<?php if (!defined('PmWiki')) exit();
# vim: set ts=4 sw=4 et:
##
## File: text2tbl.php
## Version: 2010-04-06
## SVN ID: $Id$
## Status: alpha
## Author: Peter Bowers
## Create Date: August 25, 2008
## Copyright: 2008-2010, Peter Bowers
##
## This program is free software; you can redistribute it and/or modify
## it under the terms of the GNU General Public License, Version 2, as
## published by the Free Software Foundation.
## http://www.gnu.org/copyleft/gpl.html
## This program is distributed in the hope that it will be useful,
## but WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
## GNU General Public License for more details.
##
## text2tbl.php provides a couple $ROEPatterns to convert text to simple table
## markup.
$RecipeInfo['Text2Tbl']['Version'] = '2013-09-23';
define(text2tbl, true);
SDV($t2tKeepOriginal, 1);
SDV($t2tEnableSuppressDefaultAlign, false);
# This pattern lets you define a table like this:
# TABLESTART border=1
# abc def ghi
# jkl asdfasfd asdf
# qwer sss sssssss
# TABLEEND
# You don't have to make things line up -- just make sure that there are no
# occurrences of multiple consecutive spaces unless you want them to become
# the dividers between columns.
$ROEPatterns["/TABLESTART([^\n]*)\n(.*?)\n\s*TABLEEND/se"] = "'|| '.PSS('$1').'\n'.PSS(preg_replace(array('/ {2,}/', '/^/m', '/$/m'), array(' ||', '||', ' ||'), PSS('$2')))";
# This pattern is significantly more complicated. It allows this type of
# table definition:
# (:text2tbl OPTIONS:)
# col1 SEP col2 SEP col3
# col1 SEP col2 SEP col3
# col1 SEP col2 SEP col3
# (:text2tbl:)
# OPTIONS are specified in the usual opt=val method and can include the
# following:
# re=1 (treat the separater as a regular expression pattern)
# sep=X (the separator for all columns will be X (can be multiple chars)
# sep1=X sep2=Y sep3=Z (specify potentially different separators for each
# column)
# align=lrc (specify the alignment of each column as left, right, or center
# the number of chars will equal the number of columns: llc, lrllc,etc)
# fmt=AXAXAX (specify the alignment (A) and the separator (X) of each col
# in a single option. Separators must be single-character.
# options="border=1" (specify table options - whatever is in the quotes will
# be put verbatim in the usual place on the 1st line)
# quotes=1 (specify that you want double-quoted strings to be hidden from
# possibly containing separators -- good for CSV and etc.)
# keepquotes=1 (specify that the quotes should stay in the text rather than
# being stripped out as would normally happen if quotes=1)
# Note that specifying sep1,sep2,sep3 is mutually exclusive with specifying
# sep and is mutually exclusive with specifying fmt. You can specify your
# separator in exactly one way -- if you choose more than one way then something
# will be ignored...
$ROEPatterns["/(\\(:text2ta?ble?([^:\n]*):\\))\n(.*?\n)(\s*\\(:text2ta?ble?end:\\))/se"] = "t2tText2Tbl(PSS('$2'), PSS('$3'), PSS('$1'), PSS('$4'))";
function t2tText2Tbl($args, $tabledef, $text2tbl_begin, $text2tbl_end)
{
global $KeepToken, $KPV, $t2tKeepOriginal, $t2tEnableSuppressDefaultAlign;
$opt = ParseArgs($args);
SDV($opt['keeporiginal'], $t2tKeepOriginal);
#echo "Text2Tbl(): Entering<br>\n";
