From d4ce905d1ac350f4f3ae3171bedd556097d0bcf7 Mon Sep 17 00:00:00 2001
From: Chuck Dries <chuck@chuckdries.com>
Date: Mon, 9 Aug 2021 12:25:47 -0700
Subject: [PATCH] initial deduplication implementation

---
 .gitignore   |  1 +
 index.js     | 59 +++++++++++++++++++++++++++++++++++++
 leads.json   | 82 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 package.json | 11 +++++++
 4 files changed, 153 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 index.js
 create mode 100644 leads.json
 create mode 100644 package.json

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..496ee2c
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+.DS_Store
\ No newline at end of file
diff --git a/index.js b/index.js
new file mode 100644
index 0000000..d51a185
--- /dev/null
+++ b/index.js
@@ -0,0 +1,59 @@
+const fs = require("fs");
+const { leads } = JSON.parse(fs.readFileSync("./leads.json"));
+
+// index records by ID for easy lookup
+const leadsById = {};
+// index IDs by email for easy lookup
+const leadIdssByEmail = {};
+
+const collisions = [];
+
+for (const currentLead of leads) {
+  const { _id, email } = currentLead;
+  const collidingLeadIdByEmail = leadIdssByEmail[email];
+  const collidingLead = collidingLeadIdByEmail
+    ? leadsById[collidingLeadIdByEmail]
+    : leadsById[_id];
+
+  if (collidingLead) {
+    const collision = {
+      left: collidingLead,
+      right: currentLead,
+      collidingField: collidingLeadIdByEmail ? "email" : "_id",
+    };
+    collisions.push(collision);
+    const lDate = new Date(collision.left.entryDate);
+    const rDate = new Date(collision.right.entryDate);
+    if (lDate > rDate) {
+      // existing lead is newer than current lead
+      // discard current lead by doing nothing with it
+      collision.took = "left";
+    } else {
+      // current lead is newer than existing lead, or both leads have the same date
+      // either way, take the current lead over the existing one
+      collision.took = "right";
+      if (collision.collidingField === "_id") {
+        // colliding ID - replace ID index, delete old email in email index
+        delete leadIdssByEmail[collision.left.email];
+        leadIdssByEmail[email] = _id;
+        leadsById[_id] = currentLead;
+      } else {
+        // colliding email - replace ID in email index, delete old ID index
+        leadIdssByEmail[email] = _id;
+        delete leadsById[collision.left._id];
+        leadsById[_id] = currentLead;
+      }
+    }
+  } else {
+    // no collision
+    leadsById[currentLead._id] = currentLead;
+    leadIdssByEmail[currentLead.email] = currentLead._id;
+  }
+}
+
+console.log("collisions", collisions);
+console.log("leadsById", leadsById);
+
+console.log('records processed:', leads.length)
+console.log('collisions:', collisions.length)
+console.log('output leads:', Object.keys(leadsById).length)
\ No newline at end of file
diff --git a/leads.json b/leads.json
new file mode 100644
index 0000000..8444c57
--- /dev/null
+++ b/leads.json
@@ -0,0 +1,82 @@
+{"leads":[
+{
+"_id": "jkj238238jdsnfsj23",
+"email": "foo@bar.com",
+"firstName":  "John",
+"lastName": "Smith",
+"address": "123 Street St",
+"entryDate": "2014-05-07T17:30:20+00:00"
+},
+{
+"_id": "edu45238jdsnfsj23",
+"email": "mae@bar.com",
+"firstName":  "Ted",
+"lastName": "Masters",
+"address": "44 North Hampton St",
+"entryDate": "2014-05-07T17:31:20+00:00"
+},
+{
+"_id": "wabaj238238jdsnfsj23",
+"email": "bog@bar.com",
+"firstName":  "Fran",
+"lastName": "Jones",
+"address": "8803 Dark St",
+"entryDate": "2014-05-07T17:31:20+00:00"
+},
+{
+"_id": "jkj238238jdsnfsj23",
+"email": "coo@bar.com",
+"firstName":  "Ted",
+"lastName": "Jones",
+"address": "456 Neat St",
+"entryDate": "2014-05-07T17:32:20+00:00"
+},
+{
+"_id": "sel045238jdsnfsj23",
+"email": "foo@bar.com",
+"firstName":  "John",
+"lastName": "Smith",
+"address": "123 Street St",
+"entryDate": "2014-05-07T17:32:20+00:00"
+},
+{
+"_id": "qest38238jdsnfsj23",
+"email": "foo@bar.com",
+"firstName":  "John",
+"lastName": "Smith",
+"address": "123 Street St",
+"entryDate": "2014-05-07T17:32:20+00:00"
+},
+{
+"_id": "vug789238jdsnfsj23",
+"email": "foo1@bar.com",
+"firstName":  "Blake",
+"lastName": "Douglas",
+"address": "123 Reach St",
+"entryDate": "2014-05-07T17:33:20+00:00"
+},
+{
+"_id": "wuj08238jdsnfsj23",
+"email": "foo@bar.com",
+"firstName":  "Micah",
+"lastName": "Valmer",
+"address": "123 Street St",
+"entryDate": "2014-05-07T17:33:20+00:00"
+},
+{
+"_id": "belr28238jdsnfsj23",
+"email": "mae@bar.com",
+"firstName":  "Tallulah",
+"lastName": "Smith",
+"address": "123 Water St",
+"entryDate": "2014-05-07T17:33:20+00:00"
+},
+{
+"_id": "jkj238238jdsnfsj23",
+"email": "bill@bar.com",
+"firstName":  "John",
+"lastName": "Smith",
+"address": "888 Mayberry St",
+"entryDate": "2014-05-07T17:33:20+00:00"
+}]
+}
\ No newline at end of file
diff --git a/package.json b/package.json
new file mode 100644
index 0000000..2c405b6
--- /dev/null
+++ b/package.json
@@ -0,0 +1,11 @@
+{
+  "name": "adobe-coding-challenge",
+  "version": "1.0.0",
+  "description": "",
+  "main": "index.js",
+  "scripts": {
+    "test": "echo \"Error: no test specified\" && exit 1"
+  },
+  "author": "",
+  "license": "ISC"
+}