Description
Introduction
This assignment asks you to complete the lexical analysis and parsing phases of the front-end for the
CFlat (Cb) compiler.
In addition, you should prepare one more test program in the Cb language which will run and do
something interesting. Your test program should exercise some Cb language features which are not
used in the sample program CbExample.cs which is provided with this assignment.
You are provided with an incomplete parser specification in the CbParser.y file. (In particular, the
grammar rules in that file do not allow multiple classes in the program or multiple methods in a class.)
Assignment Description
1. Your Cb test program should compile and run with the regular C# compiler.
2. You are provided with an incomplete grammar for the Cb language in a form which is accepted
by the gppg parser generator. This is the file named CbParser.y. If the command
gppg /gplex /conflicts CbParser.y > CbParser.cs
is executed, a parser module file named CbParser.cs and coded in C# is generated.
3. Your task is to provide a complete lexical analyzer (aka scanner or lexer) which works with the
generated parser. The gplex tool should be used to generate that lexer. The C# file which contains the lexer should be named CbLexer.cs.
4. You also need to provide a main program which invokes the lexer and parser on a Cb source file.
The main program should be implemented as a file named cbc.cs which contains a class with a
name of your choosing. That class must contain a static method named Main. The Main method
accepts command line options and exactly one filename argument. An example invocation of the
lexer+parser showing every option is as follows:
cbc -tokens -debug sampleprog.cs
where cbc is the name of the program being invoked (taken from the filename cbc.cs) and
sampleprog.cs is a textfile holding a sample Cb program.
The -tokens option causes the lexer to generate a listing of all the tokens which are encountered in the Cb source file. The listing should be written to a new file named tokens.txt, with
one token per line. A few lines of the listing might be
Token.Kwd_class
Token.Ident, text = “Foo”
Token “{”
where the representation of each token (e.g. the name Kwd_class) is taken from the names used
for these token types in CbParser.y (see the Tokens enum type in the CbParser.cs file).
The -debug option sets a flag which enables your own debugging output. You will find this
option to be very useful when developing the later stages of the compiler. What you output is
completely up to you. (Indeed you can output nothing at all.)
When a Cb source file is successfully parsed, there should be a succinct output message similar to
the following:
Page 2
237 lines from file cbtest.cs were parsed successfully
If a lexical error or a syntactic error is discovered, an error message which specifies the line number and the name of the source file should be generated. The default message produced by a
gplex scanner or gppg parser is quite acceptable as long as the message includes the location of
the error. (The column number within a line need not be provided.)
Advice
Your tasks can be performed in this order:
1. Add %token declarations to the CbParser.y file for all keywords mentioned in the Cb Language Specification document but which are missing.
2. Implement a scanner which recognizes all the lexical elements listed in the Cb Language Specification.
3. Compare the EBNF grammar against the grammar implemented by the CbParser.y file and
add any missing Cb language constructs to the CbParser.y file. Generate the parser code.
4. Write the cbc.cs file so that it calls the parser on the specified test file. Compile the three files
together.
5. Construct a small trivial Cb test file and check to see if your scanner+parser combination successfully recognizes it. If there are problems recognizing the input, fix the problems and try again.
Otherwise, expand your test file to use more features of the Cb language and repeat.
The Provided Materials
• The conneX website holds the most recent versions of the gplex and gppg programs, along
with full documentation for them.
• The file CbLangSpec.pdf holds a succinct specification of the Cb language. Note that Cb is
intended to be a proper subset of C#. If the meaning of any construct is not explained or unclear,
just assume the usual C# meaning.
• The file CbExample.cs holds a sample Cb program.
Submission Requirements
1. You must provide four files. The three files which implement the lexer+parser should be named
CbLexer.lex, CbParser.y and cbc.cs. The sample Cb program can be named anything as
long as the filename makes it obvious that this is a test program or sample program, and has the
suffix “.cs”. A name beginning with the letters “test” would be appropriate.
2. Important: do not submit any files generated by gplex or gppg.
3. You must combine all your files into a single compressed archive file. The only accepted formats
for the archive file are as a zipfile (and the filename must have a “.zip” suffix) or as a gzipped
Unix tar file (and the filename must have a “.tgz” suffix).
4. The project is to be completed in teams of either 2 or 3 persons. The ideal size is 2 people. All
team members must participate. To encourage and to reward active participation, you can expect
the midterm test to contain at least one question on the minutiae of Cb semantics. Team memberships can be changed after each assignment,
