54b36e4410d0631a66d474323/fd0d929f-966f-4d1a-89cd-feee5a1c5347.json > /dev/null 2> /dev/null
while [[ $(tail -1 /tmp/count) -ne $(tail -2 /tmp/count | head -1) ]]; do
  jq -r '.links[]' *.json | xargs wget -nc > /dev/null 2> /dev/null
  ls | wc -l >> /tmp/count
done
jq .content *.json | grep -Eo '\$[0-9,.]+[0-9]' | tr -d '$' | sed 's/,/./g' | awk '{sum = sum + $1} END{print sum}'
cd ..
rm -rf ./xyz /tmp/count

9064.79 😄

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment