Lab 3 : A Command Line Parser

ECE244 Programming Fundamentals 
Lab 3 : A Command Line Parser
1 Objectives
The objectives of this assignment are for you to practice: (1) the use of C++ I/O streams, including error handling, (2) writing a simple class with constructors, accessors and mutators, and (3)
dynamic allocation and de-allocation of one-dimensional arrays. You will do so through the design
of a program that parses drawing commands from the standard input, displaying appropriate error
messages if necessary, and by creating and maintaining objects that represent drawn shapes.
2 Problem Statement
The assignment consists of two main parts. In the first part, you will write a command parser
that provides a textual input interface to your program. The parser takes a sequence of commands
as input. The commands create, delete modify and display shapes to be drawn on the screen.
Each command consists of an operation keyword followed by arguments. The command and the
arguments are separated by one or more spaces. Thus, the code you will write for this part of
the assignment should take input from the standard input, parse it, verify that it is correct, print
a response or error message and either create, delete or modify Shape objects that represent the
shapes specified in the command. The command parser loops, processing input as long as input is
available.
In the second part of the assignment, you will implement a simple “database” of objects that
stores the created shapes. To do so, you will implement a class called Shape that is used to store the
shape properties. Further, you will create and maintain a dynamically allocated array of pointers
to Shape objects to keep track of Shape objects created and deleted.
In this assignment, you will not actually draw the shapes to the screen, just process the commands and maintain the database of the Shape objects.
3 String Streams
In the lectures, you were introduced to the standard input stream (handled by cin) as well as
user-created file streams (handled by ifstream objects that you create). There is one more useful
type of user-defined streams, namely string streams. These streams come handy when processing
input one line at a time, as you will do in this assignment. The rest of this section introduces you
to string streams and how to use them. You will find that they are not that much different than
using cin or ifstream.
String streams allows the extraction of input from a string, as opposed to from the keyboard
(cin) or a file (ifstream)
1
. The following example demonstrates how this may be done.
1String streams can be also used as output streams, but in this section, only their use for input is described.
Page 1 of 12
1 #include <iostream>
2 using namespace std;
3 #include <string>
4 #include <sstream>
5
6 int main () {
7 int anInteger;
8 string inputLine = "204 113 ten";
9
10 stringstream sin (inputLine);
11
12 sin >> anInteger; // Extracts 204
13 if (sin.fail()) return (-1);
14 sin >> anInteger; // Extracts 113
15 if (sin.fail()) return (-1);
16 sin >> anInteger; // Extraction fails
17 if (sin.fail()) return (-1);
18
19 return (0);
20 }
The #include <sstream> on line 4 imports the definition of string streams, allowing it to be
used in the example. The main function creates a string variable called inputLine on line 8 and
initializes it to "204 113 ten". The declaration on line 10 creates a new string stream handler
called sin (ala cin, but you can give it any other name). This stream is initialized from the
inputLine string variables created and initialized earlier2
.
Once this is done, we can use the sin handler in the same way we use cin. We can extract an
integer from the stream, as shown on line 12. The handler sin has the same set of flags that cin
has. Thus, we can check if the extraction operation failed by invoking the method sin.fail(). In
the example, the extraction succeeds and the value 204 is placed in anInteger. The same happens
for the second extraction on line 14, which extracts 113. In contrast, the third extraction on line
16 fails, the value of anInteger is not affected and main returns with exit code -1.
The above example is not very interesting because it extracts input from a string initialized by
the program and has the foreknowledge that three integers are expected. More interesting is when
we wish to extract input from a string provided by the user and we have no knowledge of how many
extractions we have to do. The example below illustrates how to do this.
In the example, the function getline() is used to read the input the user provides through the
keyboard and places the entire stream received by cin into the string inputLine. This includes all
the white spaces in the stream (see your lecture notes for details). It also appends an eof to the
stream. Thus, while a cin stream may end with an Enter or an end-of-file (eof) a string stream
always ends with an eof. In the example, we assume the user entered 204 113 10 as input. Thus,
inputLine contains "204 113 10".
