Transcript Document

C# File I/O
Line by line and token-based file input
Copyright 2008 by Pearson Education
Input/output (I/O)
using System.IO;
 Create a FileInfo object to get info about a file on disk.
(This doesn't actually create a new file on the hard disk.)
FileInfo f = new FileInfo(“Files\\example.txt");
if (f.Exists && f.Length > 1000) {
f.Delete();
}
Method name
Description
Delete()
removes file from disk
MoveTo(filename)
moves and/or changes name of file
Property name
Description
Exists
whether this file exists on disk
Name
returns file's name
Length
returns number of bytes in file
Copyright 2008 by Pearson Education
2
Type of Files
 Binary files
 http://en.wikipedia.org/wiki/Binary_file
 Can contain arbitrary patterns of bits (data)
 We will NOT be dealing with these
 Include program .EXE's, .CLASS files, image files (PNG, GIF,
JPEG, etc), video files, etc, etc.
 Text Files
 http://en.wikipedia.org/wiki/Text_file
 Have ONLY text inside them
 Includes anything you made using Notepad, source code files
(i.e., files ending in .cs, .java, .cpp), HTML web pages (.html),
XML, most-if-not-all Unix/Linux configuration files, etc, etc
Copyright 2008 by Pearson Education
3
Reading files
 To read a file, call OpenText on a FileInfo object.
FileInfo file = new FileInfo("Files\\example.txt");
TextReader name = file.OpenText();
or, even better:
TextReader name = new
StreamReader("Files\\example.txt");
Copyright 2008 by Pearson Education
4
Closing files
 You will need to dispose of the file when you’re done with it:
TextReader name = new StreamReader("Files\\example.txt");
// The program reads the file here
name.Dispose();
 C# provides using, which will call dispose for you:
using( TextReader name = new
StreamReader(@"Files\example.txt")
{
// The program reads the file here
}
)
 name.Dispose() called automatically on exit, no matter how you exit
(reach the closing }, return statement, throw an exception, etc, etc)
Copyright 2008 by Pearson Education
5
File paths
 absolute path: specifies a drive or a top "/" folder
C:/Documents/smith/hw6/input/data.csv
 Windows can also use backslashes to separate folders.
 relative path: does not specify any top-level folder
names.dat
input/kinglear.txt
 Assumed to be relative to the current directory:
FileInfo file = new FileInfo("data/readme.txt");
If our program is in
H:/hw6,
FileInfo will look for H:/hw6/data/readme.txt
 You can use
.. to move up a folder level:
with ../hw3/data/readme.txt
FileInfo will look for H:/hw3/data/readme.txt
Copyright 2008 by Pearson Education
6
Possible error w/ files
 The following program crashes when run:
public void Slide_07()
{
using (TextReader t = new
StreamReader("missing_file.txt") )
{
string s = t.ReadLine();
}
}
 The following error occurs:
Unhandled Exception: System.IO.FileNotFoundException: Could not find
file 'E:\path\to\program\missing_file.txt'.
File name: 'E:\path\to\program\missing_file.txt'
at System.IO.__Error.WinIOError(Int32 errorCode, String
maybeFullPath)
< snip – extra stuff removed for clarity  >
at PCE_StarterProject.SlideExamples.Slide_05() in
E:\path\to\program\Student_Answers.cs:line 76
at PCE_StarterProject.Program.Main(String[] args) in
E:\path\to\program\Student_Answers.cs:line 30
Copyright 2008 by Pearson Education
7
Exceptions
 exception: An object representing a runtime error.

dividing an integer by 0

calling charAt on a String and passing too large an index

trying to read a file that does not exist
 We say that a program with an error "throws" an exception.
 It is also possible to "catch" (handle or fix) an exception.

We will not be catching any exceptions in this topic
Copyright 2008 by Pearson Education
8
Input tokens
 token: A unit of user input, separated by whitespace.
 We can split a file's contents into tokens.
 If an input file contains the following:
23
3.14
"John Smith"
We can interpret the tokens as the following types:
Token
23
3.14
"John
Smith"
Copyright 2008 by Pearson Education
Type(s)
int, double, String
double, String
String
String
9
Line-Based File Parsing
 Consider a file numbers.txt that contains this text:
 (Note that ˽ means a blank space, \t means a tab)
308.2
˽˽˽˽7.4
3.9˽˽˽
˽-15.4˽˽˽
 We can think of the file's contents like so:
308.2\n˽˽˽˽7.4\n3.9˽˽˽\n\t˽-15.4˽˽˽\n\n
 We can get each line (using the TextReader.ReadLine()),
then deal with each line individually
Copyright 2008 by Pearson Education
10
File input question
 Recall the input file numbers.txt:
308.2
˽˽˽˽7.4
3.9˽˽˽
\t˽-15.4˽˽˽
 Write a program that reads the lines from the file and
prints each one followed by a < sign (so it's clear where the
line ends)(note the blank line at the end of the output)
308.2<
7.4<
3.9
<
-15.4
<
<
Copyright 2008 by Pearson Education
11
File Input Answer
 public void Slide_12()
{
using (TextReader file = new
StreamReader("Files/numbers.txt"))
{
for (int i = 0; i < 5; i++)
{
string sLine = file.ReadLine();
Console.WriteLine("{0}<", sLine );
}
}
}
Copyright 2008 by Pearson Education
12
File input question
 Recall the input file numbers.txt:
308.2
˽˽˽˽7.4
3.9˽˽˽
\t˽-15.4˽˽˽
 Write a program that reads the values from the file and
prints them along with their sum.
number = 308.2
number = 7.4
number = 3.9
number = -15.4
Sum = 304.1
Copyright 2008 by Pearson Education
13
Testing for valid input
 All basic data types offer a version of TryParse:
Method
Int32.TryParse
Double.TryParse
Description
http://msdn.microsoft.com/enus/library/f02979c7.aspx
http://msdn.microsoft.com/enus/library/994c0zb1.aspx
short.TryParse
aka
Int16.TryParse
http://msdn.microsoft.com/enus/library/9hh1awhy.aspx
 If given a valid string (Int32, Double, short, etc) each one will change
it's second parameter to be that value and return true
 If given an invalid string (Int32, Double, short, etc) each one will
change it's second parameter to be zero and return false
Copyright 2008 by Pearson Education
14
File Input Answer
 public void Slide_15() {
using (TextReader file = new
StreamReader("Files/numbers.txt"))
{
double sum = 0.0;
for( int i = 0; i < 5; i++)
{
string sLine = file.ReadLine();
double dNum;
if (Double.TryParse(sLine, out dNum))
{
sum += dNum;
Console.WriteLine("number = {0}", dNum);
}
}
Console.WriteLine("Sum = {0}", sum);
}
}
Copyright 2008 by Pearson Education
15
Reading an entire file
 Suppose we want our program to process the entire file.
(It should work no matter how many values are in the file.)
number = 308.2
number = 7.4
number = 3.9
number = -15.4
number = 4.7
number = 5.4
number = 2.8
Sum = 317
Copyright 2008 by Pearson Education
308.2
˽˽˽˽7.4
3.9˽˽˽
\t˽-15.4˽˽˽
4.7
˽5.4
2.8
16
Reading an entire file answer
 public void Slide_17()
{
using (TextReader file = new
StreamReader("Files/numbers2.txt"))
{
double sum = 0.0;
string sLine;
sLine = file.ReadLine();
while (sLine != null)
{
double dNum;
if (Double.TryParse(sLine, out dNum))
{
sum += dNum;
Console.WriteLine("number = {0}", dNum);
}
sLine = file.ReadLine();
}
Console.WriteLine("Sum = {0}", sum);
}
}
Copyright 2008 by Pearson Education
17
File input question 3
 Test the program to make sure that it handles files that
contain non-numeric tokens (by skipping them).
 For example, it should produce the same output as before
when given this input file, numbers3.txt:
308.2 hello
oops 7.4
bad 3.9 stuff
-15.4
4.7
5.4
:-) 2.8 @#*($&
Copyright 2008 by Pearson Education
308.2˽hello
˽˽˽˽oops˽7.4
bad˽3.9˽stuff
-15.4
4.7
5.4
:-)˽2.8˽˽@#*($&
18
File input answer 3
 public void Slide_19() {
char[] delimiters = { ' ', '\t' };
using (TextReader file = new StreamReader("Files/numbers3.txt"))
{
double sum = 0.0;
string sLine = file.ReadLine();
while (sLine != null) {
string[] tokensFromLine = sLine.Split(delimiters,
StringSplitOptions.RemoveEmptyEntries);
foreach (string token in tokensFromLine) {
double dNum;
if (Double.TryParse(token, out dNum)) {
sum += dNum;
Console.WriteLine("number = {0}", dNum);
break; // out of foreach
}
}
sLine = file.ReadLine();
}
Console.WriteLine("Sum = {0}", sum);
}
}
Copyright 2008 by Pearson Education
19
Election question
 Write a program that reads a file poll.txt of poll data.
 Format: State Obama% McCain% ElectoralVotes Pollster
CT 56 31 7 Oct U. of Connecticut
NE 37 56 5 Sep Rasmussen
AZ 41 49 10 Oct Northern Arizona U.
 The program should print how many electoral votes each
candidate leads in, and who is leading overall in the polls.
Obama: 7 votes
McCain: 15 votes
Overall: McCain
Copyright 2008 by Pearson Education
20
Election answer
 public void Slide_21()
{
char[] delimiters = { ' ', '\t' };
int obamaVotes = 0, mccainVotes = 0, eVotes = 0;
using (TextReader t = new StreamReader("Files/poll.txt"))
{
string sLine = t.ReadLine();
while (sLine != null)
{
string[] tokens = sLine.Split(delimiters,
StringSplitOptions.RemoveEmptyEntries);
if (tokens.Length < 5)
{
Console.WriteLine("Did not find the expected number
of items on line:\n\t{0}", sLine);
sLine = t.ReadLine();
continue;
}
Copyright 2008 by Pearson Education
21
Election answer

int obama;
if (!Int32.TryParse(tokens[1], out obama))
{
Console.WriteLine("Obama's votes not formatted
properly on line\n\t{0}", sLine);
sLine = t.ReadLine();
continue;
}
int mccain;
if (!Int32.TryParse(tokens[2], out mccain))
{
Console.WriteLine("McCain's votes not formatted
properly on line\n\t{0}", sLine);
sLine = t.ReadLine();
continue;
}
if (!Int32.TryParse(tokens[3], out eVotes))
{
Console.WriteLine("Electoral votes not formatted
properly on line\n\t{0}", sLine);
sLine = t.ReadLine();
continue;
}
Copyright 2008 by Pearson Education
22
Election answer
if (obama > mccain)
obamaVotes += eVotes;
else if (mccain > obama)
mccainVotes += eVotes;
// on tie neither candidate gets the votes

sLine = t.ReadLine();
} // end of while
}
Console.WriteLine("Obama: {0} votes", obamaVotes);
Console.WriteLine("McCain: {0} votes", mccainVotes);
if (obamaVotes > mccainVotes)
Console.WriteLine("Overall: Obama");
else
Console.WriteLine("Overall: McCain");
}
Copyright 2008 by Pearson Education
23
Token-Based File Parsing
 Consider a file numbers.txt that contains this text:
 (Note that ˽ means a blank space, \t means a tab)
308.2˽˽˽˽7.4
˽˽abc˽3.9˽˽˽
˽-15.4˽˽˽
 We can think of the file's contents like so:
308.2˽˽˽˽7.4\n˽˽abc˽3.9˽˽˽\n\t˽-15.4˽˽˽\n\n
 Get ALL the lines (using the TextReader.ReadToEnd()),
use the String.Split method to break the string into an
array of individual tokens,
then parse (convert) the tokens one by one
Copyright 2008 by Pearson Education
24
File input question 5
 Test the program to make sure that it handles files that
contain non-numeric tokens (by skipping them).
 For example, it should produce the same output as before
when given this input file, numbers4.txt:
308.2 hello oops 7.4
bad 3.9 stuff -15.4
4.7
5.4
:-) 2.8 @#*($&
Copyright 2008 by Pearson Education
25
Token-Based Parsing Answer

public void Slide_26()
{
char[] delimiters = { ' ', '\t', '\n', '\r' };
using (TextReader file =
new StreamReader("Files/numbers4.txt"))
{
double sum = 0.0;
string sFile = file.ReadToEnd();
string[] tokensFromFile = sFile.Split(delimiters,
StringSplitOptions.RemoveEmptyEntries);
foreach (string token in tokensFromFile)
{
double dNum;
if (Double.TryParse(token, out dNum))
{
sum += dNum;
Console.WriteLine("number = {0}", dNum);
}
}
Console.WriteLine("Sum = {0}", sum);
}
}
Copyright 2008 by Pearson Education
26
Visual Studio Details
 You will need to set the current working directory for
your programs
 In Solution Explorer: Right-click on project, then Properties
 Left-hand column: click on 'Debug'
 Working Directory is about ½ way down – use the ...
button to browse for 03_PCE_StudentCode

Notice that it puts an absolute path into the directory, and not
a relative directory 
 All the files you need for this week's PCEs are stored in
the 'Files' sub folder of the starter project
Copyright 2008 by Pearson Education
27