Fixed footnotes

Author: Efthimia Aivaloglou

floats/table-tokens.tex

@@ -40,5 +40,3 @@
 \bottomrule
 \multicolumn{4}{l}{\begin{footnotesize}\textsuperscript{\number\value{footnote}} A function list is available as part of the reference implementation. Lists provided by Microsoft are also available in \cite{ExcelFunctionReference} and \cite{ExcelOfficialGrammar}.\end{footnotesize}}\\
 \end{tabular}
-
-\stepcounter{footnote}

xlGrammar.tex

@@ -350,7 +350,7 @@ \section{Evaluation}
 
 In this section we explain how we implemented and evaluated the grammar using two large datasets and we discuss the obtained results and formula parse failures. In the grammar analysis in Section \ref{subsection:grammarAnalysis} we examine how frequently language features occur in the formulas of the datasets.
 
-The grammar is implemented using the Irony parser generator framework\footnote{\url{https://irony.codeplex.com/}} and the resulting parser, named XLParser, is available for download\footnote{\url{https://github.com/PerfectXL/XLParser}}. An online demo is also available\footnote{\url{http://xlparser.perfectxl.nl/demo}}.
+The grammar is implemented using the Irony parser generator framework\footnote{\url{https://irony.codeplex.com/}} and the resulting parser, named XLParser, is available for download\footnote{\url{https://github.com/PerfectXL/XLParser}}. An online demo is also available.\footnote{\url{http://xlparser.perfectxl.nl/demo}}
 
 To extract unique formulas from spreadsheets and use them as input to the parser we built a tool that opens spreadsheets using a third-party library called Gembox\footnote{\url{http://www.gemboxsoftware.com/}}. This tool reads all cells and identifies the formulas that are unique when adjusted for cell location (R1C1 representation), thus rejecting the formulas with adjusted references due to cell copying (e.g. formulas \texttt{=C1} and \texttt{=C2} are considered the same if contained in cells A1 and A2 respectively). The tool then uses each unique formula string as input to the parser.
 
@@ -513,7 +513,7 @@ \section{Conclusion}
 \label{section:conclusion}
 In this paper we (1) present a grammar for spreadsheet formulas, (2) evaluate it against over one million unique formulas, successfully parsing 99.99\%, and (3) use it to analyze the formulas in the dataset, measure the frequency of their characteristics and find uncommon grammatical cases.
 
-The grammar is compact and produces processable parse trees, suited for further manipulation and analysis. We believe that the grammar is reliable and concise enough to facilitate further research on spreadsheet formula codebases. It has already been applied in other works for analyzing formula characteristics, calculation chains and code smells and for applying formula transformations. The XLParser is published as open-source software\footnote{\url{https://github.com/PerfectXL/XLParser}} and an online demo is also available\footnote{\url{http://xlparser.perfectxl.nl/demo}}.
+The grammar is compact and produces processable parse trees, suited for further manipulation and analysis. We believe that the grammar is reliable and concise enough to facilitate further research on spreadsheet formula codebases. It has already been applied in other works for analyzing formula characteristics, calculation chains and code smells and for applying formula transformations. The XLParser\footnote{\url{https://github.com/PerfectXL/XLParser}} is published as open-source software and an online demo is also available.\footnote{\url{http://xlparser.perfectxl.nl/demo}}
 
 A point of improvement for the grammar is that its exact compatibility with the official Excel grammar is unknown.
 A comparison to the official specification could lead to either improving compatibility, or extending the number of known limitations.