Today: File I/O patterns

In a nutshell

Today is the hinge of Week 7. Monday and Tuesday set up the story (read a file into an array, do array work, write the result) and introduced new Scanner(new File(path)). Today we add the second half of that arc: choosing the right read method when the file has mixed types, walking the file in two passes to size and fill a fixed array, and writing a result back out with PrintStream. By the end of class you have everything you need to start Lab 11.

One-sentence summary. A File is a path, not the contents; a Scanner does the reading, a PrintStream does the writing, and count-allocate-fill is how we turn a file of unknown length into a fixed-size array.

By the end of today you will be able to:

• Order the hasNext* cascade correctly (int before double before general token) and explain why.
• Recognize the leftover-newline trap on sight and apply the throwaway sc.nextLine() fix.
• Construct a file PrintStream with new PrintStream(new File(path)) and explain what close() does that nothing else does.
• Walk a file twice (count, then allocate, then fill) to load an unknown-length file into a fixed array.
• Read the Lab 11 file format and identify which next* method each column needs.

The class flow

1.  Recap: search and sort (Thursday's quiz)
2.  New today: File is a path (one new thing)
3.  Four-move File lifecycle (null/blank check, exists check, wrap, close)
4.  The hasNext cascade table
5.  Count-allocate-fill: the two-pass pattern
6.  Lab 11 file format
7.  countLines, in English then in code
8.  main: the orchestration scaffold
9.  The three fillArray overloads
10. PrintStream: writing a result
11. Lab 11 launch

Today’s objectives (Day 3)

These map to objectives O3.0 through O3.6 in the week’s contract.

Today’s lessons on the site

The slides cover the live class flow; the site lessons are the searchable, re-readable version of the same material. Read in this order before lab time tonight.

• 7c: The next* Cascade and the nextLine Trap (~18 min): choosing the right read method, the hasNextInt before hasNextDouble order rule, and the leftover-newline trap.
• 7d: Writing Files with PrintStream (~13 min): System.out is already a PrintStream; pointing one at a file is the only new step; closing matters.
• 7e: Count-Allocate-Fill (~18 min): the two-pass pattern for loading an unknown-length file into a fixed array. We touch this today to launch Lab 11; we will return to it Friday.

Lab 11 launches today

Lab 11 is the first lab where the input comes from a file. The spec uses every pattern we cover today: a File path argument, a Scanner-on-File loop, the next* cascade for mixed columns, count-allocate-fill to size the array, and a PrintStream for the output report. Read 7c, 7d, and 7e tonight, then start the lab.

This week at a glance

The week’s load-bearing skills

These are the items that lab tests, quiz questions, and exam problems will all draw from:

1. new Scanner(new File(path)) for the read, new PrintStream(new File(path)) for the write. One throws FileNotFoundException on main covers both.
2. The hasNext* cascade in the right order: int → double → general token. The strictest check goes first.
3. The leftover-newline trap fix: one throwaway sc.nextLine(); between any nextInt/nextDouble and a following nextLine.
4. Count-allocate-fill: pass 1 counts tokens with a fresh Scanner, pass 2 fills a freshly allocated array with another fresh Scanner.
5. Close every Scanner and every PrintStream. The empty-output-file bug is almost always a missing close().
6. The APE four-step pattern (validate, operate, handle errors, update state). Tag the spec sentences before writing code.

Common pitfalls already showing up in office hours

• new File(path) does not open the file. It holds a path string. Reading happens when the path is handed to a Scanner.
• Working-directory trap. new File("scores.txt") resolves against the JVM launch directory, not the source file directory. Use getAbsolutePath() to debug.
• Forgetting throws FileNotFoundException. The compiler error names the missing keyword. The fix is one declaration on main, not a try/catch (that is CSCD 211).
• hasNextDouble before hasNextInt. Every int passes hasNextDouble, so the cascade swallows ints as doubles. Order matters.

After class

1. Read 7c, 7d, and 7e in full.
2. Start Lab 11. The countLines method is the first one to write; the spec is a direct application of the count phase of count-allocate-fill.
3. Drill the array-tracing problems in lesson 6f for Thursday’s quiz.