# is it a regex?
$re = @$opt['re'];
# $sep will either be a string (same for all) or an array of strings
# (each column can be different)
if (isset($opt['sep1']))
for ($i=1; isset($opt["sep${i}"]); $i++)
$sep[$i-1] = $opt["sep${i}"];
elseif (isset($opt['sep'])) $sep = $opt['sep'];
elseif (isset($opt['fmt'])) { // align-sep-align-sep-align
$align = substr($opt['fmt'], 0, 1);
for ($i=1; $i<strlen($opt['fmt']); $i+=2) {
$sep[($i-1)/2] = substr($opt['fmt'], $i, 1);
$align .= substr($opt['fmt'], $i+1, 1);
}
}
elseif ($opt['input'] == 'spaces' || $opt['input'] == 'space') {
$sep = ' {2,}';
$re = true;
} elseif ($opt['input'] == 'csv') {
$sep = ',';
$re = false;
$opt['quotes'] = true;
} elseif ($opt['input'] == 'tsv') {
$sep = '\t';
$re = true;
} else $sep = ($re?' +':','); // default to multi-spaces if re, comma non-re
if (preg_match('/^(.*?)\(:end\-of\-original\-data[^:]*:\)/s', $tabledef, $m))
$tabledef = $m[1];
$lines = preg_split("/(?<!\\\\)\n/", $tabledef);
# Determine how many columns (in the 1st line if not specified by $sep)
if (is_array($sep)) $colcnt = sizeof($sep);
else {
if ($opt['quotes'])
$x = preg_replace('/"([^"]*)"/e', "Keep(PSS('$1'),'P')", $lines[0]);
else $x = $lines[0];
if ($re) $colcnt = preg_match_all("/$sep/", $x, $junk);
else $colcnt = substr_count($x, "$sep");
}
# Make sure we have a valid $align ($colcnt occurrences of l or r or c)
if (!@$align) {
$align = '';
if ($opt['align']) {
if (strlen($opt['align']) == $colcnt+1)
$align = $opt['align'];
elseif (strlen($opt['align']) == 1)
$align = str_repeat($opt['align'],$colcnt+1);
elseif (strlen($opt['align']) < $colcnt+1)
$align = $opt['align'] . str_repeat('l', $colcnt+1-strlen($opt['align']));
else
$align = substr($align, 0, $colcnt+1);
}
if (!$align) $align = str_repeat('l', ${colcnt}+1);
}
$align = strtolower($align);
# If it's not a regex then preg_quote everything
if (!$re) {
if (is_array($sep))
foreach ($sep as &$s)
$s = preg_quote($s);
else $sep = preg_quote($sep);
}
unset($s); // don't want the reference anymore - tuf debug!
$ValidOpts = array('#'=>1, 'sep'=>1, 'align'=>1, 'fmt'=>1, 're'=>1,
'options'=>1, 'align'=>1, 'keepquotes'=>1, 'input'=>1, 'output'=>1,
'keeporiginal'=>1);
for ($i=1; $i<25; $i++) $ValidOpts['sep'.$i] = 1;
$newopt = array_diff_key($opt, $ValidOpts);
if ($newopt)
foreach ($newopt as $k=>$v)
$opt['options'] .= ($opt['options']?" ":"") . "$k=$v";
##
## Read from simple table
##
if ($opt['input'] == 'simple') {
if (preg_match("/^\s*\|\|\s*([^\|]*)$/", $lines[0], $m)) {
array_shift($lines);
$opt['options'] = ($opt['options']?$opt['options'].' ':'').$m[1];
}
#echo "lines=<pre>".print_r($lines,true)."</pre><br>\n";
foreach ($lines as $k => &$line) {
$line = preg_replace("/^\s*\|\|(.*?)(?:\|\|)?\s*$/s", "$1", $line);
$cells[$k] = explode("||", $line);
