Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
93.65% |
59 / 63 |
|
75.00% |
6 / 8 |
CRAP | |
0.00% |
0 / 1 |
| EncodedWord | |
93.65% |
59 / 63 |
|
75.00% |
6 / 8 |
28.20 | |
0.00% |
0 / 1 |
| decode | |
100.00% |
7 / 7 |
|
100.00% |
1 / 1 |
2 | |||
| findRanges | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
3 | |||
| encode | |
100.00% |
9 / 9 |
|
100.00% |
1 / 1 |
3 | |||
| decodeRun | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
2 | |||
| decodeWord | |
100.00% |
5 / 5 |
|
100.00% |
1 / 1 |
3 | |||
| convertToUtf8 | |
85.71% |
6 / 7 |
|
0.00% |
0 / 1 |
5.07 | |||
| needsEncoding | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| splitIntoChunks | |
87.50% |
21 / 24 |
|
0.00% |
0 / 1 |
9.16 | |||
| 1 | <?php |
| 2 | declare(strict_types=1); |
| 3 | /** |
| 4 | * Pop PHP Framework (https://www.popphp.org/) |
| 5 | * |
| 6 | * @link https://github.com/popphp/popphp-framework |
| 7 | * @author Nick Sagona, III <dev@noladev.com> |
| 8 | * @copyright Copyright (c) 2009-2027 NOLA Interactive, LLC. |
| 9 | * @license https://www.popphp.org/license New BSD License |
| 10 | */ |
| 11 | |
| 12 | /** |
| 13 | * @namespace |
| 14 | */ |
| 15 | namespace Pop\Mime\Part\Header; |
| 16 | |
| 17 | /** |
| 18 | * RFC 2047 encoded-word encoder/decoder - has no dependency on ext-imap. |
| 19 | * Decoding handles both B (base64) and Q (quoted-printable-style) |
| 20 | * encoded-words, since real-world mail uses both; encoding only ever |
| 21 | * produces B, which is simpler and unconditionally correct for any |
| 22 | * content shape. |
| 23 | * |
| 24 | * @category Pop |
| 25 | * @package Pop\Mime |
| 26 | * @author Nick Sagona, III <dev@noladev.com> |
| 27 | * @copyright Copyright (c) 2009-2027 NOLA Interactive, LLC. |
| 28 | * @license https://www.popphp.org/license New BSD License |
| 29 | * @version 3.0.0 |
| 30 | */ |
| 31 | final class EncodedWord |
| 32 | { |
| 33 | |
| 34 | /** |
| 35 | * Decode all RFC 2047 encoded-words found in a string. Never throws - |
| 36 | * malformed or unrecognized encoded-word-shaped text is left as literal |
| 37 | * output. |
| 38 | * |
| 39 | * @param string $text |
| 40 | * @return string |
| 41 | */ |
| 42 | public static function decode(string $text): string |
| 43 | { |
| 44 | if (!str_contains($text, '=?')) { |
| 45 | return $text; |
| 46 | } |
| 47 | |
| 48 | $pattern = '/=\?[^?\s]+\?[BbQq]\?[^?]*\?=(?:\s+=\?[^?\s]+\?[BbQq]\?[^?]*\?=)*/'; |
| 49 | |
| 50 | $result = preg_replace_callback($pattern, function ($matches) { |
| 51 | return self::decodeRun($matches[0]); |
| 52 | }, $text); |
| 53 | |
| 54 | return $result ?? $text; |
| 55 | } |
| 56 | |
| 57 | /** |
| 58 | * Find the byte ranges of every individual RFC 2047 encoded-word in a |
| 59 | * string - used by Header::fold() to treat an encoded-word as an |
| 60 | * atomic, unbreakable unit when choosing fold points. Folding between |
| 61 | * two adjacent encoded-words, at their separating whitespace, is still |
| 62 | * fine - only folding INSIDE one is not. |
| 63 | * |
| 64 | * @param string $text |
| 65 | * @return array |
| 66 | */ |
| 67 | public static function findRanges(string $text): array |
| 68 | { |
| 69 | $ranges = []; |
| 70 | if (preg_match_all('/=\?[^?\s]+\?[BbQq]\?[^?]*\?=/', $text, $matches, PREG_OFFSET_CAPTURE)) { |
| 71 | foreach ($matches[0] as $match) { |
| 72 | $ranges[] = ['start' => $match[1], 'end' => $match[1] + strlen($match[0])]; |
| 73 | } |
| 74 | } |
| 75 | return $ranges; |
| 76 | } |
| 77 | |
| 78 | /** |
| 79 | * Encode a string into one or more RFC 2047 encoded-words, if it |
| 80 | * contains anything outside printable US-ASCII. Returns the input |
| 81 | * unchanged otherwise. |
| 82 | * |
| 83 | * @param string $text |
| 84 | * @param string $charset |
| 85 | * @return string |
| 86 | */ |
| 87 | public static function encode(string $text, string $charset = 'UTF-8'): string |
| 88 | { |
| 89 | if (!self::needsEncoding($text)) { |
| 90 | return $text; |
| 91 | } |
| 92 | |
| 93 | $prefix = '=?' . $charset . '?B?'; |
| 94 | $suffix = '?='; |
| 95 | $maxEncodedLength = 75 - strlen($prefix) - strlen($suffix); |
| 96 | |
| 97 | $words = []; |
| 98 | foreach (self::splitIntoChunks($text, $maxEncodedLength) as $chunk) { |