The string inputLine is then used to build the string stream handler called sin on line 11.
Subsequently, the while loop iterates until there is no more integers in a line. sin is used to extract
2A copy of the string variable is made inside the string stream. Thus, if inputLine changes after the string stream
is created, the stream in sin does not change.
Page 2 of 12
an integer from the input (the string), as shown on line 16. The first extraction succeeds and thus
the fail and eof flags of sin are false. The integer (204 in our example) is printed to the standard
output and the moreInput flag remains true. The next two iterations of the while loop extract
the next two integers, 113 and 10, and prints them to the output.
On the next iteration of the while loop, the extraction fails because the eof is encountered.
Both the fail and the eof flags are set to true. The code checks for these flags as shown on lines
17 and 18. Since the eof flag is true, the moreInputs flag is set to false (line 19) causing the while
loop to exit and the program to terminate. Since the program immediately exits after checking the
eof flag, there is really no need to clear the flags of sin, as the comment indicates in the code on
line 20. Thus, it is safe to remove this line from the code.
1 #include <iostream>
2 using namespace std;
3 #include <string>
4 #include <sstream>
5
6 int main () {
7 int anInteger;
8 string inputLine;
9
10 getline(cin, inputLine);
11 stringstream sin(inputLine);
12
13 bool moreInput=true;
14 while (moreInput) {
15
16 sin >> anInteger;
17 if (sin.fail()) {
18 if (sin.eof()) {
19 moreInput = false;
20 sin.clear(); // Not necessary
21 }
22 else {
23 cout << "Bad input\n";
24 sin.clear(); // Not necessary
25 sin.ignore(10000,’\n’); // Again not necessary
26 }
27 }
28 else cout << "The integer read is: " << anInteger << endl;
29 }
30
31 return (0);
32 }
Now, let’s assume that the user provides " 204 113 ten" as input. The first two extraction
succeed as above. However, the third extraction fails because of the ten. The fail flag is set to true
but the eof flag is set to false. The code then checks if the reason of failure is the eof (line 18), and
this is not the case. A message is printed to instruct the user that the input is bad. When using
Page 3 of 12
cin the flags must be cleared (using cin.clear()) and the input stream must be flushed using
cin.ignore(). However, with string streams, this is unnecessary since we can simply discard the
input by reading another line from cin using getline() and then rebuilding the stream handler
using the new input (not shown in the example). This automatically resets the flags and flushes
the old input, replacing it by the new one. Thus, the two calls on lines 24 and 25 are not really
necessary and can also be removed from the code.
The use of string streams is helpful when input must be processed one line at a time and the user
is not allowed to break input across multiple lines of input, separated by Enters. String streams
allows your program to get the entire line of input, analyze it and decide if the line is valid or not.
While this can be done using cin, it is more difficult since cin allows user input to be split into
multiple lines. Indeed, this is the case for this assignment and the skeleton of the main program
(included with the lab release) shows a modified version of the above example.
4 Specifications
It is important that you follow the specifications below carefully. Where the specification says
shall or must (or their negatives), following the instruction is required to receive credit for the
assignment. If instead it says may or can, these are optional suggestions. The use of should indicates
a recommendation; compliance is not specifically required. However, some of the recommendations
may hint at a known-good way to do something or pertain to good programming style. Your code
will be marked subjectively for style, so it’s best to take the recommendations unless you have a
good reason not to.
Example input and output for the program are provided in Section 6 for your convenience.
They do not cover all parts of the specification. You are responsible for making sure your program
meets the specification by reading and applying the description below.
4.1 Coding Requirements
1. The code you will write shall be contained in only two source files named parser.cpp and
Shape.cpp. Skeletons of the two files are released with the assignment’s zip file. The zip
file also contains two .h files: globals.h and Shape.h. These files are NOT to be modified
in any way. Modifying these files often results in a mark of 0 for the assignment.
