| 1 |
% \iffalse meta-comment |
| 2 |
% |
| 3 |
% Copyright (C) 2019 by Laurence R Taylor |
| 4 |
% ----------------------------------- |
| 5 |
% |
| 6 |
% This file may be distributed and/or modified under the |
| 7 |
% conditions of the LaTeX Project Public License, either version 1.3 |
| 8 |
% of this license or (at your option) any later version. |
| 9 |
% The latest version of this license is in: |
| 10 |
% |
| 11 |
% http://www.latex-project.org/lppl.txt |
| 12 |
% |
| 13 |
% and version 1.3 or later is part of all distributions of LaTeX |
| 14 |
% version 2005/12/01 or later. |
| 15 |
% |
| 16 |
% \fi |
| 17 |
% |
| 18 |
% \iffalse |
| 19 |
% |
| 20 |
%<*driver> |
| 21 |
\documentclass{ltxdoc} |
| 22 |
\usepackage{csvmerge} |
| 23 |
\usepackage{hyperref} |
| 24 |
\hypersetup{ |
| 25 |
colorlinks=true, |
| 26 |
linkcolor=blue, |
| 27 |
filecolor=magenta, |
| 28 |
urlcolor=cyan, |
| 29 |
} |
| 30 |
\DisableCrossrefs |
| 31 |
\CodelineNumbered |
| 32 |
\RecordChanges |
| 33 |
\CheckSum{525} |
| 34 |
\begin{document} |
| 35 |
\DocInput{csvmerge.dtx} |
| 36 |
\end{document} |
| 37 |
%</driver> |
| 38 |
% \fi |
| 39 |
% |
| 40 |
% |
| 41 |
% \CharacterTable |
| 42 |
% {Upper-case \A\B\C\D\E\F\G\H\I\J\K\L\M\N\O\P\Q\R\S\T\U\V\W\X\Y\Z |
| 43 |
% Lower-case \a\b\c\d\e\f\g\h\i\j\k\l\m\n\o\p\q\r\s\t\u\v\w\x\y\z |
| 44 |
% Digits \0\1\2\3\4\5\6\7\8\9 |
| 45 |
% Exclamation \! Double quote \" Hash (number) \# |
| 46 |
% Dollar \$ Percent \% Ampersand \& |
| 47 |
% Acute accent \' Left paren \( Right paren \) |
| 48 |
% Asterisk \* Plus \+ Comma \, |
| 49 |
% Minus \- Point \. Solidus \/ |
| 50 |
% Colon \: Semicolon \; Less than \< |
| 51 |
% Equals \= Greater than \> Question mark \? |
| 52 |
% Commercial at \@ Left bracket \[ Backslash \\ |
| 53 |
% Right bracket \] Circumflex \^ Underscore \_ |
| 54 |
% Grave accent \` Left brace \{ Vertical bar \| |
| 55 |
% Right brace \} Tilde \~} |
| 56 |
% |
| 57 |
% |
| 58 |
% |
| 59 |
% \changes{v1.0}{?2019?/?07?/?17?}{Initial version} |
| 60 |
% \title{csvmerge} |
| 61 |
% \author{Laurence R Taylor} |
| 62 |
% \date{August 8, 2019} |
| 63 |
% \maketitle |
| 64 |
% |
| 65 |
% \begin{abstract} |
| 66 |
% The package |csvmerge.sty| is designed to take a bit of \TeX\ code such as a letter or a line |
| 67 |
% in a table and produce |
| 68 |
% multiple copies of it with parts of each copy filled in from a data file. |
| 69 |
% It is invoked in the usual manner from a master-file, that is a standard |
| 70 |
% \LaTeX\ file with |
| 71 |
% |
| 72 |
% \null\noindent\hskip 1in|\documentclass{whatever}|\\ |
| 73 |
% \null\hskip 1in|\usepackage{csvmerge}|\\ |
| 74 |
% \null\hskip 1.5in$\vdots$ |
| 75 |
% |
| 76 |
% The data source is a standard csv file. |
| 77 |
% The bit of \TeX\ code resides in a blueprint file. |
| 78 |
% The |\usepackage{csvmerge}| can go anywhere in the preamble. |
| 79 |
% \end{abstract} |
| 80 |
% \tableofcontents |
| 81 |
% \def\Hyperlink#1{\hyperlink{#1Description}{\tt \mm #1}} |
| 82 |
% \def\Hypertarget#1#2{\hypertarget{#1Description}{#2}} |
| 83 |
% \xdef\mm{$\backslash\hskip-5pt\backslash$} |
| 84 |
% \section{Introduction} |
| 85 |
% |
| 86 |
% The |csvmerge.sty| merges \TeX\ code with a data file. |
| 87 |
% \begin{itemize} |
| 88 |
% \item |
| 89 |
% The \TeX\ code comes from a blueprint file, which looks like |
| 90 |
% a standard \TeX\ file except it contains no code which should |
| 91 |
% go into the preamble of a normal \TeX\ file. |
| 92 |
% The blueprint file should also not contain a |\begin{document}| |
| 93 |
% nor an |\end{document}|. This stuff goes in the master-file. |
| 94 |
% \item The data file is a standard csv file. |
| 95 |
% \end{itemize} |
| 96 |
% One difference between |csvmerge.sty| and other packages which |
| 97 |
% merge |csv| data into a \TeX\ file is that the user does not need to do |
| 98 |
% much with the |csv|-file. As long as the first row contains reasonable |
| 99 |
% entries describing the contents of its column you are good to go. |
| 100 |
% Extra columns are no problem; subsequent rows can be too short |
| 101 |
% or too long; characters which are special to \TeX\ such as |$| and |&| |
| 102 |
% can appear in the data; \dots\ |
| 103 |
% \noindent More details concerning the |csv|-file can be found in subsection 2.1. |
| 104 |
% |
| 105 |
% The |tex|-file consists of standard \TeX\ code. Each row of the |csv|-file is processed |
| 106 |
% and you have access to the values of all the entries on that row as well as the row |
| 107 |
% number. You also can tell if the entry in a particular column is empty. |
| 108 |
% \noindent More details concerning the |tex|-file can be found in subsection 2.2. |
| 109 |
% |
| 110 |
% The main macro, |\mergeFields|, supports two modes. |
| 111 |
% In one mode, \emph{stream mode}, the \TeX\ code from the |tex|-file |
| 112 |
% is added to the input stream at the location from which |\mergeFields| is invoked. |
| 113 |
% In the second mode, \emph{storage mode}, the code is written to a |
| 114 |
% temporary file. You can then use the standard \LaTeX\ |\input| to read the |
| 115 |
% temporary file back into the master-file. |
| 116 |
% |
| 117 |
% Good models for the two modes are the following. Most mail-merges can be |
| 118 |
% done in streaming mode. Because \TeX\ is doing rather complicated things behind |
| 119 |
% the scenes when it is setting a table or a matrix, you are usually better off using |
| 120 |
% storage mode with your blueprint file containing the code for one line of your table/matrix. |
| 121 |
% After the data is generated you can read it back into your table. |
| 122 |
% See \hyperlink{subsubsection.B.3.1}{here} for a typical example. |
| 123 |
% |
| 124 |
% \newpage |
| 125 |
% \section{Usage} |
| 126 |
% |
| 127 |
% \DescribeMacro{\mergeFields} |
| 128 |
% The main macro is |\mergeFields| which takes one argument. |
| 129 |
% The format is |
| 130 |
%\begin{verbatim} |
| 131 |
% \mergeFields{ |
| 132 |
% tex={<tex-file>} |
| 133 |
% csv = {<csv-file>} |
| 134 |
% tmp = {<tmp-file>} |
| 135 |
% } |
| 136 |
%\end{verbatim} |
| 137 |
% The |tmp| entry is not required. If there is none, |mergeFields| operates in |
| 138 |
% \emph{streaming} mode, otherwise it acts in \emph{storage} mode. |
| 139 |
% The |<prefix-file>| entry should be a path to the corresponding sort of file. |
| 140 |
% If the file name has spaces you will need to quote it: life is easier if there are no |
| 141 |
% spaces in file names. |
| 142 |
% The format |prefix = {<prefix-file>}| is quite forgiving: all leading and trailing spaces |
| 143 |
% are stripped out so |prefix = { <prefix-file> }| works. |
| 144 |
% The three lines can be in any order and do not even need to be on separate lines. |
| 145 |
% The extensions of the three files are irrelevant: the |tex|-file contains \TeX\ code |
| 146 |
% but it can have an extension of |.txt| (or anything else or none at all) if you prefer. |
| 147 |
% All that matters is that |
| 148 |
% when \TeX\ reads the file the contents have the proper content. |
| 149 |
% If you have a simple project you may be able to complete it just by looking at |
| 150 |
% the examples in \hyperlink{appendix.B}{appendix B}. |
| 151 |
% |
| 152 |
% \subsection{The data file} |
| 153 |
