Go to TextAnalyzer Client.
This is the server application of my TextAnalyzer program. The server opens a socket to accept incoming connections from the client application.
When run, the server is ready to accept connections from the client application. Once it receives a URL from the client application, the server will parse the URL and send back the statistics to the client application, which will then display the results in a GUI (Graphical User Interface).
For this exercise, the TextAnalyzer program was split into two parts: a client application and this server application. The client application presents the user with the GUI, interacts with the server application to process the URL submitted by the user, and then displays the data processed and sent back by the server.
For additional code design notes, see the TextAnalyzer Code Desing section.
The client and server applications were built upon the existing TextAnalyzer code base. The program was not developed initially using TDD (Test Driven Development) standards. In other words, the code and expected output were not tested formally. Testing was done incrementally, that is, after each completed minor change, typically after each line of code was added, mainly to catch compile-time and runtime errors not previously detected and highlighted by the IDE.
As of version 1.7, the program includes unit tests created using Junit.
The program is a JavaFX application using version 8 of Amazon's distribution of the Open Java Development Kit (OpenJDK) Corretto 8, which includes JavaFX 8. Unit tests were created using Junit 5.
The program requires the Jsoup Java HTML parsing library.
The program assumes that an existing database user with all database privileges in the local MySQL database with username/password: textanalyzer/textanalyzer. It also assumes that it will connect to localhost using default port 3306. However, the user may change these initial configuration options by editing lines 32 to 35 in the Database class:
/src/gce/textanalyzer/controller/Database.java
32 String databaseHost = "localhost";
33 String databasePort = "3306";
34 String databaseUser = "textanalyzer";
35 String databasePass = "textanalyzer";
The server application requires no user interaction other than compilation adn execution. It must be run before the client application runs.
Option 1. Clone the repo and import it into your favorite Java IDE. Make sure that:
- The project SDK is Java 8 with project language level 8, and
- JavaFX 8, Junit 5, Jsoup, and the Oracle JDBC Driver for MySQL are installed in your system.
Option 2. Download and run TextAnalyzerServer.jar.
IMPORTANT: You must run this server BEFORE you run TextAnalyzerClient.
Not all HTML files are created equal. The program's code to convert HTML files to plain text is still rudimentary, as it is unable to identify HTML tag properties in lines without opening or closing tags.
For example, in the following code:
<div class="Popover anim-scale-in js-tagsearch-popover"
hidden
data-tagsearch-url="/qstainless/TextAnalyzer/find-symbols"
data-tagsearch-ref="master"
data-tagsearch-path="README.md"
data-tagsearch-lang="Markdown">
</div>The following will be considered as 'words' by the program because the lines do not begin with "<" nor end with ">."
hidden
data-tagsearch-url="/qstainless/TextAnalyzer/find-symbols"
data-tagsearch-ref="master"
data-tagsearch-path="README.md"That is because the program parses the target URL line by line. Lines that begin with "<" or end with ">" are ignored for purposes of counting words in them.
To resolve this issue and to avoid having to reinvent the wheel, I re-implemented the Jsoup library, which creates a clean HTML version of the target URL.
The server/client run from an IDE in macOS, Windows 10 Pro and Linux (Ubuntu 18.04). The client .jar file, however, only runs properly in macOS and Windows 10 Pro. Searching for a solution, Linux would require that Openjfx be installed alongside the JDK (in my case, Amazon Corretto 8). Unfortunately, installing Openjfx was not enough. I may revisit this in the future.
Execution time was drastically different between macOS (~12s) and Windows 10 Pro (~73s). Because I don't typically develop in a Windows environment, I can't explain the difference. The screenshot shows the server and client timestamps and shows the schema and tables properly created.
Refactor the storeWordsIntoDatabase method to detect lines that begin with "<" and ignore all following lines up to and including the next line with a closing ">." Because the parser will ignore the lines, there will be no need to call the htmlToText method. The Jsoup library seems like a viable solution.
Find a way to run the client .jar file in Linux.
The version numbering of this project does not follow most version numbering guidelines. Instead, it is limited to a two-token concept:
(major).(course module)
Version 1.11 (current) - Converted program into server/client
Version 1.10 - Added database support
Version 1.9 - Added JavaDocs
Version 1.7 - Added unit tests using JUnit
Version 1.6 - Added GUI functionality
Version 1.2 - First version