| 99 | $words[] = $prefix . base64_encode($chunk) . $suffix; |
| 100 | } |
| 101 | |
| 102 | return implode(' ', $words); |
| 103 | } |
| 104 | |
| 105 | /** |
| 106 | * Decode a run of one or more whitespace-separated encoded-words, |
| 107 | * dropping the whitespace between them per RFC 2047 §6.2 |
| 108 | * |
| 109 | * @param string $run |
| 110 | * @return string |
| 111 | */ |
| 112 | protected static function decodeRun(string $run): string |
| 113 | { |
| 114 | preg_match_all('/=\?([^?\s]+)\?([BbQq])\?([^?]*)\?=/', $run, $words, PREG_SET_ORDER); |
| 115 | |
| 116 | $decoded = ''; |
| 117 | foreach ($words as $word) { |
| 118 | $decoded .= self::decodeWord($word[1], $word[2], $word[3]); |
| 119 | } |
| 120 | |
| 121 | return $decoded; |
| 122 | } |
| 123 | |
| 124 | /** |
| 125 | * Decode a single encoded-word's charset/encoding/text into UTF-8 |
| 126 | * |
| 127 | * @param string $charset |
| 128 | * @param string $encoding |
| 129 | * @param string $encodedText |
| 130 | * @return string |
| 131 | */ |
| 132 | protected static function decodeWord(string $charset, string $encoding, string $encodedText): string |
| 133 | { |
| 134 | if (strcasecmp($encoding, 'B') === 0) { |
| 135 | $bytes = base64_decode($encodedText, true); |
| 136 | $bytes = ($bytes === false) ? '' : $bytes; |
| 137 | } else { |
| 138 | $bytes = quoted_printable_decode(str_replace('_', ' ', $encodedText)); |
| 139 | } |
| 140 | |
| 141 | return self::convertToUtf8($bytes, $charset); |
| 142 | } |
| 143 | |
| 144 | /** |
| 145 | * Convert decoded bytes to UTF-8, gracefully degrading if the charset |
| 146 | * can't be converted (no mbstring, or an unrecognized charset name) |
| 147 | * |
| 148 | * @param string $bytes |
| 149 | * @param string $charset |
| 150 | * @return string |
| 151 | */ |
| 152 | protected static function convertToUtf8(string $bytes, string $charset): string |
| 153 | { |
| 154 | if ((strcasecmp($charset, 'UTF-8') === 0) || (strcasecmp($charset, 'US-ASCII') === 0)) { |
| 155 | return $bytes; |
| 156 | } |
| 157 | |
| 158 | if (function_exists('mb_convert_encoding')) { |
| 159 | try { |
| 160 | return mb_convert_encoding($bytes, 'UTF-8', $charset); |
| 161 | } catch (\Throwable) { |
| 162 | return $bytes; |
| 163 | } |
| 164 | } |
| 165 | |
| 166 | return $bytes; |
| 167 | } |
| 168 | |
| 169 | /** |
| 170 | * Determine if a string contains anything outside printable US-ASCII |
| 171 | * and therefore needs RFC 2047 encoding |
| 172 | * |
| 173 | * @param string $text |
| 174 | * @return bool |
| 175 | */ |
| 176 | protected static function needsEncoding(string $text): bool |
| 177 | { |
| 178 | return (preg_match('/[^\x20-\x7E]/', $text) === 1); |
| 179 | } |
| 180 | |
| 181 | /** |
| 182 | * Split a UTF-8 string into byte-boundary-safe chunks, each short |
| 183 | * enough that base64-encoding it stays within $maxEncodedLength chars. |
| 184 | * Scans UTF-8 lead-byte patterns manually - no mbstring dependency |
| 185 | * needed for this direction. |
| 186 | * |
| 187 | * @param string $text |
| 188 | * @param int $maxEncodedLength |
| 189 | * @return array |
| 190 | */ |
| 191 | protected static function splitIntoChunks(string $text, int $maxEncodedLength): array |
| 192 | { |
| 193 | $chunks = []; |
| 194 | $current = ''; |
| 195 | $length = strlen($text); |
| 196 | $pos = 0; |
| 197 | |
| 198 | while ($pos < $length) { |
| 199 | $byte = ord($text[$pos]); |
| 200 | if (($byte & 0x80) === 0x00) { |
| 201 | $charLen = 1; |
| 202 | } else if (($byte & 0xE0) === 0xC0) { |
| 203 | $charLen = 2; |
| 204 | } else if (($byte & 0xF0) === 0xE0) { |
| 205 | $charLen = 3; |
| 206 | } else if (($byte & 0xF8) === 0xF0) { |
| 207 | $charLen = 4; |
| 208 | } else { |
| 209 | $charLen = 1; |
| 210 | } |
| 211 | |
| 212 | $char = substr($text, $pos, $charLen); |
| 213 | |
| 214 | if (($current !== '') && (strlen(base64_encode($current . $char)) > $maxEncodedLength)) { |
| 215 | $chunks[] = $current; |
| 216 | $current = ''; |
| 217 | } |
| 218 | |
| 219 | $current .= $char; |
| 220 | $pos += $charLen; |
| 221 | } |
| 222 | |
| 223 | if ($current !== '') { |
| 224 | $chunks[] = $current; |
| 225 | } |
| 226 | |
| 227 | return $chunks; |
| 228 | } |
| 229 | |
| 230 | } |