CSE 11 Programming Assignment 7

String Manipulation, Command line, and File I/O

Due Date: Monday, August 29, 10:00PM Pacific Time

In this assignment you will put together many of the programming features and techniques you’ve seen so far to build a useful command-line application for text manipulation.

This program is larger than the ones you’ve written so far for this course, so you should start early so you understand the scope. Use the implementation suggestions and milestones to help you get started and work incrementally.

On this assignment, we encourage you to share publicly and with other students what you think the expected output should be on particular examples. For example, you could share a java StringSearch ... command you tried out, and show the results, and check with other students if they agree on the behavior. This allows you to discuss how the assignment is supposed to work without sharing any code, and you also might want to share examples you found interesting!

We ask that you not share your code in public posts on Piazza, or in general with other students, on this PA. It’s an important PA to work through the code yourself to get the most learning out of it.

Submission checklist:

Starter code here:

https://github.com/ucsd-cse11-su222/cse11-pa7-starter

FAQ

Q: My gradescope submission is not compiling for StringSearchMilestone1/2/3/4.

Double check the name of the class. The class name should not be StringSearchMilestone1/2/3/4 but rather just StringSearch.

Q: My queries and/or transforms worked fine locally but failed on the autograder.

The autograder checks for sample tests provided in the writeup as well as some more interesting cases. Try to come up with more thorough test cases when testing locally!

Q: What does “a query of any type” in Milestone3 mean?

This means any single query (length, greater, less, contains, starts, ends, not) could be used to test the Milestone3 code. An example of a test that would not be run on Milestone3 but could be run on the final completed task is:

```
$ java StringSearch "poem.txt" "greater=13&starts='This'"
This is a short file
```

Q: What if there are multiple transforms?

The transforms should be applied in order, from left to right.

Q: Can multiple queries be included within a not query like not(greater=3&less=100)?

No, only a single query is included within the parentheses for not.

Q: Can there be a transform without there being a query?

No, if there’s a transform there will also be a query

Q: If a replace-able string overlaps, for example replacing aaa with b in aaaa, which should happen?

Replace the earliest occurrence in the part that overlaps. So this would produce ba. Check what the replace method on strings does, as well, to see if that does what you need before making it more complicated for yourself!

Task

The task is similar to the famous programs grep and sed.

In a file called StringSearch.java, you’ll write a class StringSearch with a main method that uses command-line arguments as described below. You can write as many additional methods and classes as you wish, and use any Java features you like. We have some suggestions in the program structure section later on that you can use, or not use, as you see fit.

Clarification: If just a file is provided, the program should print the file's entire contents, and if just a file and a query are provided with no transform, just the matching lines should print (see examples below).

The main method should expect 3 command-line arguments:

$ java StringSearch "<file>" "<query>" "<transform>"

The overall goal of StringSearch is to take a file of text, search for lines in the file based on some criteria, then print out the matching lines after transforming them somehow.

The <thing> syntax means, as usual, that we will be describing what kinds of syntax can go in each position in more detail.

Queries

The <query> part of the command-line should be a &-separated sequence of individual queries. The individual queries are:

Transforms

The <transform> part of the command-line should be a &-separated sequence of individual transforms. The individual transforms are:

Where you see <string> above, it should always be characters inside single quotes, like 'abc'. We chose this because it works best with command-line tools.

Where you see <number> above, it should always be a positive integer.

The <file>, <query>, and <transform> command-line arguments should always be inside double quotes. This ensures that they won’t be interpreted as commands, or parts of commands, by your terminal.

Examples

Poem by tutor Rachel Lim

The file poem.txt contains the following content:

This is a short file
It contains text and this is
Also a haiku

The file words contains a standard dictionary (it’s from Joe’s computer’s MacOS installation).

The following commands, when run at the command line, should produce the given outputs.