% The data file is a |csv| spreadsheet file. |
| 154 |
% By default the field delimiter is the comma and the |
| 155 |
% text delimiter is the double-quotes character. |
| 156 |
% This is default behavior for most spreadsheets when |
| 157 |
% asked to produce a |cvs| file. |
| 158 |
% If there are no {\tt "}'s in your |csv|-file there should be no problems. |
| 159 |
% In some cases you may need a different set of delimiters |
| 160 |
% and this is supported. |
| 161 |
% See \Hyperlink{setDelimitersTabQuote}\ below for more details. |
| 162 |
% |
| 163 |
% The entries in the first row should consist solely of \emph{alphanumeric characters, |
| 164 |
% punctuation and spaces}. |
| 165 |
% Each entry is a \emph{label} for that column. |
| 166 |
% Labels need not be unique although they usually are except sometimes there are |
| 167 |
% several columns with blank labels. |
| 168 |
% |
| 169 |
% A \emph{field-name} is extracted from a label by deleting all spaces and making all |
| 170 |
% characters lowercase: e.g. a column labeled |First Name| has a field-name |
| 171 |
% |firstname|. Commas are also deleted so a column label of |Last, First Name| |
| 172 |
% becomes a field-name of |lastfirstname|. |
| 173 |
% Distinct labels can have the same field-name although again |
| 174 |
% this seems unlikely except for blank labels. |
| 175 |
% |
| 176 |
% Associated to each field-name is a \emph{field-name-macro}. |
| 177 |
% This macro is private: its name is found by prepending |@LRT@@| to the field-name. |
| 178 |
% When |\mergeFields| is invoked, the data file is read one row at a time. |
| 179 |
% All rows after the first are data. |
| 180 |
% When a row is read, each entry is peeled off and placed in the field-name-macro |
| 181 |
% whose field-name is the one associated to the label for that column |
| 182 |
% with |@LRT@| prepended. |
| 183 |
% For example, if row one of some column contains |First Name|, the field-name is |firstname| and |
| 184 |
% the field-name-macro is |\@LRT@@firstname|. When a row after the first is read, |
| 185 |
% |\@LRT@@firstname| will expand to the entry in the |First Name| column on that row. |
| 186 |
% |
| 187 |
% When |\mergeFields| has read a row and set the field-name-macros the |tex|-file is |
| 188 |
% read and the \TeX\ code evaluated. |
| 189 |
% In streaming mode the result is added to the input stream at |
| 190 |
% the current location whereas in storage mode the result is written to the |tmp|-file. |
| 191 |
% |
| 192 |
% \DescribeMacro{End of csv-file}\Hypertarget{End of csv-file} |
| 193 |
% This process continues until the last line in the |csv|-file is processed OR until an entirely |
| 194 |
% blank line is encountered. |
| 195 |
% \vskip10pt |
| 196 |
% The request for only alphanumeric, punctuation and blanks for entries in row one is |
| 197 |
% not actually enforced but the macro |caselower| from |stringstrings| is used to |
| 198 |
% produce lowercase letters and may do weird things to characters not on the requested list. |
| 199 |
% Without studying the |stringstrings| manual carefully % |
| 200 |
% you may not be able to identify the field-names if you insist on weird characters. |
| 201 |
% |
| 202 |
% It is quite possible that several columns can have the same field-name, and hence |
| 203 |
% the same field-name-macro. In that case the macro will expand to the contents of the |
| 204 |
% column with that field-name which is furthest to the right. |
| 205 |
% |
| 206 |
% Because of the prepended |@LRT@@| you can not use these field-name-macros directly |
| 207 |
% in your |tex|-file. |
| 208 |
% The solution is to use \Hyperlink{Field}{}|{field-name}| which will expand correctly. |
| 209 |
% See the examples in appendix B. |
| 210 |
% \subsubsection{Entry issues} |
| 211 |
% Except for issues discussed here, you should be able to use any |
| 212 |
% standard |csv|-file. |
| 213 |
% |
| 214 |
% The data file is read with the usual \TeX\ issues turned off: |
| 215 |
% |#|, |$|, |&|, |@|, |%| |
| 216 |
% , |^|, |~|, and \verb!|! |
| 217 |
% can all be in your data file. |
| 218 |
% \begin{itemize} |
| 219 |
% \item |
| 220 |
% In streaming mode, only |#|, |$|, |&|, |@|, |%|, |^|, |~| print correctly. |
| 221 |
% The characters \$, \#, \%, @ and \& often occur in data and they print correctly. |
| 222 |
% \TeX\ code will not be rendered correctly so for example, |\$| in the data file produces |
| 223 |
% {$``$\tt\$} when \TeX ed. |
| 224 |
% As long as you are not too creative in your entries you should get the results you want. |
| 225 |
% \item |
| 226 |
% In storage mode the characters above end up printing correctly in the |tmp|-file |
| 227 |
% BUT when you |\input| the |tmp|-file back into your master-file these characters |
| 228 |
% will resume their usual \TeX\ meaning and wreak havoc with your code. |
| 229 |
% If you absolutely must have them, writing |\&|, |\$|, |\#|, |
| 230 |
% etc. in the |csv|-file may save you. |
| 231 |
% \end{itemize} |
| 232 |
% |
| 233 |
% \DescribeMacro{Delimiters}\null\hskip-\parindent |
| 234 |
% There is one more issue. For most csv files, the field separator is the comma. |
| 235 |
% When a field has a comma in it, the entire field is delimited by double quotes ({\tt"}). |
| 236 |
% When the field has a double quote it is converted to a pair of double quotes ({\tt""}) and so on. |
| 237 |
% If you have both commas and quotes in a field the best case scenario is that |
| 238 |
% you will get |""|'s where you were hoping for |"|'s. This probably is not what you want. |
| 239 |
% Even worse, a |",| in a data entry will end the entry prematurely at this point |
| 240 |
% and things deteriorate from there. |
| 241 |
% |
| 242 |
% There is a mechanism to deal with this. Most spreadsheets can save csv files with other |
| 243 |
% field (and text) delimiters. The second most common version of csv has a tab as the |
| 244 |
% field delimiter and quotes for fields which have a tab in a field. |
| 245 |
% |
| 246 |
% These two cases are supported out of the box. By default you get the comma-quotes version. |
| 247 |
% If you add |\setDelimitersTabQuote| before calling |\mergeFields| you will get |
| 248 |
% the tab-quotes version. |
| 249 |
% If you have a second call to |\mergeFields| and the data file is comma-quotes, |
| 250 |
% adding |\setDelimitersCommaQuote| before |\mergeFields| will restore comma-quotes. |
| 251 |
% See \hyperlink{setDelimitersCommaQuoteDescription}{setDelimters\dots} |
| 252 |
% for a more thorough discussion. |
| 253 |
% |
| 254 |
% |
| 255 |
% \subsection{The blueprint file} |
| 256 |
% |
| 257 |
% The blueprint file is a normal \TeX\ file with macros defined as above from the |csv|-file. |
| 258 |
% It should not contain a \verb!\begin{document}! or an \verb!\end{document}! |
| 259 |
% nor anything that needs to be expanded in the preamble. |
| 260 |
% Such things belong in the master file. |
| 261 |
% |
| 262 |
% For mail-merges, the |tex|-file should end with a \verb!\newpage! |
| 263 |
% to guarantee that each copy of the message appears on a |
| 264 |
% different page. |
| 265 |
% If you need custom sizes or headers or you have page numbers that you want to |
| 266 |
% reset for each copy, you should do this in the |tex|-file. |
| 267 |
% |