# Now check alignment and store override in an array...
$lastcol=0;
for ($i = 0; $i < sizeof($cells[$k]); $i++) {
if ($x = t2tOptions($cells[$k][$i]))
$options[$k][$col] = $x;
if (!$cells[$k][$i] || preg_match("/^_{2,}$/", $cells[$k][$i])) {
if ($colspan[$k][$lastcol])
$colspan[$k][$lastcol]++;
else
$colspan[$k][$lastcol] = 2;
$cells[$k][$i] = '';
continue;
}
if (preg_match("/(\+{2,})(\s*)$/", $cells[$k][$i], $m)) {
# Get rid of the +++
$cells[$k][$i] = preg_replace("/\+{2,}(\s*)$/", '$1', $cells[$k][$i]);
$rowspan[$k][$i] = strlen($m[1]);
}
if (preg_match("/^\^{2,}$/", $cells[$k][$i])) {
$cells[$k][$i] = '';
continue;
}
if (substr($cells[$k][$i], -1, 1) == ' ') {
if (substr($cells[$k][$i], 0, 1) == ' ')
$a = 'c';
else
$a = 'l';
} elseif (substr($cells[$k][$i], 0, 1) == ' ')
$a = 'r';
else
$a = 'l';
if ($a != $align[$i]) {
#echo "Overriding $k-$i: :".str_replace(" ", "X", $cells[$k][$i]).": $a<br>\n";
$align_override[$k][$i] = $a;
}
$lastcol = $i;
}
if ($i > $colcnt) $colcnt = $i;
}
$colcnt--; // some weird offset I've got going
} else {
##
## Read from (:text2tbl ...:) "markup"
##
if ($opt['quotes']) {
$rpat = "/$KeepToken(\\d+P)$KeepToken/e";
$rrep = '$KPV[\'$1\']';
}
if (!is_array($sep)) $s = $sep;
foreach ($lines as $k => &$line) {
if (!$line) continue;
if ($opt['quotes'])
$line = preg_replace('/"([^"]*)"/e', "Keep(PSS('$1'), 'P')", $line);
for ($col=0; $col<=$colcnt; $col++) {
if (is_array($sep))
$s = $sep[$col];
if ($col == $colcnt) {
if ($x = t2tOptions($line))
$options[$k][$col] = $x;
$cells[$k][$col] = $line;
} elseif (preg_match("/^((.*?)".$s.")/", $line, $m)) {
if ($x = t2tOptions($m[2]))
$options[$k][$col] = $x;
$line = substr($line, strlen($m[1]));
$cells[$k][$col] = $m[2];
}
# Note: we can't give this message because a rowspan could
# result in a column "missing" which is totally valid
#else echo "OOPS! No match... sep=$s, line=$line<br>\n";
}
}
}
#echo "cnt=$colcnt, cells=<pre>".print_r($cells,true)."</pre><br>\n";
#echo "align_override=<pre>".print_r($align_override,true)."</pre><br>\n";
$lines = array();
if ($opt['output'] == 'advanced') {
$aligntbl=array('l'=>' align=left', 'r'=>' align=right', 'c'=>' align=center');
if ($t2tEnableSuppressDefaultAlign) $aligntbl['l'] = '';
$lines[] = "(:table $opt[options]:)";
$newtbl='';
for ($j=0; $j<sizeof($cells); $j++) {
for ($col=0; $col<=$colcnt; $col++) {
if (!$cells[$j][$col]) continue;
$a = (isset($align_override[$j][$col]) ? $align_override[$j][$col] : strtolower(substr($align, $col, 1)));
#echo "a=$a ($aligntbl[$a])<br>\n";
$line = "(:cell" . ($col==0?'nr':'') . $aligntbl[$a];
if (substr($cells[$j][$col], 0, 1) == '!') {
$cells[$j][$col] = substr($cells[$j][$col], 1);
$line .= " style='text-align:center; font-weight:bold;'";
}
if ($options[$j][$col])
$line .= ' '.$options[$j][$col];
if ($opt['attr'.($col+1)])
$line .= ' '.$opt['attr'.($col+1)];
if ($opt['attr'])
$line .= ' '.$opt['attr'];
if ($colspan[$j][$col])
$line .= " colspan=".$colspan[$j][$col];
if ($rowspan[$j][$col])
$line .= " rowspan=".$rowspan[$j][$col];
$line .= ":)";
$cell = trim($cells[$j][$col]);
if ($opt['quotes']) {
if ($opt['keepquotes'])
$cell = preg_replace($rpat, '\'"\'.'.$rrep.'.\'"\'', $cell);
else
$cell = preg_replace($rpat, $rrep, $cell);
}
$lines[] = $line.$cell;
}
}
$lines[] = "(:tableend:)";
} else {
$newtbl = "|| $opt[options]\n";
for ($j=0; $j<sizeof($cells); $j++) {
for ($col=0; $col<=$colcnt+1; $col++) {
if ($col == 0 || strtolower(substr($align, $col-1, 1)) == 'r')
$left = '';
else $left = ' ';
if ($col < $colcnt+1 && strtolower($align{$col}) == 'l')
$right = '';
else $right = ' ';
if ($col == 0) {
#echo "1: ".substr($cells[$j][$col], 0, 1). "<br>\n";
$line = "||${right}".$cells[$j][$col];
if (substr(ltrim($cells[$j][$col]), 0, 1) == '!')
