Java - Processing input and output

PROCESSING INPUT AND OUTPUT
PROGRAMMAZIONE CONCORRENTE E DISTR.
Università degli Studi di Padova
Dipartimento di Matematica
Corso di Laurea in Informatica, A.A. 2015 – 2016
rcardin@math.unipd.it

Programmazione concorrente e distribuita
SUMMARY
 Introduction
 Obtaining streams
 Reading / writing bytes
 Character encodings
 Text input / output
 Random access files
 Processing files
 URLs
 Serialization
2Riccardo Cardin

INTRODUCTION
 Java provides an elegant API to read and write
data in binary and text format
 The API let’s you work in the same way with files,
directories, web pages and so on
 A source from which one can read bytes is called an
input stream
 Bytes can come from a file, a network connection or an array
in memory
 A destination for bytes in an output stream
 In contrast readers and writers consume and produce
sequences of characters
3Riccardo Cardin

OBTAINING STREAMS
 A stream can be obtained from different sources
 From a file
 From an URL
 From an array of bytes
 Once the stream is obtained we can I/O from and to
the source using the same API
4Riccardo Cardin
InputStream in = Files.newInputStream(path);
OutputStream out = Files.newOutputStrem(path);
URL url = new URL("http://rcardin.github.io");
InputStream in = url.openStream();
byte[] bytes = /* ... */;
InputStream in = new ByteArrayInputStream(bytes);
OutputStream out = new ByteArrayOutputStream();
bytes = out.toByteArray();

READING BYTES
 The type to read bytes from a source is
java.io.InputStream
 An input stream can read a single byte
 Cast to byte only after you’ve checked that is not -1
 Or it can read in bulk into a byte array
5Riccardo Cardin
InputStream in = /* ... */;
// A byte is an integer from 0 to 255. -1 is reserved to the
// end of the input stream
int b = in.read();
bytes[] bytes = /* ... */;
// Returns the number of actual read bytes
int actualBytesRead = in.read(bytes);
// Writes the bytes from a position of the array, for at max
// ‘length’ number of bytes
actualBytesRead = in.read(bytes, start, length);

WRITING BYTES
 The type to write bytes to a target is
java.io.OutputStream
 Output stream can write a single byte or a bytes array
 When done, you have to close streams
 Streams implement AutoClosable and so you can use a try-
with-resources statements
6Riccardo Cardin
OutputStream in = /* ... */;
int b = /* ... */;
out.write(b);
byte[] bytes = /* ... */;
out.write(bytes);
out.write(bytes, start, length)
try (OutputStream out = /* ... */) {
out.write(bytes);
}

OBTAINING STREAMS
7Riccardo Cardin

CHARACTER ENCODINGS
 In many cases you will interpret bytes as a
sequence of characters
 How such characters were encoded?
 Java uses Unicode standard for characters
 Each char, or code point, has a 21-bit integer number
 UTF-8 encondes each Unicode code point into a
sequence of one to four bytes
 ASCII character set are represented with only 1 byte
 UTF-16 encodes Unicode code points into one or two
16-bit values
 This encoding is the default used in Java strings
8Riccardo Cardin

CHARACTER ENCODINGS
 There is no realiable way to automatically detect
the encoding from a stream of bytes
 You should always explicitly specify the encoding
using StandardCharsets class
 Use the Charset object when reading or writing text
 If you specify nothing, the computer default charset is used
9Riccardo Cardin
// Values are of type Charset
StandardCharsets.UTF_8
StandardCharsets.UTF_16
StandardCharsets.UTF_16BE
StandardCharsets.UTF_16LE
StandardCharsets.ISO_8859_1
StandardCharsets.US_ASCII
String str = new String(bytes, StandardCharsets.UTF_8)

TEXT INPUT
 To read text input use a Reader
 Obtain a reader from an input stream using an
InputStreamReader decorator
 It is not very convenient to read a char at time
 If you want to read an input line by line, usa the
decorator class BufferedReader
10Riccardo Cardin
// Here you can view the "onion" structure
Reader in = new InputStreamReader(new InputStream(/*...*/, charset);
// Reads a code unit between 0 and 65536, or -1
int ch = in.read();
try (BufferedReader reader =
new BufferedReader(new InputStreamReader(
new InputStream(/*...*/)) {
// A null is returned when the stream is done
String line = reader.readLine();
}

TEXT OUTPUT
 To write text us the class Writer
 To turn an output stream to a writer use
OutputStreamWriter decorator
 The write method writes strings to the stream
 Class PrintWriter adapts the interface of writers to
the print, println and printf used with System.out
 StringWriter lets you to write a stream into a String
 You can combine it with a PrintWriter
11Riccardo Cardin
Writer out = new OutputStreamWriter(
new OutputStream(/*...*/), charset);
out.write("Any string");
PrintWriter wrt = new PrintWriter(out, "UTF-8");
wrt.println("Any string");

TEXT OUTPUT
12Riccardo Cardin

DEALING WITH BINARY DATA
 The DataInput type reads from binary source
 The DataOutput type writes in binary format
 Dealing with binary data is useful beacuse it is fixed in
width and efficient
 No parsing operation is needed
 The classes DataInputStream and
DataOutputStream adapt stream to the interface
13Riccardo Cardin
byte readByte();
char readChar();
int readInt();
long readLong();
// ...
DataInput in = new DataInputStream(Files.newInputStream(path));
DataOutput out = new DataOutputStream(Files.newOutputStream(path));

RANDOM ACCESS FILES
 The type RandomAccessFile lets you read or
write data anywhere in a file
 Implements DataInput / DataOutput interfaces
 Use "r" to open in read mode, "rw" in read-write
 Such a file has a pointer that indicates the position of
the next byte
 Use the seek method to position the pointer inside the file
14Riccardo Cardin
RandomAccessFile file = new RandomAccessFile(path.toString(), "rw");
// Read the next integer in the file
int value = file.readInt();
// Move the pointer
file.seek(file.getFilePointer(), 4);
// Write an integer to the file
file.writeInt(value + 1);
Writing moves the
pointer to the next
sequence of bytes

FILE LOCKING
 If more than a program tries to modify the same
file, it can easily become damaged
 The FileLock solves the problem
 Use the lock method to lock a file
 Use the tryLock method to verify if a file is locked
 Returns null if the lock is not available
 Use a FileChannel to obtain the lock
 The file remains locked until the lock or the channel is closed
 It is best to use try-with-resources
15Riccardo Cardin
FileChannel channel = FileChannel.open(path);
FileLock lock = channel.lock();
// or
FileLock lock = channel.tryLock()

FILE CREATION
 A Path is a sequence of directory names,
optionally followed by a file name
 If the first component is a root element, then the
path is absolute, otherwise is relative
 Root elements may be «/» or «C:»
 The Paths class is a companion class with a lot of utilities
 An InvalidPathException will be thrown if the path is
not valid in the given filesystem
 A Path does not have to correspond to a file that
actually exists
16Riccardo Cardin
Path abs = Paths.get("/", "home", "rcardin");
Path rel = Paths.get("home", "rcardin");
Path another = Paths.get("/home/rcardin");

FILE CREATION
 Create a directory
 Create only the last part of the path
 Create also intermediate directory as well
 Create a file
 Throws an exception if the file already exists
 Check for existence and creation are an atomic operation
17Riccardo Cardin
Files.createDirectory(path);
Files.createDirectories(path);
Files.createFile(path);
// Checks if a file already exists
Files.exists(path);
Files.isDirectory(); // Test if the path is a directory
Files.isRegularFile(); // Test if the path is a regular file

FILE CREATION
 It’s possible to create temporary file / folders
 Names of temp files and folders are generated
randomly
 It is possible to give a prefix, a suffix and a target path
 If no path is given, the default SO temporary folder is used
 The file / folder will not be deleted automatically on
JVM termination
 Call File.deleteOnExit() or specify the attribute
StandardOpenOption.DELETE_ON_CLOSE during creation
18Riccardo Cardin
Files.createTempFile(dir, prefix, suffix);
Files.createTempDirectory(dir, prefix)
Files.createTempDirectory(dir, prefix,
StandardOpenOption.DELETE_ON_CLOSE);

OPERATIONS ON FILES
 Abstractions on file serve to allow you to
operate on them
 It’s possible to copy or move a file or an empty dir.
 If the StandardCopyOperation.REPLACE_EXISTING is
not set, the above operations will fail if the target exists
 With StandardCopyOperation.REPLACE_EXISTING you
can specify to maintain the source file if the operation fails
 It’s possible to delete a file or an empty directory
19Riccardo Cardin
Files.copy(fromPath, toPath);
// Moves a file or an empty directory
Files.move(fromPath, toPath);
Files.delete(path)
boolean deleted = Files.deleteIfExists(path)