| 268 |
% To use the data from the |csv|-file you need to write |\Field{some-field-name}| wherever |
| 269 |
% you want the data in your |tex|-file. |
| 270 |
% Some field-names may be rather long so if you are going to use them a lot in your |
| 271 |
% |tex|-file, feel free to write something like |\def\sfn{\Field{some-field-name}}| at the |
| 272 |
% top of your |tex|-file and then use |\sfn| thereafter. |
| 273 |
% |
| 274 |
% One additional issue occurs in storage mode. You may very well want \TeX\ code in your |
| 275 |
% |tex|-file. When |mergeFields| reads a line of data, you want the field-name-macros to |
| 276 |
% expand to their current values but you often do not want your \TeX\ code to expand. |
| 277 |
% It is good practice stick a |\noexpand| in front of each \TeX\ macro. |
| 278 |
% Each line is expanded once when the line is written to the |tmp| file, so the |\noexpand| |
| 279 |
% disappears and you are left with just your \TeX\ macros. |
| 280 |
% |
| 281 |
% If you do define macros like |\sfn| as above, you will need |
| 282 |
% to write\\ |\noexpand\def\noexpand\sfn{\Field{some-field-name}}|. |
| 283 |
% \StopEventually{\PrintIndex} |
| 284 |
% |
| 285 |
% \section{List of public macros} |
| 286 |
% There are the 6 public macros: links are to descriptions of the |@LRT@|-versions which |
| 287 |
% actually do the work. |
| 288 |
% |
| 289 |
% \null\hskip-10pt \begin{tabular}{lll} |
| 290 |
% \Hyperlink{mergeFields}% |
| 291 |
% &\Hyperlink{Field}% |
| 292 |
% &\hbox to 0pt{\Hyperlink{ifFieldEmpty}\hss}% |
| 293 |
% \\ |
| 294 |
% \Hyperlink{setDelimitersCommaQuote}&\Hyperlink{setDelimitersTabQuote}% |
| 295 |
% &\Hyperlink{makeMePublic}\\ |
| 296 |
% \end{tabular} |
| 297 |
% \begin{itemize} |
| 298 |
% \item |\mergeFields{tex={tex-file}csv={csv-file}}| or |
| 299 |
% \\|\mergeFields{tex={tex-file}csv={csv-file}{tmp={tmp-file}}|\\ |
| 300 |
% sets up to do the merge. |
| 301 |
% \hyperlink{section.2}{Section 2} is devoted to a detailed discussion. |
| 302 |
% \item |\Field{field-name}| expands to the current value of the field-name-macro. |
| 303 |
% \item |\ifFieldEmpty{field-name to test}{do if empty}{do if not}| |
| 304 |
% \item The two |\setDelimiters| set the delimiters to match those used in the |csv| file: |
| 305 |
% comma-quotes is the default. |
| 306 |
% \item |\makeMePublic| must be immediately followed by |\@LRT@| or you will get a |
| 307 |
% |Use of \makeMePublic doesn't match its definition.| error. Then you can write |
| 308 |
% |\makeMePublic \@LRT@foo\becomes\Foo| to make the public macro |\Foo| that |
| 309 |
% you just define equal to the private macro |\@LRT@foo|. You may not write |
| 310 |
% |\makeMePublic{\@LRT@foo}\becomes\Foo| or other variants: the first non-white-space |
| 311 |
% character \TeX\ sees after |\makeMePublic| needs to be the start of |\@LRT@|. |
| 312 |
% |
| 313 |
% As a useful example, |
| 314 |
% |\makeMePublic\@LRT@Row\becomes\Row| means you can get your hands on |
| 315 |
% the row number using |\Row|. |
| 316 |
% \end{itemize} |
| 317 |
% |
| 318 |
% \section{Implementation} |
| 319 |
% |
| 320 |
% \begin{macro}{Header} |
| 321 |
% First we have the standard package header stuff. |
| 322 |
% The line numbers in the typeset \jobname.dtx file match the line numbers |
| 323 |
% in the \jobname.sty file |
| 324 |
% generated by \TeX ing the \jobname.ins file. |
| 325 |
% \setcounter{CodelineNo}{15} |
| 326 |
% \begin{macrocode} |
| 327 |
\NeedsTeXFormat{LaTeX2e}[2005/12/01] |
| 328 |
\ProvidesPackage{csvmerge} |
| 329 |
[2019/07/17 v1.0 merges TeX code with csv data] |
| 330 |
|
| 331 |
% \end{macrocode} |
| 332 |
% \end{macro} |
| 333 |
% \begin{macro}{Required Packages} |
| 334 |
% These required packages are part of the standard \href{https://www.tug.org/texlive/}{{\tt TeX Live}} |
| 335 |
%and \href{https://miktex.org}{{\tt MiKTEX}} distributions. |
| 336 |
% \begin{macrocode} |
| 337 |
\RequirePackage{stringstrings} |
| 338 |
\RequirePackage{trimspaces} |
| 339 |
\RequirePackage{etoolbox} |
| 340 |
|
| 341 |
% \end{macrocode} |
| 342 |
% \end{macro} |
| 343 |
% \begin{macro}{Options} |
| 344 |
% |csvmerge| supports an option which is rarely needed. All the private macros |
| 345 |
% (except \hyperlink{exceptionDescription}{one}) in |
| 346 |
% this package have |@LRT@| or |@lrt@| somewhere in their name so are very |
| 347 |
% unlikely to produce name conflicts. In the very unlikely chance that one or more of |
| 348 |
% the six public macros has a name conflict, |
| 349 |
% this can be alleviated by putting |\usepackage[foo]{csvmerge}| |
| 350 |
% in your master file. Here |foo| can be any string of alphabetical characters. |
| 351 |
% Then the names of all six public macros are invoked with a leading |foo|: |
| 352 |
% |\mergeFields| no longer exists but becomes |\foomergeFields|. The evident change happens |
| 353 |
% with the other five macros. |
| 354 |
% |
| 355 |
% The next two lines of code merely save the option for use |
| 356 |
% \hyperlink{optionsDescription}{later} at the end. |
| 357 |
% \begin{macrocode} |
| 358 |
\DeclareOption*{\xdef\@LRT@CO{\CurrentOption}} |
| 359 |
\ProcessOptions\relax |
| 360 |
|
| 361 |
% \end{macrocode} |
| 362 |
% \end{macro} |
| 363 |
% \begin{macrocode} |
| 364 |
\makeatletter |
| 365 |
|
| 366 |
% \end{macrocode} |
| 367 |
% \begin{macro}{\mergeFields} |
| 368 |
% \Hypertarget{mergeFields} |
| 369 |
% |\@LRT@mergeFields| actually initiates the work. |
| 370 |
% It initializes some macros for later use. |
| 371 |
% Then it collects the |prefix={<prefix-file>}| arguments using |\@LRT@getFilePaths| |
| 372 |
% and checks to be sure it has a |csv|-file |
| 373 |
% and a |tex|-file. If not it bails. |
| 374 |
% Then it checks if it has a |tmp| file and sets the file reading subroutine, |
| 375 |
% |\@LRT@Input| to |\@LRT@streaming| is there is no |tmp| file and to |
| 376 |
% |\@LRT@storage| if there is. If there is a |tmp| file, it is opened. |
| 377 |
% Then |\@LRT@mergeData| is called which actually does the merging. |
| 378 |
% Finally, if there is a |tmp| file, it is closed. |
| 379 |
% \begin{macrocode} |
| 380 |
\newcommand{\@LRT@mergeFields}[1]{% |
| 381 |
\gdef\@LRT@csv{}% |
| 382 |
\gdef\@LRT@tex{}% |
| 383 |
\gdef\@LRT@tmp{}% |
| 384 |
\setcounter{\@LRT@macroPrefix Row}{2}\relax% |
| 385 |
\gdef\@LRT@csvBlankLine{}% |
| 386 |
\@LRT@getFilePaths#1=@% |
| 387 |
\if@LRT@empty{\@LRT@csv}{\@LRT@missingArgumentMessage{csv}}{}% |
| 388 |
\if@LRT@empty{\@LRT@tex}{\@LRT@missingArgumentMessage{tex}}{}% |
| 389 |
\if@LRT@empty{\@LRT@tmp}% |
| 390 |
{\global\let\@LRT@Input\@LRT@streaming}% |
| 391 |
{\global\let\@LRT@Input\@LRT@storage}% |
| 392 |
\if@LRT@empty{\@LRT@tmp}{}{% |
| 393 |
\immediate\openout\@lrt@tmpFile=\@LRT@tmp\relax}% |
| 394 |
\@LRT@mergeData% |
| 395 |
\if@LRT@empty{\@LRT@tmp}{}{\immediate\closeout\@lrt@tmpFile\relax}% |
| 396 |
} |
| 397 |
|
| 398 |
% \end{macrocode} |
| 399 |
% \end{macro} |
| 400 |
% \Hypertarget{macroPrefix}{} |
| 401 |
% \begin{macro}{\@LRT@macroPrefix} |
| 402 |
% \begin{macrocode} |
| 403 |
\gdef\@LRT@macroPrefix{@LRT@@} |
| 404 |
|
| 405 |
% \end{macrocode} |
| 406 |
% \end{macro} |
| 407 |
% \Hypertarget{Field}{} |
| 408 |
% \begin{macro}{\Field} |
| 409 |
% |\@LRT@Field| is just a wrapper for a |\csname ... \endcsname|. |