However, you may make use of helper functions to split up the code for readability and
to make it easier to re-use. These functions (and their prototypes) must be in one of the
aforementioned two .cpp files. That is, you must not add any new .h or .cpp files.
2. Input and output must be done only using the C++ standard library streams cin and cout.
3. The stream input operator >> and associated functions such as fail() and eof() shall be
used for all input. C-style IO such as printf and scanf shall not be used.
4. Strings shall be stored using the C++ library type string, and operations shall be done
using its class members, not C-style strings.
5. C-library string-to-integer conversions (including but not limited to atoi, strtol, etc) shall
not be used.
Page 4 of 12
Argument Description, type, and range
name a string consisting of any non-whitespace characters3
; except strings that
represent commands, shape types or the reserved word all.
type a string that represents the type of a shape and must be one of: ellipse,
circle, rectangle or triangle
loc a positive integer (0 or larger) that represents the location of the shape in
either the x or y dimension
size a positive integer (0 or larger) that represents the size of the a shape in
either the x or y dimension
value a positive integer (0 or larger) that represents the maximum number of
shapes in the database
angle a positive integer between 0 and 360 that represents angle of rotation of a
shape
Table 1: Acceptable input arguments
4.2 Command Line Input
Input will be given one command on one line at a time. The entire command must appear on one
line. All input must be read using the C++ standard input cin. The program shall indicate that
it is ready to receive user input by prompting with a greater-than sign followed by a single space
(> ); see Section 6 for an example. Input shall always be accepted one line at a time, with each
line terminated by a newline character4
. If there is an error encountered when parsing a line, the
program shall print an error message (see Section 4.3), the line shall be discarded, and processing
shall resume at the next line. The program shall continue to accept and process input until an
end-of-file (eof) condition is received5
.
Each line of valid input shall start with a command name, followed by zero or more arguments,
each separated by one or more space characters. The number and type of arguments accepted
depend on the command. The arguments and their permissible types/ranges are shown below in
Table 1.
Command Arguments Output if Command is Valid
maxShapes value New database: max shapes is value
create name type loc loc size size Created name: type loc loc size size
move name loc loc Moved name to loc loc
rotate name angle Rotated name by angle degrees
draw name Drew name: type loc loc size size
draw all Drew all shapes
delete name Deleted shape name
delete all Deleted: all shapes
Table 2: Valid commands and arguments and their output
The valid commands, their arguments, and their output if the command and its arguments are
3 Whitespace characters are tab, space, newline, and related characters which insert “white space”; they mark the
boundaries between values read in by operator<<. All other characters (digits, letters, underscore, symbols, etc.)
are non-whitespace characters.
4A newline character is input by pressing Enter.
5
eof is automatically provided when input is redirected from a file. It can also be entered at the keyboard by
pressing Ctrl-D.
Page 5 of 12
Error message Cause
invalid command The first word entered does not match one of the valid commands
invalid argument The argument is not of the correct type. For example, a floating point
number or a string may have been entered instead of an integer where
an integer is expected.
invalid shape name The name used for a shape is a reserved word (e.g., a command name
or a shape type)
shape name exists A shape with the name name exists in the database, i.e., has once
been created and has not been deleted
shape name not found A shape with the name name specified in a command does not exist
invalid shape type The type used for a shape is not one of the allowed types
invalid value The value specified in a command is invalid. For example, a less than
0 value for a loc argument, a rotation angle not between 0 and 360,
or the two size arguments of the circle shape are not equal.
too many arguments More arguments were given than expected for a command
too few arguments Fewer arguments were given than expected for a command
shape array is full An attempt to create more shapes than the argument given to the
maxShapes command
Table 3: List of errors to be reported, in priority order
all legal are shown below in Table 2. Notice that the last two commands (draw and delete) can be
run in two ways (depending on their argument): with a specific shape name, or with the keyword
all. In the case of the draw all command, the program prints not only the message shown in the
table, but also all the shapes in the database (see the example in Section 6).
Also notice that for the circle shape, the two size arguments must be equal, or an error
message is printed, as described in Section 4.3.