OPERATIONS ON FILES
 It is possible to walk through a file tree
 Use the walkFileTree method
 The FileVisitor interface defines method that are
called prior, during and post directory visit
 The class SimpleFileVisitor gives a wrapper
implementation
 The FileVisitResult defines how to continue the walking
process
20Riccardo Cardin
// The second argument is an implementation of FileVisitor
Files.walkFileTree(startingDir, fv);
FileVisitResult postVisitDirectory(T dir, IOException exc)
FileVisitResult preVisitDirectory(T dir, BasicFileAttributes attrs)
FileVisitResult visitFile(T file, BasicFileAttributes attrs)
FileVisitResult visitFileFailed(T file, IOException exc)

OPERATIONS ON FILES
21Riccardo Cardin
public static class PrintFiles extends SimpleFileVisitor<Path> {
// Print information about each type of file.
@Override
public FileVisitResult visitFile(Path file,
BasicFileAttributes attr) {
if (attr.isSymbolicLink()) {
System.out.format("Symbolic link: %s ", file);
} else if (attr.isRegularFile()) {
System.out.format("Regular file: %s ", file);
} else {
System.out.format("Other: %s ", file);
}
return CONTINUE;
}
// Print each directory visited.
@Override
public FileVisitResult postVisitDirectory(Path dir,
IOException exc) {
System.out.format("Directory: %s%n", dir);
return CONTINUE;
}
}

OPERATIONS ON FILES
22Riccardo Cardin

URL CONNECTIONS
 The simplest method to read from an URL is to
get an input stream from it
 Using URLConnection it is possible to retrieve
additional resources from URL and to write to it
 The type has a subtype for each type of URL
 Send data to the server using an output stream
23Riccardo Cardin
// Opens an input stream on the URL
InputStream in = url.openStream();
// From an HTTP URL returns an HttpURLConnection
URLConnection con = url.openConnection();
con.setDoOutput(true);
try (OutputStream out = con.getOutputStream()) {
// Write to out
}

SERIALIZATION
 Representation of an object as a sequence of
bytes, that includes data as well as type
 The process is JVM independent
 The type must implement java.io.Serializable
 Marker interface with no methods
 All fields in the class must be serializable. If a field is not, it
must be marked as transient
 Transient fields are not serialized and are lost intentionally
 A transient variable cannot be final or static
24Riccardo Cardin
public class Employee implements Serializable {
private String name;
private double salary;
// ...
}

SERIALIZATION
 Serialization process uses the stream API
 To serialize objects, use an ObjectOutputStream
 The writeObject method serializes the object
 To deserialize objects, use an ObjectInputStream
 The readObject does the magic
 Serialization process is recursive on object attributes
 Name of class and name / values of attributes are saved
25Riccardo Cardin
ObjectOutputStream out =
new ObjectOutputStream(Files.newOutputStream(path));
Employee paul = new Employee("Paul", 29000D);
out.writeObject(paul);
ObjectInputStream in =
new ObjectInputStream(Files.newInputStream(path));
Employee paul = (Employee) in.readObject();

SERIALIZATION
 Serialization works fine with duplicated objects
 Each object gets a serial number when it is saved
 An ObjectOutputStream checks if an object was previously
written with the same serial number. In that case, it just write out
the serial number
 The ObjectInputStream works conversely
 It is possible to tweak the serialization process
 Redefine readObject and writeObject methods
 Use defaultWriteObject and defaultReadObject on streams
26Riccardo Cardin
// Used to deserialize an object
private void readObject(ObjectInputStream in)
throws IOException, ClassNotFoundException
// Used to serialize an object
private void writeObject(ObjectOutputStream out) throws IOException

SERIALIZATION
 Versioning
 If you use serialization for long term persistence, you
need to consider what happens when classes evolve
 How do we deserialize objects into new versions of a class?
 Serialization supports versioning
 Assign a serialVersionUID to a Serializable class. The
version uid is written with other information
 When the class change in an incompatible way, modify the
version uid
 If uids are different, an InvalidClassException is thrown
 Default uid is generated from hash code of the class
27Riccardo Cardin
Private static final long serialVersionUID = 1L;

SERIALIZATION
28Riccardo Cardin

EXAMPLES
29Riccardo Cardin
https://github.com/rcardin/pcd-snippets

REFERENCES
 Chap. 9 «Processing Input and Output», Core Java for the
Impatient, Cay Horstmann, 2015, Addison-Wesley
 Walking the File Tree
https://docs.oracle.com/javase/tutorial/essential/io/walk.html
30Riccardo Cardin

Java - Processing input and output

Recommended

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Java - Processing input and output (20)

More from Riccardo Cardin (7)

Recently uploaded (20)

Java - Processing input and output