| 410 |
% Notice it adds the |\@LRT@macroPrefix| prefix to the argument so |
| 411 |
% the input is always the field-name and the output is always the value of |
| 412 |
% the corresponding field-name-macro. |
| 413 |
% Because of the third |@| there is no danger that these macros defined from |
| 414 |
% entries in the |csv|-file will conflict with other macros defined in this package |
| 415 |
% since none of them have a |@@| in their names. |
| 416 |
% \begin{macrocode} |
| 417 |
\def\@LRT@Field#1{% |
| 418 |
\csname\@LRT@macroPrefix#1\endcsname% |
| 419 |
} |
| 420 |
|
| 421 |
% \end{macrocode} |
| 422 |
% \end{macro} |
| 423 |
% |
| 424 |
% \begin{macro}{\ifFieldEmpty} |
| 425 |
% \Hypertarget{ifFieldEmpty}{} |
| 426 |
% |\ifFieldEmpty| takes three arguments. |
| 427 |
% The first argument is a field-name and if the content of the corresponding |
| 428 |
% macro is empty then the second argument is executed: otherwise the |
| 429 |
% third argument is executed. |
| 430 |
% |
| 431 |
% \begin{macrocode} |
| 432 |
\long\gdef\@LRT@ifFieldEmpty#1#2#3{\ifbool{\@LRT@macroPrefix#1}{#2}{#3}} |
| 433 |
|
| 434 |
% \end{macrocode} |
| 435 |
% \end{macro} |
| 436 |
% |
| 437 |
% \begin{macro}{\setDelimitersCommaQuote} |
| 438 |
%\end{macro} |
| 439 |
% \begin{macro}{\setDelimitersTabQuote} |
| 440 |
% \Hypertarget{setDelimitersCommaQuote}{} |
| 441 |
% \Hypertarget{setDelimitersTabQuote}{} |
| 442 |
% These two |setDelimiters| are similar. |
| 443 |
% Recall how delimiters work. A row in a |csv|-file with comma delimiters |
| 444 |
% looks like |foo1,foo2,foo3,...| so the delimiter is the comma and if |
| 445 |
% |\@LRT@getFieldDelimited| is put in front of this string it will peal off |
| 446 |
% |foo1|. Since |^^I| is \TeX\ for the tab, |
| 447 |
% a row in a csv file with tab delimiters |
| 448 |
% looks like |foo1^^Ifoo2^^Ifoo3^^I...| and so the first two lines of |
| 449 |
% |\@LRT@setDelimitersCommaQuote| and |\@LRT@setDelimitersTabQuote| |
| 450 |
% look the same after switching |,| and |^^I|. |
| 451 |
% |
| 452 |
% There is a wrinkle in the tab case. The tab normally looks to \TeX\ like white space and |
| 453 |
% so by the time the line reaches |\@LRT@getFieldDelimited| the line looks like |
| 454 |
% |foo1foo2foo3...| and there is no way to break off |foo1|. The |\catcode`\^^I12\relax| |
| 455 |
% insures |\@LRT@getFieldDelimited| sees the line as |foo1^^Ifoo2^^Ifoo3^^I...|.. |
| 456 |
% This means that the catcode of |^^I| must also be 12 when each line is read and the line |
| 457 |
% and the macro |\gdef\@LRT@catcodes{|\\|\catcode`\^^I12\relax}| insures this. |
| 458 |
% In the comma delimited version no catcode heroics are needed so |\@LRT@catcodes| |
| 459 |
% is empty. |
| 460 |
% |
| 461 |
% If some field contains one or more |\@LRT@fieldDelimiterBack|'s then the |csv|-file will use |
| 462 |
% the text-delimiter at the front and back. When quote is the text delimiter and |foo2| contains |
| 463 |
% a comma but |foo1| and |foo3| do not, the comma delimited string looks like |
| 464 |
% |foo1,"foo2",foo3,...| and the tab delimited string looks like |
| 465 |
% |foo1^^I"foo2"^^Ifoo3^^I...|. |
| 466 |
% |
| 467 |
% The text delimiters have a front one, the double-quotes in both versions, and the back one which is |
| 468 |
% the front text delimiter followed by the back field delimiter. |
| 469 |
% |
| 470 |
% The macro |\@LRT@blankFieldItem| is used to construct what a blank line looks like. |
| 471 |
% In the comma delimited case, you need a string of commas of the same length as |
| 472 |
% the number of entries on line one of the |csv|-file. In the tab delimited case, the blank line |
| 473 |
% turns out to be a blank macro. |
| 474 |
% |
| 475 |
% The |\@LRT@catcodes| macro is a collection of catcodes that need to be changed: |
| 476 |
% nothing for the comma delimited case and |^^I| for the tab delimited case. |
| 477 |
% |
| 478 |
% See \hyperlink{newDelimiterAppendix}{Appendix A} for more discussion on writing your own delimiters. |
| 479 |
% \begin{macrocode} |
| 480 |
{{% |
| 481 |
\gdef\@LRT@setDelimitersCommaQuote{ |
| 482 |
\gdef\@LRT@fieldDelimiterBack{,} |
| 483 |
\gdef\@LRT@getFieldDelimited##1,{\@LRT@getFirstArg{##1}} |
| 484 |
\gdef\@LRT@textDelimiterFront{"} |
| 485 |
\gdef\@LRT@getTextDelimited"##1",{\@LRT@getFirstArg{##1}} |
| 486 |
\gdef\@LRT@blankFieldItem{,} |
| 487 |
\gdef\@LRT@catcodes{} |
| 488 |
}% |
| 489 |
}} |
| 490 |
|
| 491 |
{{\catcode`\^^I12\relax% |
| 492 |
\gdef\@LRT@setDelimitersTabQuote{ |
| 493 |
\gdef\@LRT@fieldDelimiterBack{^^I} |
| 494 |
\gdef\@LRT@getFieldDelimited##1^^I{\@LRT@getFirstArg{##1}} |
| 495 |
\gdef\@LRT@textDelimiterFront{"} |
| 496 |
\gdef\@LRT@getTextDelimited"##1"^^I{\@LRT@getFirstArg{##1}} |
| 497 |
\gdef\@LRT@blankFieldItem{} |
| 498 |
\gdef\@LRT@catcodes{\catcode`\^^I12\relax} |
| 499 |
}% |
| 500 |
}} |
| 501 |
|
| 502 |
% \end{macrocode} |
| 503 |
% \end{macro} |
| 504 |
% |
| 505 |
% \begin{macro}{\@LRT@makeMePublic} |
| 506 |
% \hypertarget{exceptionDescription}{} |
| 507 |
% \hypertarget{makeMePublicDescription}{} |
| 508 |
% The next line of code defines the only macro which does not have |@LRT@| |
| 509 |
% in its name. It seems sufficiently weird that there should be no name conflicts. |
| 510 |
% Notice that |\makeatletter| is still in effect so |@| has catcode 11. |
| 511 |
% |
| 512 |
% Then we switch to |\makeatother| so |@| has catcode 12. |
| 513 |
% The next line expands to |\@LRT@makeMePublic| with the initial |
| 514 |
% |\@LRT@| having its |@|'s of catcode 11 so it is a private macro. |
| 515 |
% In the |\@LRT@| that you see, the |@|'s have catcode 12 so it is really |
| 516 |
% the macro |\@| with extra stuff. Since it should never be expanded |
| 517 |
% it doesn't really matter. Then we drop back into |\makeatletter| to define |
| 518 |
% more private macros. |
| 519 |
% |
| 520 |
% The net result is that you can write \\ |
| 521 |
% \null\hskip1in |\makeMePublic \@LRT@foo\becomes\newMacro| \\in your master-file and |
| 522 |
% the private macro |\@LRT@foo| becomes the same as |\newMacro| and you can |
| 523 |
% use |\newMacro| in your |tex|-file. |
| 524 |
% \begin{macrocode} |
| 525 |
\expandafter\gdef\csname1238LRTLRTsvbneLRT\endcsname{@LRT@} |
| 526 |
\makeatother |
| 527 |
\expandafter\gdef\csname\csname1238LRTLRTsvbneLRT\endcsname |
| 528 |
makeMePublic\endcsname\@LRT@#1\becomes#2{\global\letcs{#2}{@LRT@#1}}% |
| 529 |
\makeatletter |
| 530 |
|
| 531 |
% \end{macrocode} |
| 532 |
% \end{macro} |
| 533 |
% \begin{macro}{private macros} |
| 534 |
% \end{macro} |
| 535 |
% \begin{macro}{\@LRT@Row} |
| 536 |
% \Hypertarget{Row}{} |
| 537 |
% The macro |\@LRT@Row| expands to the number of the row of the |csv|-file currently |
| 538 |
% being read. To get the |\xdef| to expand correctly we needed to define (and did) |
| 539 |
% |\@LRT@macroPrefix| first. Recall that the first line of actual data is row 2. |
| 540 |
% \begin{macrocode} |
| 541 |
\xdef\@LRT@Row{\noexpand\number\noexpand\value{\@LRT@macroPrefix Row}} |
| 542 |
|
| 543 |
% \end{macrocode} |
| 544 |
% \end{macro} |
| 545 |
% \begin{macro}{\if@LRT@empty}\Hypertarget{if@LRT@empty}{} |
| 546 |
% Takes three arguments: if |#1| expands to empty do the code in |#2|; |
| 547 |
% otherwise do the code in |#3|. |
| 548 |
% \begin{macrocode} |
| 549 |
\newcommand{\if@LRT@empty}[3]{\ifthenelse{\equal{#1}{}}{#2}{#3}} |
| 550 |
|
| 551 |
% \end{macrocode} |
| 552 |
% \end{macro} |
| 553 |
% \begin{macro}{\@LRT@fieldName} |
| 554 |
% Takes an alphanumeric string, uses |\caselower| from the |stringstrings| package |
| 555 |
% to convert the string to all lower case; then it strips the spaces from the answer |
| 556 |
% using |\convertchar|, also from the |stringstrings| package. Next it strips commas from |
| 557 |
% the answer. |
| 558 |
% The final |\@LRT@TheString| makes the answer global. |
| 559 |
% \begin{macrocode} |
| 560 |
\gdef\@LRT@fieldName#1{% |
| 561 |
\caselower[q]{#1}% |
| 562 |
\convertchar[q]{\thestring}{ }{}% |
| 563 |
\convertchar[q]{\thestring}{,}{}% |
| 564 |
\xdef\@LRT@TheString{\thestring}% |
| 565 |
} |
| 566 |
|
| 567 |
% \end{macrocode} |
| 568 |
% \end{macro} |
| 569 |
% \begin{macro}{\@LRT@getFilePaths} |
| 570 |
% |\@LRT@getFilePaths#1=#2@| is a recursive macro which works through the |
| 571 |
% |prefix={<prefix-file>}| argument list. |
| 572 |
% The first argument, |#1|, picks up one |prefix={<prefix-file>}| and |#2| contains the rest of the arguments. |
| 573 |
% If |#1| is not empty it is passed to |\@LRT@extractKeyValue| which collects the |
| 574 |
% prefix and file path. |
| 575 |
% \begin{macrocode} |
| 576 |
\gdef\@LRT@getFilePaths#1=#2@{% |
| 577 |
\xdef\@LRT@hashKey{\trim@spaces{#1}}% |
| 578 |
\if@LRT@empty{\@LRT@hashKey}{}{ |
| 579 |
\def\@LRT@test{tex}\ifthenelse{\equal{\@LRT@hashKey}{\@LRT@test}}{}% |
| 580 |
{\def\@LRT@test{csv}\ifthenelse{\equal{\@LRT@hashKey}{\@LRT@test}}{}% |
| 581 |
{\def\@LRT@test{tmp}\ifthenelse{\equal{\@LRT@hashKey}{\@LRT@test}}{}% |
| 582 |
{\@LRT@unknownArgumentMessage{#1}}}}% |
| 583 |
\@LRT@extractKeyValue#2@% |
| 584 |
\expandafter\@LRT@getFilePaths\@LRT@doWeStillHaveStuffToDo @}} |
| 585 |
|
| 586 |
% \end{macrocode} |
| 587 |
% \end{macro} |
| 588 |
% \begin{macro}{\@LRT@extractKeyValue} |
| 589 |
% |\@LRT@hashKey| contains the prefix and |\@LRT@hashValue| contains the file path. |
| 590 |
% If the prefix is not one of |tex|, |csv| or |tmp| you will be warned. |
| 591 |
% \begin{macrocode} |
| 592 |
\def\@LRT@extractKeyValue#1#2@{% |
| 593 |
\edef\@LRT@hashValue{\trim@spaces{#1}}% |
| 594 |
\xdef\@LRT@doWeStillHaveStuffToDo{\trim@spaces{#2}}% |
| 595 |
\expandafter\xdef\csname @LRT@\@LRT@hashKey\endcsname{\@LRT@hashValue}% |
| 596 |
} |
| 597 |
|
| 598 |
% \end{macrocode} |
| 599 |
% \end{macro} |
| 600 |
% \begin{macro}{\@LRT@SecondifFieldEmpty} |
| 601 |
% |\@LRT@SecondifFieldEmpty| checks for an empty macro directly. |
| 602 |
% The previously defined \Hyperlink{ifFieldEmpty} uses a different method so both are needed at |
| 603 |
% various points in the code. |
| 604 |
% \begin{macrocode} |
| 605 |
\newcommand{\@LRT@SecondifFieldEmpty}[3]{% |
| 606 |
\edef\@LRT@fieldValue{#1}% |
| 607 |
\ifx\@LRT@fieldValue\@empty\relax{#2}\else{#3}\fi\relax} |
| 608 |
|
| 609 |
% \end{macrocode} |
| 610 |
% \end{macro} |
| 611 |
% \begin{macro}{\@LRT@getCommaFieldDelimited} |
| 612 |
% The macros |\@LRT@getCommaFieldDelimited|and |
| 613 |
% |\@LRT@getFieldDelimited| are the same when the field delimiter is the comma. |
| 614 |
% The list of field-names is always comma delimited and so |
| 615 |
% we need a |\@LRT@getCommaFieldDelimited| to iterate through it. |
| 616 |
% \begin{macrocode} |
| 617 |
\gdef\@LRT@getCommaFieldDelimited#1,{\@LRT@getFirstArg{#1}} |
| 618 |
|
| 619 |
% \end{macrocode} |
| 620 |
% \end{macro} |
| 621 |
% \begin{macro}{New counter} The counter containing the row number. |
| 622 |
% \begin{macrocode} |
| 623 |
\newcounter{\@LRT@macroPrefix Row} |
| 624 |
|
| 625 |
% \end{macrocode} |
| 626 |
% \end{macro} |
| 627 |
% \begin{macro}{New read/write's} |
| 628 |
% \begin{macrocode} |
| 629 |
\newread\@lrt@csvFile% |
| 630 |
\newread\@lrt@texFile% |
| 631 |
\newwrite\@lrt@tmpFile% |
| 632 |
|
| 633 |
% \end{macrocode} |
| 634 |
% \end{macro} |
| 635 |
% \begin{macro}{Messages} |
| 636 |
% \begin{macrocode} |
| 637 |
\gdef\@LRT@justProcessedMessage{% |
| 638 |
\immediate\typeout{}% |
| 639 |
\immediate\typeout{Just processed row % |
| 640 |
\@LRT@Row\space of the data file.}% |
| 641 |
\immediate\typeout{}% |
| 642 |
\stepcounter{\@LRT@macroPrefix Row}\relax% |
| 643 |
} |
| 644 |
|
| 645 |
\gdef\@LRT@failedToOpenMessage#1{% |
| 646 |
\immediate\typeout{}% |
| 647 |
\immediate\typeout{File #1 failed to open.}% |
| 648 |
\immediate\typeout{}\fi% |
| 649 |
\end{document}% |
| 650 |
} |
| 651 |
|
| 652 |
\gdef\@LRT@NoMoreDataMessage{% |
| 653 |
\immediate\typeout{No more data.}% |
| 654 |
\immediate\typeout{}% |
| 655 |
} |
| 656 |
|
| 657 |
\gdef\@LRT@missingArgumentMessage#1{ |
| 658 |
\immediate\typeout{}% |
| 659 |
\immediate\typeout{I was never given a path to the #1 file.}% |
| 660 |
\immediate\typeout{}% |
| 661 |
\end{document}% |
| 662 |
} |
| 663 |
|
| 664 |
\gdef\@LRT@unknownArgumentMessage#1{ |
| 665 |
\immediate\typeout{}% |
| 666 |
\immediate\typeout{I do not recognize an argument of type #1.}% |
| 667 |
\immediate\typeout{}% |
| 668 |
} |
| 669 |
|
| 670 |
% \end{macrocode} |
| 671 |
% \end{macro} |
| 672 |
% \begin{macro}{Inputs} |
| 673 |
% When |mergeFields| starts up, it knows immediately whether it is in streaming or in storage |
| 674 |
% mode. It sets |\@LRT@Input| to |\@LRT@streaming| or |\@LRT@storage|. |
| 675 |
% Each time a new line of data in the |csv|-file is read the |
| 676 |
% the macros are updated to reflect that data on the current line. |
| 677 |
% |\@LRT@streaming| uses the \LaTeX\ command |\input| to add the |tex|-file to the |
| 678 |
% stream. |
| 679 |
% |\@LRT@storage| works through the data in the |tex|-file. |
| 680 |
% First it opens the |tex| file; then it |
| 681 |
% reads the |tex|-file one line at a time; expands that line; and |
| 682 |
% writes the answer to a the |tmp|-file. When the |eof| for the |tex|-file is reached, |
| 683 |
% the |tex|-file is closed. |
| 684 |
% |
| 685 |
% |\@LRT@catcodeMagic| is used in two places to set the catcodes of certain characters |
| 686 |
% to values that won't cause trouble. |
| 687 |
% |
| 688 |
% |\@LRT@texNotAtEOF{#1}| is used to tighten up the code looping over the lines in |
| 689 |
% the |tex|-file in |\@LRT@storage|. |
| 690 |
% \begin{macrocode} |
| 691 |
\gdef\@LRT@streaming{\input{\@LRT@tex}} |
| 692 |
|
| 693 |
\gdef\@LRT@catcodeMagic{% |
| 694 |
\catcode`\^^M9\relax% |
| 695 |
\catcode`\#12\relax |
| 696 |
\catcode`\$12\relax |
| 697 |
\catcode`\&12\relax |
| 698 |
\catcode`\^12\relax |
| 699 |
\catcode`\_12\relax |
| 700 |
\catcode`\%12\relax |
| 701 |
} |
| 702 |
|
| 703 |
\gdef\@LRT@texNotAtEOF#1{% |
| 704 |
\ifeof\@lrt@texFile{\global\booltrue{@LRT@texEOF}}\else{#1}\fi% |
| 705 |
} |
| 706 |
|
| 707 |
\gdef\@LRT@storage{{% |
| 708 |
\@LRT@catcodeMagic% |
| 709 |
\openin\@lrt@texFile=\@LRT@tex\relax% |
| 710 |
\@LRT@texNotAtEOF{\global\boolfalse{@LRT@texEOF}}% |
| 711 |
\unlessboolexpr{bool{@LRT@texEOF}}{% |
| 712 |
\read\@lrt@texFile to \@LRT@preItem% |
| 713 |
\@LRT@texNotAtEOF{% |
| 714 |
\immediate\write\@lrt@tmpFile{\@LRT@preItem}\relax}% |
| 715 |
}{}% |
| 716 |
\immediate\closein\@lrt@texFile% |
| 717 |
}} |
| 718 |
|
| 719 |
% \end{macrocode} |
