3 namespace Org\Mxchange\CoreFramework\Filesystem\Input\Csv;
5 // Import framework stuff
6 use Org\Mxchange\CoreFramework\Filesystem\Text\BaseInputTextFile;
7 use Org\Mxchange\CoreFramework\Stream\Filesystem\CsvInputStreamer;
10 use \InvalidArgumentException;
12 use \UnexpectedValueException;
15 * A CSV file input class for writing CSV files
17 * @author Roland Haeder <webmaster@ship-simu.org>
19 * @copyright Copyright (c) 2007, 2008 Roland Haeder, 2009 - 2021 Core Developer Team
20 * @license GNU GPL 3.0 or any newer version
21 * @link http://www.ship-simu.org
23 * This program is free software: you can redistribute it and/or modify
24 * it under the terms of the GNU General Public License as published by
25 * the Free Software Foundation, either version 3 of the License, or
26 * (at your option) any later version.
28 * This program is distributed in the hope that it will be useful,
29 * but WITHOUT ANY WARRANTY; without even the implied warranty of
30 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
31 * GNU General Public License for more details.
33 * You should have received a copy of the GNU General Public License
34 * along with this program. If not, see <http://www.gnu.org/licenses/>.
36 class CsvInputFile extends BaseInputTextFile implements CsvInputStreamer {
38 * Protected constructor
42 private function __construct () {
43 // Call parent constructor
44 parent::__construct(__CLASS__);
48 * Creates an instance of this File class and prepares it for usage
50 * @param $infoInstance An instance of a SplFileInfo class
51 * @return $fileInstance An instance of this File class
53 public final static function createCsvInputFile (SplFileInfo $infoInstance) {
55 $fileInstance = new CsvInputFile();
57 // Init this abstract file
58 $fileInstance->initFile($infoInstance);
60 // Return the prepared instance
65 * Reads a line from CSV file and returns it as an indexed array. Please
66 * note that strings *must* be always in double-quotes, else any found
67 * column separators will be parsed or they may be interpreted incorrectly.
69 * @param $columnSeparator Character to use separting columns
70 * @param $expectedMatches Expected matches, 0 is default and means flexible
71 * @return $lineArray An indexed array with the read line
72 * @throws InvalidArgumentException If a parameter is invalid
73 * @throws UnexpectedValueException If the array count is not matching expected count
75 public function readCsvFileLine (string $columnSeparator, int $expectedMatches = 0) {
77 /* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] columnSeparator=%s,expectedMatches=%d - CALLED!', __METHOD__, __LINE__, $columnSeparator, $expectedMatches));
78 if (strlen($columnSeparator) === 0) {
79 // No empty column separator
80 throw new InvalidArgumentException('columnSeparator cannot be empty.');
81 } elseif ($expectedMatches < 0) {
82 // Below zero is not valid
83 throw new InvalidArgumentException(sprintf('expectedMatches=%d is below zero', $expectedMatches));
86 // Read raw line and trim anything unwanted away
87 $data = trim($this->getPointerInstance()->readLine());
90 /* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] data(%d)=%s', __METHOD__, __LINE__, strlen($data), $data));
92 // Yes, then skip below code
93 /* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] Read data is an empty line - EXIT!', __METHOD__, __LINE__));
98 $lineArray = $this->parseDataToIndexedArray($data, $columnSeparator);
100 // Is the expected count found?
101 /* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] expectedMatches=%d,lineArray()=%d', __METHOD__, __LINE__, $expectedMatches, count($lineArray)));
102 if (($expectedMatches > 0) && (count($lineArray) !== $expectedMatches)) {
103 // Invalid line found as strict count matching is requested
104 throw new UnexpectedValueException(sprintf('lineArray()=%d has not expected count %d', count($lineArray), $expectedMatches));
108 /* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] lineArray()=%d - EXIT!', __METHOD__, __LINE__, count($lineArray)));
113 * Parses given data into an array
115 * @param $data Raw data e.g. returned from readLine()
116 * @param $columnSeparator Character to use separting columns
117 * @return $lineArray An indexed array with the read line
119 private function parseDataToIndexedArray (string $data, string $columnSeparator) {
121 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] data()=%d,columnSeparator=%s - CALLED!', __METHOD__, __LINE__, strlen($data), $columnSeparator));
124 // Whether the parser reads a quoted string (which may contain the column separator again)
130 // Now parse the line
131 for ($idx = 0; $idx < strlen($data); $idx++) {
133 $char = substr($data, $idx, 1);
135 // Is the column separator found and not within quotes?
136 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] idx=%d,char=%s ...', __METHOD__, __LINE__, $idx, $char));
137 if (($isInQuotes === false) && ($char == $columnSeparator)) {
138 // Add this line to the array
139 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] Adding column=%s ...', __METHOD__, __LINE__, $column));
140 array_push($lineArray, $column);
142 // Clear variable ...
143 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] line[]=%d - After add!', __METHOD__, __LINE__, count($lineArray)));
148 } elseif ($char == chr(34)) {
149 // $column must be empty at this point if we are at starting quote
150 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] column=%s ...', __METHOD__, __LINE__, $column));
151 assert(($isInQuotes === true) || (empty($column)));
153 // Double-quote found, so flip variable
154 $isInQuotes = (!$isInQuotes);
156 // Skip double-quote (escaping of them is not yet supported)
157 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] isInQuotes=%d ...', __METHOD__, __LINE__, intval($isInQuotes)));
161 // Add char to column
162 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] Adding char=%s ...', __METHOD__, __LINE__, $idx, $char));
166 // Is there something outstanding?
167 if (!empty($column)) {
168 // Then don't forget this. :-)
169 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] Adding column=%s ...', __METHOD__, __LINE__, $column));
170 array_push($lineArray, $column);
173 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] line[]=%d - After add!', __METHOD__, __LINE__, count($lineArray)));
177 //* NOISY-DEBUG: */ self::createDebugInstance(__CLASS__, __LINE__)->debugOutput(sprintf('[%s:%d:] line[]=%d - EXIT!', __METHOD__, __LINE__, count($lineArray)));