I have a generated database and need to look up values

Question

A database in a format that can be freely defined had entities with the following attributes:

ID, HEXNUMBER, LATEXFORMULA, LITERAL.

• The ID is unique and can contain numbers, dots and dashed. Example: MystructOne.val16_A

• HEXNUMBER is always an address like 0xD0007246.

• LATEXFORMULA will look like so: $p = \frac{57426}{10^8} \mathrm{[V]} \cdot i$

• The LITERAL will contain type information: UINT(t16)

I am free to save the data however I want, but the contents is fixed. How can I save them so that I can access them in LaTeX with something like the following:

\custom{MystructOne.val16_A}{ID}
\custom{MystructOne.val16_A}{HEXNUMBER}
\custom{MystructOne.val16_A}{LATEXFORMULA}
\custom{MystructOne.val16_A}{LITERAL}

and at one point I would like to generate a table using something like the following:

foreach key in keysource
{
    \custom{key}{ID}
    \custom{key}{HEXNUMBER}
    \custom{key}{LATEXFORMULA}
    \custom{key}{LITERAL}
}

What can I do in LaTeX to access data like that?

EDIT #2 (replaced previous edit)

This is how far i have gotten now:

\documentclass{article}
\usepackage{hyperref}

\newcounter{keylabel}
\newcommand*{\keylabel}[2]{%
    \leavevmode%
    \raisebox{2ex}[0pt][0pt]{%
    \renewcommand*{\thekeylabel}{#1}%
    \refstepcounter{keylabel}%
    \label{#2}%
    }%
    #1%
}

\usepackage[T1]{fontenc}

\usepackage{csvtools}
\setcsvseparator{;}
\def\verbID{\texttt{\edef\tmp{\insertID}\expandafter\detokenize\expandafter{\tmp}}}
\newcommand{\DATARow}[4]{\keylabel{#1}{#1} & #2 & #3 & #4\\[.5ex]}
%The keylabel throws a lot of errors here

\begin{document}

%%% - Creating the Table
\CSVtotabular{stackoverflow.csv}%
  {|c|c|c|c|}%
  {\DATARow{ID}{REGISTER}{TYPE}{FORMULA}}%
  {\DATARow{\verbID}{\insertREGISTER}{\insertTYPE}{\insertFORMULA}}%
  {\DATARow{\verbID}{\insertREGISTER}{\insertTYPE}{\insertFORMULA}}

\newpage  

%%% - Referencing something in the Table:
\noindent
%\ref{A_B_C_D}\\  - Does not work at all
%\ref{ONE}\\  - Does not work at all
%\ref{TWO.FOUR}\\  - Does not work at all
%\ref{Thirty[2]}\\  - Does not work at all
%\ref{TreeHouse}\\  - Does not work at all

\end{document}

The final step will be to be able to set Keylables and reference them. I want the referencing to work like in this answer here.

But i am getting errors and warnings.

• I can not set the keys, some envireonment collision i dont understand.
• I was not able to test this, but i think i need a \verbRef.
• I need to supress the warnings in texlipse that will occur since the keys can only be set during compile time. I dont know how to do that.

These final items are the ones in need to clear before i can reasonably say that i am looking up the values that i need.

Here is a demo csv file that can be used with this problem:

ID;REGISTER;TYPE;FORMULA
A_B_C_D;0xD000720C;UINT(16);$Y = \frac{57426}{10^8} \cdot X$
ONE;0xD020720C;UINT(8);$Y = 109 \cdot X$
TWO.FOUR;0xD080720C;INT(16);$Y = \frac{57426}{X}$
Thirty[2];0xD009720C;INT(8);$Y = X$
TreeHouse;0xD200720C;UINT(32);$Y = \frac{X}{10^8} \cdot X$

Answer 1

This seems to do what you want, at least for what I can understand.

