Blogs

Extending the NFA generator Previously we implemented a regular expression matching algorithm that generated an NFA from an input regular expression, and simulated a string over the NFA to check for acceptance. The regular expression was required to be in postfix form and the algorithm only supported union $|$ , concatenation $.$ , Kleene star $*$ , one-or-more $+$ , and zero-or-one $?$ operations. While this was sufficient for simple applications, building a lexer would require support for character ranges (for example $[0\text{-}9]$, $[a\text{-}z]$) and escape characters (for example one may wish to use $*$ for the multiply operator rather than Kleene star)....

Introduction The first step in building a compiler would be to build a lexical analyzer, shortened to ’lexer’. To define what a lexer does we may understand it as a function: $$\text{Lexer :: string} \to [\text{token}]$$ Where a token can be thought of as the tuple $\text{token} = \langle \text{class, string} \rangle$. A token class, loosely speaking, corresponds to a set of strings. For example we may class strings as identifier tokens, integer tokens, keyword tokens or whitespace tokens....

Notes from learning measure theory. Covers measurable spaces and functions. These are notes and therefore liable to inaccuracies.

First in a series about building a DBMS on the Raspberry Pi Pico. This article covers the first step in this process: programming flash memory

Notes from learning measure theory. Establishes the need for sigma algebras. These are notes and therefore liable to inaccuracies.