Private Investigator is a command line application that reads the input text file and generates the investigation report in the output file.
The main class PrivateInvestigatorApplication implements the CommandLineRunner interface and is responsible for starting the Spring application. The application reads in a list of sentences from a file specified in inputFilePath property, passes the sentences to PrivateInvestigator for processing, and writes the results to an output file specified in outputFilePath property.
PrivateInvestigator has several dependencies including InvestigationResultHolder, Sentences, and UUIDGenerator. The initiateInvestigation method in PrivateInvestigator loops through each sentence in the list and performs a validation check. If a sentence is valid, it creates a SentenceHolder object with the sentence, and passes it to the check method in InvestigationResultHolder. The check method generates a unique key based on the words in the sentence, with one word skipped, and maps the key to the sentence's ID.
It updates the similarSentences which is a list of similarSentenceMapKey and similarSentenceMapping which is a global Map of mapping between similarSentenceMapKey and Set of sentence IDs. similarSentenceMapKey is the concatenated value of unique Sentence combination with one word skipped and the index of the skipped word using pipe symbol |.
InvestigationResultHolder keeps track of similar sentences by mapping their keys to the sentence IDs. When all sentences have been processed, the writeResults method in InvestigationResultFormatter formats the results and writes them to an output file. The results consist of a string with expected formatted output.
Ex:
01-01-2012 19:45:00 Naomi is getting into the car
02-01-2012 09:13:15 George is getting into the car
03-01-2012 11:22:40 Mike is getting into the car
The changing word was: Naomi, George, Mike
-
Navigate to the project directory:
cd private-investigator -
Run the following command to build the project:
./gradlew clean build -
Once the build is successful, run the following command to start the application:
./gradlew :bootRun
The code has a time complexity of O(n^2) as it loops through each sentence in the input and then loops through each possible combination of words in a sentence. The space complexity is O(n) as it uses a HashMap and a Set to store the mapping of sentence combinations with sentence IDs and a set of unique sentence combinations, respectively.
-
Load Balancing: If the application is receiving a large number of requests, we can distribute the load among multiple instances of application running in different servers. This will help reduce the burden on a single server and increase the overall capacity of the application.
-
Caching: Implementing caching can significantly improve the performance of the application. We can push the processed result into the cache and process those messages via a cron job which can read the messages from cache, writes to a file or database and then deletes it from the cache.
-
Messaging: Rather than processing all the input at once. We can push the input data into a messaging system such as Kafka or RabbitMQ and process those messages at fixed rate.
-
Parallel Processing: We can divide the workload into smaller parts and process them in parallel. This can help reduce the overall processing time and increase the speed of the application.
3. If you had two weeks to do this task, what would you have done differently? What would be better?
I would integrate the system with a database and store the results in a table Expose API to upload input file(s), initiate the investigation and write the results.
It would have been more better to integrate it with messaging system, a cron job and have a listener. Listener function would push the input data to a topic if there will be any change in the input file and the cron job will process the input from the same topic and update the result in a database or a file.
input.txtis located in thesrc/resource/data/directory. The input file contains a list of sentences ordered by time, one sentence per row. Each sentence has a date-time stamp and a description of the activity being performed by the person being followed.
The application will run and generate the report in the
output.txtfile located in thesrc/resource/data/directory.