The program shall verify that the command and arguments are correctly formatted and within
range, and that a command is followed by the correct number of arguments. The handling of
command names shall be case-sensitive.
The first line of input to your program will always be the maxShapes command. It creates
a new (empty) database of shapes with the specified maximum allowed number of shapes. You
shall assume that this command will not have any errors in it. The maxShapes command may be
given multiple times. A subsequent maxShapes command shall create a new database with the new
maximum allowed number of shapes (see Section 4.5 for more details).
If there is an error, a message shall be displayed as described in Section 4.3. Otherwise, a
successful command produces a single line of output on the C++ standard output, cout, as shown
in Table 2. The values in italics in Table 2 must be replaced with the values given by the command
argument. Strings must be reproduced exactly as entered. Where locs or sizes are printed, they
shall appear on the order entered in the command.
4.3 Error Checking
The program must check that the input is valid. It must be able to identify and notify the user
of the following input errors, in order of priority. Where multiple errors exist on one input line,
only one should be reported: the one that occurs first as the line is read from left to right. If more
than one error could be reported for a single argument in the line, only the error occuring first in
Table 3 should be reported.
Page 6 of 12
Errors shall cause a message to be printed to cout, consisting of the text “Error:” followed by
a single space and the error message from Table 3. In the messages, italicized values such as name
should be replaced by the value causing the error. Error message output must comply exactly
(content, case, and spacing) with the table below to receive credit. There are no trailing spaces
following the text.
The program is not required to deal with errors other than those listed in Table 3. The following
are some clarification on the errors.
1. The commands and the shape types are case sensitive. Thus, while a shape cannot be named
all, draw or triangle, it can be named All, Draw or triAngle.
2. For every line of input, your program must output something. Either a message indicating
success (Table 2) or an error (Table 3). So for an empty line of input (nothing or just
whitespace) your program should print Error: invalid command.
3. Only the first error from the left should be reported per line of input. In the case of missing/extra arguments, these are errors in the arguments that are missing/extra and should be
reported only if the preceding arguments were valid.
4. You should let the extraction operator (>>) do the work for you. Recall that the operator
stops when the next character cannot be converted to the destination type. This may or may
not be a white space. Learn how to use the cin.peek() method explained in Section 5 below.
4.4 The shape Class
The shape class holds the properties of a shape, including its name, type, location, size and
rotation. The definition of the class appears in Shape.h, which is re-produced in Figure 1. Examine
the file and read through the comments to understand the variables and methods of the class. You
must implement this class in the file Shape.cpp.
4.5 The Database
The program shall keep track of all shapes using an array whose elements are pointers to Shape
objects. The array should be dynamically allocated after the first line of the input to have a size
that matches exactly the argument given to the maxShapes command. The array elements shall
all be initialized to NULL. This array is declared in parser.cpp. An integer variable shapeCount
is used to track the actual number of Shape objects stored in the database. Figure 2 depicts what
the database may look like during program execution.
All shapes shall be stored in the array (i.e., by having the pointer element of the array point to
a Shape object) starting at element 0 for the first shape added and incrementing from there. When
a Shape object is deleted, the memory allocated to the object must be freed and the element of the
array that used to point to the Shape object must be assigned the value NULL When a new Shape
object is added after another one is deleted, it must be added at location shapeCount. Thus, you
must not “pack” the array after deletions or reuse “deleted” locations.
When the maxShapes command is issued after a database has been created with an earlier
maxShapes command, all the shapes created so far must be deleted. Further, the existing dynamically allocated array must be de-allocated. Only then can a new array, with the new shapeCount
value be created. It is critical that your program deletes Shape objects and the database array so
as no memory leaks occur. Indeed, it should delete all the memory it allocates with new before
it exits. In this assignment, memory leaks will be checked for and reported by exercise and the
Page 7 of 12
1 //
2 // Shape.h
3 // lab3
4 //
5 // Modified by Tarek Abdelrahman on 2020-10-04.