$line = "||!${right}".substr($cells[$j][$col], 1);
else
$line = "||${right}".$cells[$j][$col];
} elseif ($col == $colcnt+1) {
#echo "2: ".substr($cells[$j][$col], 0, 1). "<br>\n";
$line = "${line}${left}||";
} else {
#echo "3: ".substr($cells[$j][$col], 0, 1). "<br>\n";
if (substr(ltrim($cells[$j][$col]), 0, 1) == '!')
$line = "${line}${left}||!${right}".
substr($cells[$j][$col], 1);
else
$line = "${line}${left}||${right}".$cells[$j][$col];
}
}
if ($opt['quotes']) {
if ($opt['keepquotes'])
$line = preg_replace($rpat, '\'"\'.'.$rrep.'.\'"\'', $line);
else
$line = preg_replace($rpat, $rrep, $line);
}
$lines[] = $line;
}
}
if ($opt['keeporiginal']) {
$newtbl = $text2tbl_begin ."\n". $tabledef . "(:end-of-original-data (THE TABLE BELOW WILL BE OVER-WRITTEN EVERY TIME THIS PAGE IS SAVED!) Specify keeporiginal=0 if you want to get rid of the text above.:)\n". $newtbl . implode("\n",$lines) ."\n". $text2tbl_end;
} else
$newtbl .= implode("\n",$lines);
return($newtbl);
}
function t2tOptions(&$string)
{
if (preg_match("/^\s*\(:t2t\s([^:]*):\)(.*)$/", $string, $m)) {
$opt = $m[1];
$string = $m[2]; // as a side effect, strip this (:t2t ...:) markup
return($opt);
} else
return('');
}
# If keep=1 was specified then the original data is maintained in the source,
# but it must not be displayed. This markup makes the original data (and the
# text2tbl markup) invisible.
Markup('t2t-comment', '<{$var}',
"/\(:text2ta?ble?[^\n]*:\).*?\(:end-of-original-data[^:]*:\)\n(.*?)\n\(:text2ta?ble?end:\)/s",
PSS('$1'));
# This will implement a "real" markup which will convert any of the usual
# types of structured text (as implemented in Text2Tbl() above) but will not
# alter the source
Markup('flextable', '<t2t-comment',
"/(\\(:flexta?ble?([^:\n]*):\\))\n(.*?\n)(\s*\\(:flexta?ble?end:\\))/se",
"FlexTbl(PSS('$2'), PSS('$3'), PSS('$1'), PSS('$4'))");
function FlexTbl($args, $tabledef, $text2tbl_begin, $text2tbl_end)
{
if (!strstr($args, 'keeporiginal')) $args .= " keeporiginal=0";
return(t2tText2Tbl($args, $tabledef, $text2tbl_begin, $text2tbl_end));
}