Cleaned-up README and added more examples

README.md

@@ -2,6 +2,8 @@
 
 This demo is used by Martijn Visser in his [Flink Forward talk 'Only SQL: Empower data analysts end-to-end with Flink SQL'](https://www.flink-forward.org/global-2021/conference-program#only-sql--empower-data-analysts-end-to-end-with-flink-sql)
 
+[![Twitter Follow](https://img.shields.io/twitter/follow/MartijnVisser82?style=social&logo=twitter)](https://twitter.com/MartijnVisser82) [![GitHub Follow](https://img.shields.io/github/followers/MartijnVisser?style=social&logo=github)](https://github.com/MartijnVisser)
+
 ## Docker
 
 We'll use Docker Compose to start all necessary services to run the demos. It will start the following services:
@@ -14,7 +16,7 @@ We'll use Docker Compose to start all necessary services to run the demos. It wi
 * Elasticsearch 7.15.0, accessible via http://localhost:9200 (or http://elasticsearch:9200 via Docker networking)
 * http-server: a simple static HTTP server, accessible via http://localhost:8080
 
-![Demo overview](overview.png "Demo overview")
+![Demo only-sql-overview](only-sql-overview.png "Demo overview")
 
 ## Starting the demo
 
@@ -34,7 +36,7 @@ docker exec -it `docker ps -q --filter "ancestor=ftisiot/flink_sql_cli:1.13.2"`
 
 ## Explore all website behaviour
 
-Any visit to one of the webpages is sent to the Kafka topic `pageview`. In order to explore them, we first need to register this Kafka topic as a TABLE. 
+Any visit to one of the webpages is sent to the Kafka topic `pageview`. In order to explore them, we first need to register this Kafka topic as a table. 
 
 ```sql
 --Create table pageviews:
@@ -56,7 +58,7 @@ CREATE TABLE pageviews (
 );
 ```
 
-Any cookie that belongs to our website, is also sent to the topic. We are specifically interested in a cookie called `identifier`. We're going to register a VIEW, which returns this value by applygin a regular expressing on the incoming data. 
+Any cookie that belongs to the domain `localhostd` (which is where our website runs), is also sent to the topic. We are specifically interested in a cookie called `identifier`. We're going to register a view, which returns this value by applying a regular expressing on the incoming data. 
 
 ```sql
 --Create view which already extracts the identifier from the cookies
@@ -79,20 +81,32 @@ By now running queries on the view while visiting a webpage, you will see data a
 SELECT * from web_activities;
 ```
 
-![Flink SQL Client Results](flink-sql-client-results-01.png "Flink SQL Client Results")
+![Flink SQL Client Results](only-sql-results-01.png "Flink SQL Client Results")
+
+In this example, we're interested in users who visit our homepage more than 3 times in 10 seconds. In order to achieve that, 
+we're first going to select all identifiers (which contains the value of the cookie `identifier`) who are visiting the url that 
+matches with `http://localhost:8080`. We're expanding our selection with the `TUMBLE` function which creates fixed windows of 
+10 seconds. Last but not least, we're using the `HAVING` function to make sure that only the identifiers that are occurring more 
+than 3 times are being selected. The end result looks like this:
 
 ```sql 
 --Get the users who visit the homepage more than 3 times in 10 seconds
 SELECT  
     `identifier`,
     CONCAT('You Have Seen The Homepage ', CAST(COUNT(*) as VARCHAR), ' Times')  
 FROM web_activities
-WHERE `url` = 'http://localhost:8080/' 
+WHERE `url` = 'http://localhost:8080/' OR `url` = 'http://127.0.0.1:8080/'
 GROUP BY 
     TUMBLE(ts, INTERVAL '10' SECONDS), identifier
 HAVING COUNT(*) > 3;
 ```
 
+We've just created the list of `identifier` that meet our requirement of visiting the homepage more than 3 times in 10 seconds.
+We now want to act on this data. In order to achieve that, we're going to send the list of `identifer` to our Elasticsearch sink. 
+Our website checks if there's any result in the Elasticsearch results and if so, it will display the notification. 
+
+To send the data to Elasticsearch, we first have to create another table like we've done before. We use the following DDL:
+
 ```sql
 --Create a sink to display a notification
 CREATE TABLE notifications (
@@ -108,6 +122,9 @@ CREATE TABLE notifications (
 );
 ```
 
+When that table is created, we'll re-use the previous SQL that returns the list of `identifier` and send those results to the 
+previously created table. 
+
 ```sql
 INSERT INTO notifications (`identifier`, `notification_id`, `notification_text`)
     SELECT  
@@ -121,6 +138,22 @@ GROUP BY
 HAVING COUNT(*) > 3;
 ```
 
+> :warning: The default value of the cookie `identifier` is `anonymous`. No notifications will be displayed if the value is `anonymous`. 
+>
+> In order to change the value, you need to open the Developer Tools via either Cmd + Opt + J (on Mac) or Ctrl + Shift + J (on Windows)
+> 
+> In the opened console, you then need to type `document.cookie="identifier=YourIdentifier"` to change the value of the `identifier` cookie. 
+
+If you've changed the value of your `identifier` cookie, and you refresh the page more than 3 times in 10 seconds, a notification will be displayed to you. 
+
+![Displaying a personal notification](only-sql-results-02.png "Displaying a personal notification")
+
+Another common use case in SQL is that you need join data from multiple sources. In the next example, we will display a notification
+to a user of our website who is calling us. We're mocking that call by using the [simulate call](http://localhost:8080/simulate-call.html) 
+link from the menu on our website. That page will send data to a separate topic which we'll use with our existing website behaviour data. 
+
+The first thing that we'll do is create another table, so we can connect to the data. 
+
 ```sql
 CREATE TABLE text_to_speech (
     `identifier` STRING,
@@ -137,6 +170,15 @@ CREATE TABLE text_to_speech (
 );
 ```
 
+The following SQL statement will do a couple of things:
+
+1. It will create a new notification by insert into our Elasticsearch sink, including a link to a page
+2. For all identifiers that are making a call
+3. While they have been active on the website 10 seconds prior to making the call
+4. If their call is about how to contribute
+
+The third step is using an interval join, which means that the join result is only there if the join condition and a time constraint is met.
+
 ```sql
 INSERT INTO notifications (`identifier`, `notification_id`, `notification_text`, `notification_link`)
 SELECT  
@@ -150,19 +192,4 @@ AND w.ts BETWEEN t.ts - INTERVAL '10' SECONDS AND t.ts
 AND t.category = 'how-to-contribute';
 ```
 
-### For debugging purposes
-
-```bash
-# List all Kafka topics
-docker-compose exec broker kafka-topics \
---list \
---bootstrap-server broker:29092
-```
-
-```bash
-# See data on topic:
-docker-compose exec broker kafka-console-consumer \
---bootstrap-server broker:29092 \
---topic notifications \
---from-beginning
-```
+![Displaying a notification with link](only-sql-results-03.png "Displaying a personal notification on how to contribute")