\documentclass{article}
\usepackage[T1]{fontenc}
\usepackage{csvtools}

\setcsvseparator{;}

\def\IDs{}
\let\IDdo\relax
\applyCSVfile{johancsv.csv}
 {%
  \expandafter\xdef\csname joh@\insertID@@@ID\endcsname{\detokenize\expandafter{\insertID}}%
  \expandafter\gdef\csname joh@\insertID@@@REGISTER\expandafter\endcsname\expandafter{\insertREGISTER}%
  \expandafter\gdef\csname joh@\insertID@@@TYPE\expandafter\endcsname\expandafter{\insertTYPE}%
  \expandafter\gdef\csname joh@\insertID@@@FORMULA\expandafter\endcsname\expandafter{\insertFORMULA}%
  \xdef\IDs{\IDs\IDdo{\insertID}}
 }

\newcommand{\use}[2]{\csname joh@#1@@@#2\endcsname}
\newcommand{\eachID}[1]{\def\IDdo##1{#1}\IDs}

\begin{document}

\use{A_B_C_D}{FORMULA}

\bigskip

\use{TreeHouse}{TYPE}

\bigskip

\newcounter{formulas}
\eachID{%
  \stepcounter{formulas}\noindent
  \theformulas\quad\texttt{\use{#1}{ID}}
  \use{#1}{FORMULA}\par\medskip
}

\end{document}

For each line of the CSV file, we associate to each ID some macros that can later be accessed with

\use{<ID>}{<FIELD>}

You can even say

\use{A_B_C_D}{ID}

for printing the ID without worrying about underscores (provided the T1 encoding is chosen); an example is in the \eachID at the end.

The macro \eachID cycles through the list of ID's and performs the actions specified in the argument, where #1 stands for the current ID.

Answer 2

I'd do this in LuaTex, because Lua can do this kind of custom record extraction more simply than Tex can.

The Lua code below is quite simple (I'll test it if you provide an example data set):

t={} -- The table we store the result in
for line in io.lines "filename" do
  id, hex, lstr, lit = line:match "^([^,]+),%s*(0x%x+),%s*(.*),%s*[^,]+$"
  t[1+#t] = {id=id, hex=hex, lstr=lst, lit=lit}
end

and fills the table t with named fields for the four comma-separated field in the input file; the code checks that the hex field actually takes the form of a hex number. It forbids commas in any field but the Latex code field.

The above Lua can be embedded in a Lua Latex document using \directlua of one of the Lua wrappers, and your \custom macro can also be defined to access table records using \directlua.

Answer 3

Don't use csvtools it's obsolete. Here's an answer that uses the replacement package datatool:

\documentclass{article}

\usepackage[T1]{fontenc}
\usepackage{datatool}
\usepackage{hyperref}

\begin{filecontents}{test.csv}
ID;REGISTER;TYPE;FORMULA
A_B_C_D;0xD000720C;UINT(16);$Y = \frac{57426}{10^8} \cdot X$
ONE;0xD020720C;UINT(8);$Y = 109 \cdot X$
TWO.FOUR;0xD080720C;INT(16);$Y = \frac{57426}{X}$
Thirty[2];0xD009720C;INT(8);$Y = X$
TreeHouse;0xD200720C;UINT(32);$Y = \frac{X}{10^8} \cdot X$
\end{filecontents}

\DTLsetseparator{;}
\DTLloaddb{data}{test.csv}

% syntax: \useentry{ID value}{field}
\newcommand{\useentry}[2]{%
 % fetch row for given ID field:
 \dtlgetrowforvalue{data}{\dtlcolumnindex{data}{ID}}{#1}%
 % access required field from current row:
 \dtlgetentryfromcurrentrow{\thisval}{\dtlcolumnindex{data}{#2}}%
 \thisval
}

\newcounter{keylabel}
\newcommand*{\keylabel}[2]{%
    \leavevmode%
    \raisebox{2ex}[0pt][0pt]{%
    \renewcommand*{\thekeylabel}{#1}%
    \refstepcounter{keylabel}%
    \label{#2}%
    }%
    #1%
}

\makeatletter
\DeclareRobustCommand*{\dispID}[1]{%
  \def\thisID{#1}%
  \@onelevel@sanitize\thisID
  \texttt{\thisID}%
}
\makeatother

\newcommand{\DATARow}[4]{%
 \keylabel{\dispID{#1}}{#1} & #2 & #3 & #4\\[.5ex]%
}


\begin{document}

Reference entries in the database:

\useentry{A_B_C_D}{FORMULA}

\useentry{Thirty[2]}{REGISTER}

Display all entries in a table:

\begin{tabular}{llll}
ID & Register & Type & Formula\\
\DTLforeach*{data}{\ID=ID,\Register=REGISTER,\Type=TYPE,\Formula=FORMULA}%
{%
  \expandafter\DATARow\expandafter{\ID}{\Register}{\Type}{\Formula}%
}
\end{tabular}

Reference something in the table:

\ref{A_B_C_D}
\ref{ONE}
\ref{TWO.FOUR}
\ref{Thirty[2]}
\end{document}

This produces: