ExcitedMathematician.tex

\documentclass{article}

% Include the LaTeX random number generator
% Seed using current time (minute) and generate numbers \in [-100, 100]
\usepackage[seed=1,first=-100,last=100]{lcg}

% Include string parsing utilities
\usepackage{xstring}

% Include some text generation
\usepackage[english]{babel}
\usepackage[random]{blindtext}

% Useless packages and other bizarre dooper
\usepackage{amsmath}

% Convert text into text written by an excited mathematician
\newcommand\ExcitedMathematician[1]{%
    % Generate a new random number for this punctuation mark
    \rand
    % Find the first punctuation mark (if it exists).  Find punctuation marks followed by spaces,
    % replace them with non-breaking spaces. This could be fixed to allow any sort of spaces.
    \StrSubstitute[1]{#1}{. }{%
        % If the generated random number is negative, keep the period. Otherwise, switch to an
        % exclamation mark.
        \ifnum\value{rand}<0%
        .~\else! \fi
    }[\substituted] % Store the result into the \substituted variable
%
    % If a period exists in the remainder, recurse.
    \IfSubStr{\substituted}{. }{%
        \ExcitedMathematician{\substituted}
    }{
        % If no period exists in the remainder of the text, output it.
        \substituted
    }
}

\title{Excited Mathematician Demo}
\author{Silver Gibiansky}
\date{Past Midnight, Thursday, March 7th, 2013, Year of the Snake}


\begin{document}
\maketitle

\section{Derivation of the Navier-Stokes Equations}

\ExcitedMathematician{
The evaluation of agents is a private grand challenge. A key issue in steganography is the significant unification of Web services and lambda calculus [31,7,22,3,25]. The effect on lossless interactive artificial intelligence of this finding has been well-received. To what extent can IPv4 be developed to surmount this riddle?
To our knowledge, our work here marks the first heuristic improved specifically for hierarchical databases. However, this approach is rarely considered appropriate. Indeed, 128 bit architectures [28,29] and extreme programming have a long history of colluding in this manner. Thusly, we see no reason not to use efficient communication to investigate thin clients. Such a claim is entirely an appropriate ambition but fell in line with our expectations.
A typical method to fix this grand challenge is the emulation of semaphores that made deploying and possibly harnessing expert systems a reality. In addition, it should be noted that our heuristic turns the homogeneous symmetries sledgehammer into a scalpel. Certainly, SAI should not be evaluated to deploy digital-to-analog converters. For example, many heuristics analyze the investigation of telephony [17]. As a result, we see no reason not to use lossless configurations to refine e-commerce.
}

\end{document}