- German Robayo (14-10924)
- Gustavo Castellanos (14-10192)
A lexer is a tool that allows us to analyze a sequence of characters and extract tokens from it (lexicographic analysis) for its subsequent parsing.
It was used as a programming language Haskell and as a generator tool for lexicographical analyzers Alex. The use of Alex was fundamental, as we focused on only determining regular expressions to extract the tokens from the file.
Alex provides several wrappers, but the one of our interest was posn because it provides the same functionality as the basic but adding the column and row of each detected token.
To implement the show of each token, a type of data TkObject was created that can be seen as an ordered pair whose first coordinate is a token and the second is its AlexPosn. Thus, it was avoided having to write a show for each token.
To filter errors, a token called TkErr was created. When we invoke alexScanTokens we have an array of TkObject's. To see if there is an error, we only filter the TkObjects and if the resulting array is empty, there are no errors and each token is printed in the described format. Otherwise, only errors are printed.
None. Although certain functions could be imported, it was decided, to practice the language more, that we ourselves implement those functions.
A perser is a computer program that analyzes a string of symbols (syntactically) according to the rules of a formal grammar.
It was used as programming language Haskell and as a parser generating tool Happy. The use of Happy focused us only on determining the grammar of the language.
Through Happy, the grammar that defines the BasicTrans language was formed. This had to be modified during the development of the parser to eliminate conflicts shift/reduce and reduce/reduce.
A Parsed data type was defined, which instantiates the Monad typeclass, in order to know if an error occurred or not while parsing.
The ToStr typeclass was defined, so that the Abstract Syntactic Tree could be printed on the console, which is instantiated by the types of data that define the analyzed tokens.
To compile the interpreter:
stack buildTo run the interpreter:
stack exec BasicTran-Interpreter-exe <archive>