New Relic Coding Challenge

Solution

Assumptions

• The text or file provided won't have words that will be followed by_ (bug on the reguler expression)
• Considering the following the chars as being used by contractions ( '’‘ )

Build

Requirements

• Java JDK version >=1.8 <=11.x

How to build the project and generate a jar

• Build project.

$ ./gradlew build

• Create shadow JAR file.

$ ./gradlew shadow

Run Application

• Run application from jar file generated on the previous step. (Please use java versions between 8 and 11)

• For file input please run:

java -jar ./build/libs/NRchallenge.jar moby_dick.txt 

• For Stdin input please run:

cat moby_dick.txt | java -jar ./build/libs/NRchallenge.jar 

• Once application has finished processing succesfully you will see the list of 3 words sequences and its respective counts, for example:

The count will be done over the stdin input
of the whale - 93

the sperm whale - 89

the white whale - 76

#Run Tests

The small subset of tests was included to test parts of the functionalities

./gradlew clean test

> Task :test

com.nr.challenge.ParserTest > TestFileParserInvalidFile PASSED

com.nr.challenge.ParserTest > TestUnwantedCharsContractions2 PASSED

com.nr.challenge.ParserTest > TestUnwantedCharsTrim PASSED

com.nr.challenge.ParserTest > TestStringParser PASSED

com.nr.challenge.ParserTest > TestUnwantedCharsContractions PASSED

BUILD SUCCESSFUL in 2s
3 actionable tasks: 2 executed, 1 up-to-date

What could have been done better

• BUG -> Regular expression could be tuned to handle all special chars. (related to words that start or end with _)
• BUG -> There is no continuity when there is more than 1 file passed as arguments. So in this scenario, it is not considering the last sequences of words when a file ends and another starts (it is working as expected when the input comes from stdin)
• Performance could have been improved.
• There could be more unit tests (including one to validate unicode chars).