| 720 |
% \end{macro} |
| 721 |
% \begin{macro}{\@LRT@dataLineIterator} |
| 722 |
% This macro looks to see if the next character is |\@LRT@textDelimiterFront|. |
| 723 |
% If it is the entry is collected using |\@LRT@getTextDelimited| and if not we collect it |
| 724 |
% using |\@LRT@getFieldDelimited|. |
| 725 |
% \begin{macrocode} |
| 726 |
\gdef\@LRT@dataLineIterator{% |
| 727 |
\expandafter\@ifnextchar\@LRT@textDelimiterFront% |
| 728 |
{\@LRT@getTextDelimited}{\@LRT@getFieldDelimited}} |
| 729 |
|
| 730 |
% \end{macrocode} |
| 731 |
% \end{macro} |
| 732 |
% \begin{macro}{\@LRT@getFirstArg} |
| 733 |
% This is the |\@LRT@getFirstArg| from the |\@LRT@getFieldDelimited| and \\ |
| 734 |
% |\@LRT@getTextDelimited| macros. It collects the first field and then |
| 735 |
% uses |\@LRT@getSecondArg| to collect the rest of the line from the |csv|-file. |
| 736 |
% \begin{macrocode} |
| 737 |
\gdef\@LRT@getFirstArg#1{\gdef\@LRT@firstArg{#1}\@LRT@getSecondArg} |
| 738 |
\gdef\@LRT@getSecondArg#1@{\gdef\@LRT@restArg{#1}} |
| 739 |
|
| 740 |
% \end{macrocode} |
| 741 |
% \end{macro} |
| 742 |
% \begin{macro}{\@LRT@getLOFirstArg} |
| 743 |
% These next two macros do the same thing with the first line of the |csv|-file. |
| 744 |
% The difference is that these two macros put their answers in a different pair of |
| 745 |
% macros than the previous two so we can iterate along both the first line (field-names) |
| 746 |
% and the current line of the |csv|-file. |
| 747 |
% \begin{macrocode} |
| 748 |
\gdef\@LRT@getLOFirstArg#1{\gdef\@LRT@ffirstArg{#1}\@LRT@getLOSecondArg} |
| 749 |
\gdef\@LRT@getLOSecondArg#1@{\gdef\@LRT@frestArg{#1}} |
| 750 |
|
| 751 |
% \end{macrocode} |
| 752 |
% \end{macro} |
| 753 |
% \begin{macro}{\@LRT@getLineOne} |
| 754 |
% This is the version of |\@LRT@getFirstArg| which is used on the first line of the |csv|-file. |
| 755 |
% \begin{macrocode} |
| 756 |
\gdef\@LRT@fieldNameIterator#1,{\@LRT@getLOFirstArg{#1}} |
| 757 |
|
| 758 |
% \end{macrocode} |
| 759 |
% \end{macro} |
| 760 |
% \begin{macro}{\@LRT@setFirstLineAndBlankLine} |
| 761 |
% This macro collects the first line of the |csv|-file and packages it for later use: |
| 762 |
% |\@lrt@firstline| is the first line when we are done. |
| 763 |
% We also construct |\@LRT@csvBlankLine| which is used to check if |
| 764 |
% a line of the |csv|-file is blank. |
| 765 |
% \begin{macrocode} |
| 766 |
\providebool{@LRT@iterator} |
| 767 |
\gdef\@LRT@setFirstLineAndBlankLine#1{% |
| 768 |
\gdef\@lrt@firstline{}% |
| 769 |
\gdef\@LRT@csvBlankLine{} |
| 770 |
\edef\@LRT@restArg{#1\@LRT@fieldDelimiterBack} |
| 771 |
\booltrue{@LRT@iterator}% |
| 772 |
\whileboolexpr{bool{@LRT@iterator}}{% |
| 773 |
\expandafter\@LRT@dataLineIterator\@LRT@restArg @% |
| 774 |
\@LRT@fieldName{\@LRT@firstArg}% |
| 775 |
\xdef\@lrt@firstline{\@lrt@firstline\@LRT@TheString,}% |
| 776 |
\if@LRT@empty{\@LRT@restArg}% |
| 777 |
{\boolfalse{@LRT@iterator}}{\booltrue{@LRT@iterator}}% |
| 778 |
\ifboolexpr{bool{@LRT@iterator}}% |
| 779 |
{\xdef\@LRT@csvBlankLine{\@LRT@csvBlankLine\@LRT@blankFieldItem}}{}% |
| 780 |
}{}% |
| 781 |
} |
| 782 |
|
| 783 |
% \end{macrocode} |
| 784 |
% \end{macro} |
| 785 |
% \begin{macro}{\@LRT@getAndProcessNextLine} |
| 786 |
% This is the recursive macro which processes each line of the |csv|-file |
| 787 |
% until there are no more. |
| 788 |
% The code in the first |{{| \dots |}}| pair |
| 789 |
% reads a line from the |csv|-file: the last line here converts |
| 790 |
% the answer from local to global. |
| 791 |
% The code in the second |{{| \dots |}}| pair checks |
| 792 |
% if the line is \hyperlink{End of csv-fileDescription}{blank} or not. |
| 793 |
% |\@LRT@exitGetAndProcessNextLine| exits the recursion and |
| 794 |
% |\@LRT@repeatGetAndProcessNextLine| reloops. |
| 795 |
% \begin{macrocode} |
| 796 |
\gdef\@LRT@getAndProcessNextLine{% |
| 797 |
\unlessboolexpr{bool{@LRT@csvEOF}}{% |
| 798 |
{{% |
| 799 |
\@LRT@catcodeMagic% |
| 800 |
\@LRT@catcodes |
| 801 |
\read\@lrt@csvFile to\@LRT@x@nextLine% |
| 802 |
\ifeof\@lrt@csvFile\relax{\global\booltrue{@LRT@csvEOF}}\else% |
| 803 |
{\global\boolfalse{@LRT@csvEOF}}\fi% |
| 804 |
\xdef\@LRT@nextLine{\@LRT@x@nextLine}% |
| 805 |
}}% |
| 806 |
{{% |
| 807 |
\if@LRT@empty{\@LRT@nextLine}% |
| 808 |
{\global\let\@LRT@next\@LRT@exitGetAndProcessNextLine\relax}% |
| 809 |
{\ifx\@LRT@csvBlankLine\@LRT@nextLine\relax% |
| 810 |
{\global\let\@LRT@next\@LRT@exitGetAndProcessNextLine\relax}\else% |
| 811 |
{\global\let\@LRT@next\@LRT@repeatGetAndProcessNextLine\relax}\fi}% |
| 812 |
\@LRT@next% |
| 813 |
}}% |
| 814 |
}{}% |
| 815 |
} |
| 816 |
|
| 817 |
% \end{macrocode} |
| 818 |
% \end{macro} |
| 819 |
% \begin{macro}{\@LRT@exitGetAndProcessNextLine} |
| 820 |
% This just exits. |
| 821 |
% \begin{macrocode} |
| 822 |
\gdef\@LRT@exitGetAndProcessNextLine{% |
| 823 |
\closein\@lrt@csvFile\relax\global\booltrue{@LRT@csvEOF}} |
| 824 |
|
| 825 |
% \end{macrocode} |
| 826 |
% \end{macro} |
| 827 |
% \begin{macro}{\@LRT@repeatGetAndProcessNextLine} |
| 828 |
% This sets up to process the line we just read. |
| 829 |
% \begin{macrocode} |
| 830 |
\gdef\@LRT@repeatGetAndProcessNextLine{% |
| 831 |
\@LRT@processNextLine{\@LRT@nextLine\@LRT@fieldDelimiterBack}} |
| 832 |
|
| 833 |
% \end{macrocode} |
| 834 |
% \end{macro} |
| 835 |
% \begin{macro}{\@LRT@setMacrosForOneEntry} |
| 836 |
% This bit of code sets the macro and the if-empty boolean for a single item: |
| 837 |
% \#1 is the field-name and \#2 is the entry in that (row,column). |
| 838 |
% \begin{macrocode} |
| 839 |
\gdef\@LRT@setMacrosForOneEntry#1#2{% |
| 840 |
\expandafter\xdef\csname\@LRT@macroPrefix#1\endcsname{#2}% |
| 841 |
\@LRT@SecondifFieldEmpty{#2}% |
| 842 |
{\global\booltrue{\@LRT@macroPrefix#1}}% |
| 843 |
{\global\boolfalse{\@LRT@macroPrefix#1}}% |
| 844 |
} |
| 845 |
|
| 846 |
% \end{macrocode} |
| 847 |
% \end{macro} |
| 848 |
% \begin{macro}{\@LRT@processNextLine} |
| 849 |
% This code actually processes the line from the |csv|-file. |
| 850 |
% The first three lines are setup. |
| 851 |
% The two |\edef|'s do the following: |\@LRT@restArg{#1}| has the entire current row; and |
| 852 |
% |\@LRT@frestArg| has a comma separated list of field-names. |
| 853 |
% Setting |\booltrue{@LRT@iterator}| means the iteration will start. |
| 854 |
% The two |\expandafter| lines do the iteration: the current values are in |\ffirstArg| for |
| 855 |
% the field-name and |\firstArg| for the entry in the current row. |
| 856 |
% The line |\@LRT@setMacrosForOneEntry| sets the contents of the field-name-macro. |
| 857 |
% Finally the |
| 858 |
% |\if@LRT@empty| line checks if we are done. |
| 859 |
% |
| 860 |
% The |\edef| between the two |\expandafter|'s adds a blank entry to the back of |
| 861 |
% |\@LRT@restArg| so if the row is too short in the |csv|-file it will still scan correctly. |
| 862 |
% |
| 863 |
% The |\@LRT@Input| outputs the result either to the stream or to the |tmp|-file |
| 864 |
% and |\@LRT@justProcessedMessage| writes to the terminal which row was just processed. |
| 865 |
% If something suddenly goes wrong you know which row you were working on |
| 866 |
% when things messed up. |
| 867 |
% \begin{macrocode} |
| 868 |
\gdef\@LRT@processNextLine#1{% |
| 869 |