$ java StringSearch "poem.txt"
This is a short file
It contains text and this is
Also a haiku
$ java StringSearch "poem.txt" "greater=13"
This is a short file
It contains text and this is
$ java StringSearch "poem.txt" "not(contains='short')"
It contains text and this is
Also a haiku
$ java StringSearch "poem.txt" "greater=13&starts='This'"
This is a short file
$ java StringSearch "poem.txt" "contains='his'" "last=10"
short file
nd this is
$ java StringSearch "poem.txt" "contains=' a '" "upper&first=18"
THIS IS A SHORT FI
ALSO A HAIKU
$ java StringSearch "poem.txt" "greater=3&less=100&not(ends='z')" "replace='i';'I'"
ThIs Is a short fIle
It contaIns text and thIs Is
Also a haIku
$ java StringSearch "poem.txt" "greater=3&less=100&not(ends='u')" "replace='i';'I'"
ThIs Is a short fIle
It contaIns text and thIs Is
$ java StringSearch "words" "contains='no'&starts='x'&not(contains='xeno')" "lower"
xanthocyanopsia
xanthocyanopsy
xanthocyanopy
xanthomelanous
xoanon
xylenol
xyloquinone
xylorcinol

Milestones

You must submit the following milestones with your submission. You should save them in files called StringSearchMilestone1.java, StringSearchMilestone2.java, and so on. This serves to help you break the program into small chunks of progress, and also when you come to help hours you can let us know which milestone you are working on.

We recommend working on the first milestone directly in StringSearch.java. Once it’s working, you can save to the StringSearchMilestone1.java file to record your progress and keep working in StringSearch.java. You can then work towards the second, copy your work over when done with that milestone, and so on to build up your submission.

You can upload any of the milestones to Gradescope to get grading feedback on them. If you finish multiple milestones at once, you can always copy your more advanced solution as the earlier milestone files – you’re only adding, not changing features as you go through the milestones. We do ask that you submit all the milestone files because it makes things much more consistent for us to review across submissions.

Milestone 1

Your program should take in one argument that is the name of the file to read, and print out all the lines in that file in order.

Milestone 2

Your program should take in the name of the file and a single contains query, and print all the lines that match that single contains query.

Milestone 3

Your program should take in the name of the file and a single query of any type, and print all the lines that match that single query.

Milestone 4

Your program should take in the name of the file and a single query of any type and a single transform of any type, and print all the lines that match that single query, transformed by that single transform.

After milestone 4, complete the full task as described above.

Suggestions

Implementation Ideas

You may find it useful to write or use some of the following components while implementing your program. These are not required, and you may find only some of them useful to you, or you might find all of them useful to you. We do think they are all useful, so we aren’t trying to trick you or anything 🙂. If you do choose some other approach, during help hours we may guide you towards these if we think your choices won’t lead to a successful design.

Reading Files

Part of this PA requires that you read in the contents of a file. You can include this code snippet at the top of your file. There are one or two things in this file-reading code that are new-ish Java features. You can use the method getLines without understanding them all. We will discuss them by the end of the quarter. You can also use the .split("\n") strategy from class if you’d rather use all code that we’ve seen before; that will work as well.

import java.nio.file.*;
import java.io.IOException;

class FileHelper {
  /*
    Takes a path to a file and returns all of the lines in the file as an
    array of strings, printing an error if it failed.
  */
  static String[] getLines(String path) {
    try {
      return Files.readAllLines(Paths.get(path)).toArray(String[]::new);
    }
    catch(IOException e) {
      System.err.println("Error reading file " + path + ": " + e);
      return new String[]{"Error reading file " + path + ": " + e};
    }
  }
}

Submission and Check-off

Submit your 5 code files (4 milestones and final StringSearch.java) to pa7 on Gradescope.

There will be an automatic set of tests that run on Gradescope while the assignment is out.

Extensions

These are not for credit, but you may find them interesting to try on your own.

  1. Add a new transform of your own design.

  2. Add a new query of your own design.

  3. All of the queries above are joined in the style of AndQuery from the ImageQuery reading, where they all need to be true to match a line. Extend your implementation to accept both & and | as separators between queries, where | indicates joining queries by or rather than and. Describe the design and details of how to read in and interpret a mix of & and | within a query to a user.

  4. Support multiple queries (including more not queries) within the parentheses for not.