1<?php
2namespace TYPO3\CMS\Core\Utility;
3
4/*
5 * This file is part of the TYPO3 CMS project.
6 *
7 * It is free software; you can redistribute it and/or modify it under
8 * the terms of the GNU General Public License, either version 2
9 * of the License, or any later version.
10 *
11 * For the full copyright and license information, please read the
12 * LICENSE.txt file that was distributed with this source code.
13 *
14 * The TYPO3 project - inspiring people to share!
15 */
16
17use TYPO3\CMS\Core\IO\CsvStreamFilter;
18
19/**
20 * Class with helper functions for CSV handling
21 */
22class CsvUtility
23{
24    /**
25     * whether to passthrough data as is, without any modification
26     */
27    public const TYPE_PASSTHROUGH = 0;
28
29    /**
30     * whether to remove control characters like `=`, `+`, ...
31     */
32    public const TYPE_REMOVE_CONTROLS = 1;
33
34    /**
35     * whether to prefix control characters like `=`, `+`, ...
36     * to become `'=`, `'+`, ...
37     */
38    public const TYPE_PREFIX_CONTROLS = 2;
39
40    /**
41     * Convert a string, formatted as CSV, into an multidimensional array
42     *
43     * This cannot be done by str_getcsv, since it's impossible to handle enclosed cells with a line feed in it
44     *
45     * @param string $input The CSV input
46     * @param string $fieldDelimiter The field delimiter
47     * @param string $fieldEnclosure The field enclosure
48     * @param int $maximumColumns The maximum amount of columns
49     * @return array
50     */
51    public static function csvToArray($input, $fieldDelimiter = ',', $fieldEnclosure = '"', $maximumColumns = 0)
52    {
53        $multiArray = [];
54        $maximumCellCount = 0;
55
56        if (($handle = fopen('php://memory', 'r+')) !== false) {
57            fwrite($handle, $input);
58            rewind($handle);
59            while (($cells = fgetcsv($handle, 0, $fieldDelimiter, $fieldEnclosure)) !== false) {
60                $maximumCellCount = max(count($cells), $maximumCellCount);
61                $multiArray[] = preg_replace('|<br */?>|i', LF, $cells);
62            }
63            fclose($handle);
64        }
65
66        if ($maximumColumns > $maximumCellCount) {
67            $maximumCellCount = $maximumColumns;
68        }
69
70        foreach ($multiArray as &$row) {
71            for ($key = 0; $key < $maximumCellCount; $key++) {
72                if (
73                    $maximumColumns > 0
74                    && $maximumColumns < $maximumCellCount
75                    && $key >= $maximumColumns
76                ) {
77                    if (isset($row[$key])) {
78                        unset($row[$key]);
79                    }
80                } elseif (!isset($row[$key])) {
81                    $row[$key] = '';
82                }
83            }
84        }
85
86        return $multiArray;
87    }
88
89    /**
90     * Takes a row and returns a CSV string of the values with $delim (default is ,) and $quote (default is ") as separator chars.
91     *
92     * @param string[] $row Input array of values
93     * @param string $delim Delimited, default is comma
94     * @param string $quote Quote-character to wrap around the values.
95     * @param int $type Output behaviour concerning potentially harmful control literals
96     * @return string A single line of CSV
97     */
98    public static function csvValues(array $row, string $delim = ',', string $quote = '"', int $type = self::TYPE_REMOVE_CONTROLS)
99    {
100        $resource = fopen('php://temp', 'w');
101        if (!is_resource($resource)) {
102            throw new \RuntimeException('Cannot open temporary data stream for writing', 1625556521);
103        }
104        $modifier = CsvStreamFilter::applyStreamFilter($resource, false);
105        array_map([self::class, 'assertCellValueType'], $row);
106        if ($type === self::TYPE_REMOVE_CONTROLS) {
107            $row = array_map([self::class, 'removeControlLiterals'], $row);
108        } elseif ($type === self::TYPE_PREFIX_CONTROLS) {
109            $row = array_map([self::class, 'prefixControlLiterals'], $row);
110        }
111        fputcsv($resource, $modifier($row), $delim, $quote);
112        fseek($resource, 0);
113        $content = stream_get_contents($resource);
114        return $content;
115    }
116
117    /**
118     * Prefixes control literals at the beginning of a cell value with a single quote
119     * (e.g. `=+value` --> `'=+value`)
120     *
121     * @param mixed $cellValue
122     * @return bool|int|float|string|null
123     */
124    protected static function prefixControlLiterals($cellValue)
125    {
126        if (!self::shallFilterValue($cellValue)) {
127            return $cellValue;
128        }
129        $cellValue = (string)$cellValue;
130        return preg_replace('#^([\t\v=+*%/@-])#', '\'${1}', $cellValue);
131    }
132
133    /**
134     * Removes control literals from the beginning of a cell value
135     * (e.g. `=+value` --> `value`)
136     *
137     * @param mixed $cellValue
138     * @return bool|int|float|string|null
139     */
140    protected static function removeControlLiterals($cellValue)
141    {
142        if (!self::shallFilterValue($cellValue)) {
143            return $cellValue;
144        }
145        $cellValue = (string)$cellValue;
146        return preg_replace('#^([\t\v=+*%/@-]+)+#', '', $cellValue);
147    }
148
149    /**
150     * Asserts scalar or null types for given cell value.
151     *
152     * @param mixed $cellValue
153     */
154    protected static function assertCellValueType($cellValue): void
155    {
156        // int, float, string, bool, null
157        if ($cellValue === null || is_scalar($cellValue)) {
158            return;
159        }
160        throw new \RuntimeException(
161            sprintf('Unexpected type %s for cell value', gettype($cellValue)),
162            1625562833
163        );
164    }
165
166    /**
167     * Whether cell value shall be filtered, applies to everything
168     * that is not or cannot be represented as boolean, integer or float.
169     *
170     * @param mixed $cellValue
171     * @return bool
172     */
173    protected static function shallFilterValue($cellValue): bool
174    {
175        return $cellValue !== null
176            && !is_bool($cellValue)
177            && !is_numeric($cellValue)
178            && !MathUtility::canBeInterpretedAsInteger($cellValue)
179            && !MathUtility::canBeInterpretedAsFloat($cellValue);
180    }
181}
182