\edef\@LRT@restArg{#1}% |
| 870 |
\edef\@LRT@frestArg{\@lrt@firstline}% |
| 871 |
\booltrue{@LRT@iterator}% |
| 872 |
\whileboolexpr{bool{@LRT@iterator}}{% |
| 873 |
\expandafter\@LRT@fieldNameIterator\@LRT@frestArg @% |
| 874 |
\edef\@LRT@restArgA{\@LRT@restArg\@LRT@fieldDelimiterBack} |
| 875 |
\expandafter\@LRT@dataLineIterator\@LRT@restArgA @% |
| 876 |
\@LRT@setMacrosForOneEntry{\@LRT@ffirstArg}{\@LRT@firstArg}% |
| 877 |
\if@LRT@empty{\@LRT@frestArg}% |
| 878 |
{\boolfalse{@LRT@iterator}}{\booltrue{@LRT@iterator}}% |
| 879 |
}{}% |
| 880 |
\@LRT@Input\relax% |
| 881 |
\@LRT@justProcessedMessage% |
| 882 |
} |
| 883 |
|
| 884 |
% \end{macrocode} |
| 885 |
% \end{macro} |
| 886 |
% \begin{macro}{\@LRT@mergeData} |
| 887 |
% This macro opens the |csv|-file, reads the first row. |
| 888 |
% The code between the |{{ ... }}| makes a comma separated list of field-names |
| 889 |
% in the order in which they occur in the |csv|-file. It also creates a macro which |
| 890 |
% expands to what a blank line in the |csv|-file will look like. It also defines one |
| 891 |
% boolean for each field-name. |
| 892 |
% This requires an iteration over |\@lrtfirstline|. |
| 893 |
% |
| 894 |
% Because \TeX\ |\newif|'s are not global, we need to define the ones we need as soon |
| 895 |
% as possible, which is right here. |
| 896 |
% |
| 897 |
% Then |\@LRT@mergeData| calls |\@LRT@getAndProcessNextLine| which |
| 898 |
% reads and processes the subsequent lines of the |csv| file. |
| 899 |
% |
| 900 |
% Finally it puts out a no-more-data message to let you know |\mergeFields| is finished. |
| 901 |
% \begin{macrocode} |
| 902 |
\gdef\@LRT@mergeData{% |
| 903 |
\providebool{@LRT@texEOF}% |
| 904 |
\immediate\openin\@lrt@csvFile=\@LRT@csv\relax% |
| 905 |
\providebool{@LRT@csvEOF}% |
| 906 |
\boolfalse{@LRT@csvEOF}% |
| 907 |
\ifeof\@lrt@csvFile\@LRT@failedToOpenMessage{\@LRT@csv}\fi% |
| 908 |
{{\@LRT@catcodes |
| 909 |
\read\@lrt@csvFile to\@lrt@firstlineX% |
| 910 |
\@LRT@setFirstLineAndBlankLine{\@lrt@firstlineX} |
| 911 |
}}% |
| 912 |
\booltrue{@LRT@iterator} |
| 913 |
\edef\@LRT@restArg{\@lrt@firstline}% |
| 914 |
\whileboolexpr{bool{@LRT@iterator}}{% |
| 915 |
\expandafter\@LRT@getCommaFieldDelimited\@LRT@restArg @% |
| 916 |
\providebool{\@LRT@macroPrefix\@LRT@firstArg}% |
| 917 |
\if@LRT@empty{\@LRT@restArg}% |
| 918 |
{\boolfalse{@LRT@iterator}}{\booltrue{@LRT@iterator}}% |
| 919 |
}{}% |
| 920 |
\@LRT@getAndProcessNextLine% |
| 921 |
\@LRT@NoMoreDataMessage% |
| 922 |
} |
| 923 |
|
| 924 |
% \end{macrocode} |
| 925 |
% \end{macro} |
| 926 |
% \Hypertarget{options}{} |
| 927 |
% \begin{macro}{Options code} |
| 928 |
% Here is the code which |
| 929 |
% examines the option string and maps the public macro names to the |
| 930 |
% private codes which actually do the work. |
| 931 |
% \begin{macrocode} |
| 932 |
\gdef\@LRT@weirdOption#1{\@ifundefined{@LRT@CO}% |
| 933 |
{\global\csletcs{#1}{@LRT@#1}}{{\global\csletcs{\@LRT@CO #1}{@LRT@#1}}}% |
| 934 |
} |
| 935 |
|
| 936 |
\@LRT@weirdOption{mergeFields}% |
| 937 |
\@LRT@weirdOption{Field}% |
| 938 |
\@LRT@weirdOption{ifFieldEmpty}% |
| 939 |
\@LRT@weirdOption{setDelimitersCommaQuote}% |
| 940 |
\@LRT@weirdOption{setDelimitersTabQuote}% |
| 941 |
\@LRT@weirdOption{makeMePublic}% |
| 942 |
|
| 943 |
% \end{macrocode} |
| 944 |
% \end{macro} |
| 945 |
% \begin{macro}{Set the default delimiters} |
| 946 |
% Here \Hypertarget{defaults}{is} |
| 947 |
% where the default delimiters are set. |
| 948 |
% \begin{macrocode} |
| 949 |
\@LRT@setDelimitersCommaQuote |
| 950 |
|
| 951 |
% \end{macrocode} |
| 952 |
% \end{macro} |
| 953 |
% \begin{macro}{Finish up} |
| 954 |
% \begin{macrocode} |
| 955 |
\makeatother |
| 956 |
\endinput |
| 957 |
% \end{macrocode} |
| 958 |
% \end{macro} |
| 959 |
% \Finale |
| 960 |
% \newpage |
| 961 |
% \appendix |
| 962 |
% \section{Writing new delimiter macros}\hypertarget{newDelimiterAppendix}{} |
| 963 |
% If the field delimiter is not something \TeX\ sees as white space, life is fairly easy. |
| 964 |
% In your master-file, copy and paste one of the code examples below. |
| 965 |
% Change to suit your needs. |
| 966 |
% |
| 967 |
% \noindent*********************************************************************** |
| 968 |
% |
| 969 |
% \begin{verbatim} |
| 970 |
% {{% |
| 971 |
% \makeatletter |
| 972 |
% \gdef\setMyDelimiters{% |
| 973 |
% \gdef\@LRT@fieldDelimiterBack{,} |
| 974 |
% \gdef\@LRT@getFieldDelimited##1,{\@LRT@getFirstArg{##1}} |
| 975 |
% \gdef\@LRT@textDelimiterFront{"} |
| 976 |
% \gdef\@LRT@textDelimiterBack{",} |
| 977 |
% \gdef\@LRT@getTextDelimited"##1",{\@LRT@getFirstArg{##1}} |
| 978 |
% \gdef\@LRT@blankFieldItem{,} |
| 979 |
% \gdef\@LRT@catcodes{} |
| 980 |
% }% |
| 981 |
% \makeatother |
| 982 |
% }} |
| 983 |
% \end{verbatim} |
| 984 |
% |
| 985 |
% \noindent*********************************************************************** |
| 986 |
% |
| 987 |
% Replace all occurrences of {\tt ,} by your field delimiter. Replace all occurrences of {\tt "} |
| 988 |
% with your text deliminator |
| 989 |
% and you should be ready to go. |
| 990 |
% Just add |\setMyDelimiters| in your master-file before using |\mergeFields|. |
| 991 |
% |
| 992 |
% If the field delimiter is a white space character such as a tab, then copy and paste |
| 993 |
% the next example |
| 994 |
% |
| 995 |
% \noindent*********************************************************************** |
| 996 |
% |
| 997 |
% \begin{verbatim} |
| 998 |
% \makeatletter |
| 999 |
% {{\makeatletter |
| 1000 |
% \catcode`\^^I12\relax% |
| 1001 |
% \gdef\@LRT@setDelimitersTabQuote{ |
| 1002 |
% \gdef\@LRT@fieldDelimiterBack{^^I} |
| 1003 |
% \gdef\@LRT@getFieldDelimited##1^^I{\@LRT@getFirstArg{##1}} |
| 1004 |
% \gdef\@LRT@textDelimiterFront{"} |
| 1005 |
% \gdef\@LRT@textDelimiterBack{"^^I} |
| 1006 |
% \gdef\@LRT@getTextDelimited"##1"^^I{\@LRT@getFirstArg{##1}} |
| 1007 |
% \gdef\@LRT@blankFieldItem{} |
| 1008 |
% \gdef\@LRT@catcodes{\catcode`\^^I12\relax} |
| 1009 |
% }% |
| 1010 |
% }} |
| 1011 |
% \makeatother |
| 1012 |
% }} |
| 1013 |
% \end{verbatim} |
| 1014 |
% |
| 1015 |
% \noindent*********************************************************************** |
| 1016 |
% |
| 1017 |
% Use Chapter 8 of the \TeX\ book to find how to write your delimiter and replace |
| 1018 |
% all occurrences of |\^^I| with the code for your delimiter. |
| 1019 |
% Then replace all occurrences of {\tt "} with your text deliminator |
| 1020 |
% and hope for the best. |
| 1021 |
% Try leaving |\@LRT@blankFieldItem| blank and hope for the best. |
| 1022 |
% Again add |\setMyDelimiters| in your master-file before using |\mergeFields|. |
| 1023 |
% |
| 1024 |
% \section{Some examples} |
| 1025 |
% \subsection{\TeX\ code in a table} |
| 1026 |
% A typical line in a tabular environment looks like |item1 & item2& item3\\|. |
| 1027 |
% Suppose you have 4 columns in your |csv|-file with entries |Last Name|, |
| 1028 |
% |First name|, |Numerator| and |Denominator| and you want |item1| to be |
| 1029 |
% |Last Name|, |item2| to be |First name| and |item3| to be |\frac{Numerator}{Denominator}|. |
| 1030 |
% Here is the needed code in your |tex|-file: |
| 1031 |
% |
| 1032 |
% \begin{verbatim} |
| 1033 |
% \Field{lastname}&\Field{firstname}&% |