6 // Created by Tarek Abdelrahman on 2018-08-25.
7 // Copyright 2018-2020 Tarek Abdelrahman.
8 //
9 // Permission is hereby granted to use this code in ECE244 at
10 // the University of Toronto. It is prohibited to distribute
11 // this code, either publicly or to third parties.
12
13 // *********** ECE244 Student: DO NOT MODIFY THIS FILE ***********
14
15 #ifndef Shape_h
16 #define Shape_h
17
18 #include <iostream>
19 #include <string>
20 using namespace std;
21
22 class Shape {
23 private:
24 string name; // The name of the shape
25 string type; // The type of the shape (see globals.h)
26 int x_location; // The location of the shape on the x-axis
27 int y_location; // The location of the shape on the y-axis
28 int x_size; // The size of the shape in the x-dimension
29 int y_size; // The size of the shape in the y-dimension
30 int rotation = 0; // The rotations of the shape (integer degrees)
31 public:
32 // Build a Shape object with its properties
33 Shape(string n, string t, int x_loc, int x_sz, int y_loc, int y_sz);
34
35 // Accessors
36 string getType(); // Returns the type
37 string getName(); // Returns the name of the shape
38 int getXlocation(); // Returns location of the shape on the x-axis
39 int getYlocation(); // Returns location of the shape on the y-axis
40 int getXsize(); // Returns the size of the shape in the x-dimension
41 int getYsize(); // Returns the size of the shape in the y-dimension
42
43 // Mutators
44 void setType(string t); // Sets the type (see globals.h)
45 // No error checking done inside the method
46 // The calling program must ensure the type
47 // is correct
48 void setName(string n); // Sets the name of the shape
49 void setXlocation(int x_loc); // Sets location of the shape on the x-axis
50 void setYlocation(int y_loc); // Sets location of the shape on the y-axis
51 void setXsize(int x_sz); // Sets size of the shape in the x-dimension
52 void setYsize(int y_sz); // Sets size of the shape in the y-dimension
53
54 void setRotate(int angle); // sets the rotation of the shape
55
56 // Utility methods
57 void draw(); // Draws a shape; for this assignment it
58 // only prints the information of the shape
59 };
60
61 #endif /* Shape_h */
62
Figure 1: Defintion of the class shape
Page 8 of 12
autotester, but there is no penalty for having memory leaks. In future assignments there will be
penalties for leaking memory.
A good way to check if you have deleted all the memory you allocated with new is to run the
valgrind memory checking program. A tutorial on valgrind is released with this assignment.
You are encouraged to learn and use this tool. It is used by exercise to check for memory leaks
in your code.
0 1 max_shapes-1
shape shape shape
shapesArray
Figure 2: A depction of the database
5 Hints
• You can check a stream for end-of-file status using the eof member function.
• The ignore member function in iostream may be useful to you if you need to ignore the
remainder of a line.
• To save typing, you can create one or more test files and pipe them to your program. You
can create a text file using a text editor (try gedit, gvim, or the NetBeans editor). If your
file is called test.txt, you can then send it to your program by typing main < test.txt.
Building a good suite of test cases is important when developing software.
• If you want to look ahead (“peek”) at what character would be read next without actually
reading it, peek() does that. For instance, if you type “Hello” then each time you run peek()
you will get ’H’. If you read a single character, it will return ’H’ but then subsequent calls to
peek() will return ’e’.
• When interacting with your program from the keyboard, Ctrl-D will send an end-of-file (eof)
marker.
• Reading from cin removes leading whitespace. When reading strings, it discards all whitespace characters up to the first non-whitespace character, then returns all non-whitespace
characters until it finds another whitespace. For integers (numbers), it skips whitespace and
reads to the first non-digit (0-9) character.
• Remember you can use the debugger to pause the program, step through it, and view variables
(including strings).
• If you decide to pass the string stream you created to a function, remember that string streams
(and other types of streams for that matter) can only be passed by reference, not by value.
A suggested (but not mandatory) structure for your code appears in the skeleton parser.cpp
file released within the assignment’s zip file.
Page 9 of 12
6 Examples
6.1 A Short Example
The program when first started, ready to receive input:
>
Now the user types a command (ending with Enter) to create a new database of size 100.
> maxShapes 100
To which the program should respond with the message indicating the successful creation of new
database with 100 entries.
New database: max shapes is 100
The user then creates a new ellipse called my circle located at x and y positions of 30 and 40 and
with a x and y sizes of 10 and 10.
> create my_circle ellipse 30 40 10 10
To which the program should respond with the message for a successful creation of a shape:
Created my_circle: ellipse 30 40 10 10
6.2 Full session
The following is an example session. Note that the text from the prompt (> ) up to the end of the
line is typed by the user, whereas the prompt and line without a prompt are program output.
> maxShapes 4
New database: max shapes is 4
> create my_circle ellipse 50 65 20 20
Created my_circle: ellipse 50 65 20 20
> create my_square rectangle 100 150 60 60
Created my_square: rectangle 100 150 60 60
> create a_circle circle 120 200 30 40
Error: invalid value
> create my_triangle triangle 40 75 -90 90
Error: invalid value
> create my_rectangle rectangle 100 275 90 180
Created my_rectangle: rectangle 100 275 90 180
> create ellipse rectangle 100 275 90 180
Error: invalid shape name
> create my_rectangle triangle 70 50 10 5
Error: shape my_rectangle exists
> create second_triangle triangle 70 50 10 5
Created second_triangle: triangle 70 50 10 5
> move my_circle
Page 10 of 12
Error: too few arguments
> mve my_circle
Error: invalid command
> move my_circle 70 90
Moved my_circle to 70 90
> rotate my_rectangle 90 100
Error: too many arguments
> rotate my_rectangle 100
Rotated my_rectangle by 100 degrees
> rotate my_rectangle 400
Error: invalid value
> draw my_trinagle
Error: shape my_trinagle not found
> draw my_circle
Drew my_circle: ellipse 70 90 20 20
> draw all
Drew all shapes
my_circle: ellipse 70 90 20 20
my_square: rectangle 100 150 60 60
my_rectangle: rectangle 100 275 90 180
second_triangle: triangle 70 50 10 5
> delete my_square
Deleted shape my_square
> draw all
Drew all shapes
my_circle: ellipse 70 90 20 20
my_rectangle: rectangle 100 275 90 180
second_triangle: triangle 70 50 10 5
> delete all
Deleted: all shapes
> draw all
Drew all shapes
>
7 Procedure
Create a sub-directory called lab3 in your ece244 directory, and set its permissions so no one else
can read it. Download the lab3 release.zip file, un-zip it and place the resulting files in the
lab3 directory. There are two .cpp files in which you will add your code. The first is parser.cpp
in which you will write the command parser code. The second file is Shape.cpp in which you will
implement the class Shape. Both files are in the directory parser. You must not rename these files
or add more files. There is also a NetBeans project to help you get started with NetBeans.
The release also contains two include files globals.h and Shape.h. You may NOT modify
these files to add to or delete from their content. Modifying the files commonly results in a mark
of zero for the assignment. In addition, there is a Makefile that is used by NetBeans to separately
compile your project. Do not modify this file either.
Write and test the program to conform to the specifications laid out in Section 4. The hints in
Page 11 of 12
Section 5 may help get you started, and the example sessions in Section 6 may be used for testing.
The ~ece244i/public/exercise command will also be helpful in testing your program. You
should exercise the executable, i.e., parser.exe, using the command:
~ece244i/public/exercie 3 parser.exe
As with previous assignments, some of the exercise test cases will be used by the autotester
during marking of your assignment. We will not provide all the autotester test cases in exercise,
however, so you should create additional test cases yourself and ensure you fully meet the specification listed above.
8 Deliverables
Submit the parser.cpp and Shape.cpp files as lab 3 using the command
~ece244i/public/submit 3
Page 12 of 12