| 1034 |
% $\noexpand\frac{\Field{numerator}}{\Field{denominator}}$\noexpand\\ |
| 1035 |
% \end{verbatim} |
| 1036 |
% You might have expected to see a |\noexpand| in front of the \$'s but remember \$ is one of |
| 1037 |
% those characters which goes through unscathed to the |tmp|-file. |
| 1038 |
% It certainly doesn't hurt to write |\noexpand$|. When you read the |tmp|-file back in |
| 1039 |
% you do want |$| to resume its meaning of making math mode and it does. |
| 1040 |
% |
| 1041 |
% The |\noexpand| in front of the |\frac| and the |\\| are needed. |
| 1042 |
% You don't want to try to expand either one |
| 1043 |
% until the |tmp|-file is read back in. |
| 1044 |
% |
| 1045 |
% Assuming that the file |merge.tex| contains the code above and that the file |data.csv| contains your data, |
| 1046 |
% the following code typesets the table. |
| 1047 |
% \begin{verbatim} |
| 1048 |
% \documentclass[12pt]{article} |
| 1049 |
% \RequirePackage{csvmerge} |
| 1050 |
% \begin{document} |
| 1051 |
% \mergeFields{ |
| 1052 |
% tmp={example.tmp} |
| 1053 |
% tex ={merge.tex} |
| 1054 |
% csv = {data.csv} |
| 1055 |
% } |
| 1056 |
% \begin{tabular}{lll} |
| 1057 |
% \input{example.tmp} |
| 1058 |
% \end{tabular} |
| 1059 |
% \end{document} |
| 1060 |
% \end{verbatim} |
| 1061 |
% The |csv|-file needs to contain the four field-names that appear in the |tex|-file, |
| 1062 |
% but it is fine if it contains additional columns. |
| 1063 |
% |
| 1064 |
% \subsection{\TeX\ code in a mail-merge} |
| 1065 |
% Here is a snippet of code to do a header in a mail-merge business letter. |
| 1066 |
% The field-names from the |csv|-file should be obvious. The interesting point |
| 1067 |
% is the usage of |\ifFieldEmpty|. |
| 1068 |
% \begin{verbatim} |
| 1069 |
% \Field{firstname} \ifFieldEmpty{middleinitial}{} |
| 1070 |
% {\Field{middleinitial}.~}\Field{lastname}\par |
| 1071 |
% \ifFieldEmpty{streetline1}{}{\Field{streetline1}\par} |
| 1072 |
% \ifFieldEmpty{streetline2}{}{\Field{streetline2}\par} |
| 1073 |
% \ifFieldEmpty{streetline3}{}{\Field{streetline3}\par} |
| 1074 |
% \Field{city}, \Field{stateprovince} \Field{postalcode} |
| 1075 |
% \end{verbatim} |
| 1076 |
% |
| 1077 |
% The |middleinitial| column is a middle initial with no period. Some recipients will have no middle initial |
| 1078 |
% but you don't want just a floating period. Also you don't want an additional space. |
| 1079 |
% Knuth also suggests that names should be \TeX ed with |First M.~Last| [The \TeX\ book, Ch. 6]. |
| 1080 |
% The |\ifFieldEmpty| with |streetline1| is probably redundant but often |streetline2| will be empty. |
| 1081 |
% If it is not, we want a new line after it, but if it is empty we don't want a blank new line. |
| 1082 |
% |
| 1083 |
% In storage mode you need a |\noexpand| before the |~|, as well as before the |\par|'s. |
| 1084 |
% \subsection{Some master-files} |
| 1085 |
% \subsubsection{A table} |
| 1086 |
% We are doing a table so we want storage mode. |
| 1087 |
% We also want a horizontal line after each row of the table. |
| 1088 |
% The master-file: |
| 1089 |
% \begin{verbatim} |
| 1090 |
% \documentclass[12pt]{amsart} |
| 1091 |
% \RequirePackage{csvmerge} |
| 1092 |
% \begin{document} |
| 1093 |
% \mergeFields{ |
| 1094 |
% tmp= { ./table.tmp} |
| 1095 |
% csv = {./data.csv} |
| 1096 |
% tex ={./table.tex } |
| 1097 |
% } |
| 1098 |
% \begin{tabular}{|l|l|l|l|l|} |
| 1099 |
% \hline% |
| 1100 |
% \input{./table.tmp} |
| 1101 |
% \end{tabular} |
| 1102 |
% \end{document} |
| 1103 |
% \end{verbatim} |
| 1104 |
% |
| 1105 |
% The |tex|-file |table.tmp|: |
| 1106 |
% \begin{verbatim} |
| 1107 |
% % one table line |
| 1108 |
% \Field{firstname} &\Field{id} & \Field{city} &\Field{lastname}& |
| 1109 |
% \Field{postalcode} \noexpand\\ |
| 1110 |
% \noexpand\hline |
| 1111 |
% \end{verbatim} |
| 1112 |
% |
| 1113 |
% \subsubsection{A mail-merge} |
| 1114 |
% We are doing a mail-merge so we can use streaming mode. |
| 1115 |
% Since the |tex|-file name has spaces, we have quoted it. |
| 1116 |
% The master-file: |
| 1117 |
% \begin{verbatim} |
| 1118 |
% \documentclass[12pt]{letter} |
| 1119 |
% \RequirePackage{./csvmerge} |
| 1120 |
% \begin{document} |
| 1121 |
% \mergeFields{ tex = {"./Second letter.template" } |
| 1122 |
% csv ={ ./Invites.csv}} |
| 1123 |
% \end{document} |
| 1124 |
% \end{verbatim} |
| 1125 |
% |
| 1126 |
% The |tex|-file |Second letter.template|: |
| 1127 |
% |
| 1128 |
% \begin{verbatim} |
| 1129 |
% %First page of letter |
| 1130 |
% \setcounter{page}{1} |
| 1131 |
% |
| 1132 |
% \Field{firstname} \ifFieldEmpty{middleinitial}{}% |
| 1133 |
% {\Field{middleinitial}.~}\Field{lastname}\par |
| 1134 |
% \ifFieldEmpty{streetline1}{}{\Field{streetline1} \par} |
| 1135 |
% \ifFieldEmpty{streetline2}{}{\Field{streetline2}\par} |
| 1136 |
% \ifFieldEmpty{streetline3}{}{\Field{streetline3}\par} |
| 1137 |
% \Field{city}, \Field{stateprovince} \Field{postalcode} |
| 1138 |
% |
| 1139 |
% \vskip.25in |
| 1140 |
% Dear \Field{preferredfirstname},\par |
| 1141 |
% \vskip10pt |
| 1142 |
% How come no RSVP? |
| 1143 |
% \vskip10pt |
| 1144 |
% Bilbo Baggins |
| 1145 |
% \newpage |
| 1146 |
% \end{verbatim} |
| 1147 |
% |
| 1148 |
% Here is an equivalent version using storage mode. |
| 1149 |
% The master-file: |
| 1150 |
% \begin{verbatim} |
| 1151 |
% \documentclass[12pt]{letter} |
| 1152 |
% \RequirePackage{./csvmerge} |
| 1153 |
% \begin{document} |
| 1154 |
% \mergeFields{ tex = {"./Second letter.template" } |
| 1155 |
% csv ={ ./Invites.csv}tmp={letters}} |
| 1156 |
% \input{letters} |
| 1157 |
% \end{document} |
| 1158 |
% \end{verbatim} |
| 1159 |
% |
| 1160 |
% We must adjust the |tex|-file |Second letter.template|: |
| 1161 |
% |
| 1162 |
% \begin{verbatim} |
| 1163 |
% %First page of letter |
| 1164 |
% \noexpand\setcounter{page}{1} |
| 1165 |
% |
| 1166 |
% \Field{firstname} \ifFieldEmpty{middleinitial}{}% |
| 1167 |
% {\Field{middleinitial}.\noexpand~}\Field{lastname}\noexpand\par |
| 1168 |
% \ifFieldEmpty{streetline1}{}{\Field{streetline1} \noexpand\par} |
| 1169 |
% \ifFieldEmpty{streetline2}{}{\Field{streetline2}\noexpand\par} |
| 1170 |
% \ifFieldEmpty{streetline3}{}{\Field{streetline3}\noexpand\par} |
| 1171 |
% \Field{city}, \Field{stateprovince} \Field{postalcode} |
| 1172 |
% |
| 1173 |
% \noexpand\vskip.25in |
| 1174 |
% Dear \Field{preferredfirstname},\par |
| 1175 |
% \noexpand\vskip10pt |
| 1176 |
% How come no RSVP? |
| 1177 |
% \noexpand\vskip10pt |
| 1178 |
% Bilbo Baggins |
| 1179 |
% \noexpand\newpage |
| 1180 |
% \end{verbatim} |
| 1181 |
% |
| 1182 |
% Usually using storage mode when you don't need it is wasteful since you have to create and |
| 1183 |
% save a |tmp|-file. It can come in handy if your letter code has lots of field macros and |
| 1184 |
% you can't get one |tex|-file to not have overfull/underfull boxes and the like after the |
| 1185 |
% various expansions. The |tmp|-file contains all the letters and you can fine tune each one and |
| 1186 |
% then just do |
| 1187 |
% \begin{verbatim} |
| 1188 |
% \documentclass[12pt]{letter} |
| 1189 |
% \begin{document} |
| 1190 |
% \input{letters} |
| 1191 |
% \end{document} |
| 1192 |
% \end{verbatim} |
| 1193 |
|
| 1194 |
|
| 1195 |
|
| 1196 |
|
| 1197 |
|
| 1198 |
|