fgetcsv
(PHP 4, PHP 5)
fgetcsv — Читает строку из файла и производит разбор данных CSV
Описание
- handle
- Корректный файловый указатель на файл, успешно открытый при помощи fopen(), popen(), или fsockopen().
- length (Необязательный)
- Должен быть больше самой длинной строки (в символах), найденной в CSV-файле (включая завершающий символ конца строки). Он стал необязательным в PHP 5. Если этот аргумент не указан (или равен 0 в версиях PHP 5.0.4 и выше), максимальная длинна строки не ограничена, но функция работает немного медленнее.
- delimiter (Необязательный)
- Устанавливает разделитель поля (только один символ). По умолчанию это запятая.
- enclosure (Необязательный)
- Устанавливает символ ограничителя поля (только один символ). По умолчанию это двойная кавычка. Добавлен в PHP 4.3.0.
Данная функция похожа на функцию fgets(), с той разницей, что она производит анализ строки на наличие записей в формате CSV и возвращает найденные поля в качестве массива.
Функция fgetcsv() возвращает FALSE в случае ошибки, а также по достижению конца файла.
Замечание: Пустая строка CSV-файла будет возвращена в качестве массива, содержащего единственный элемент null, ошибки в данном случае не возникнет.
Пример #1 Чтение и вывод на экран содержания CSV-файла
<?php
$row = 1;
$handle = fopen("test.csv", "r");
while (($data = fgetcsv($handle, 1000, ",")) !== FALSE) {
$num = count($data);
echo "<p> $num полей в строке $row: <br /></p>\n";
$row++;
for ($c=0; $c < $num; $c++) {
echo $data[$c] . "<br />\n";
}
}
fclose($handle);
?>
fgetcsv() стала правильно обрабатывать двоичные данные начиная с версии PHP 4.3.5
Замечание: Эта функция принимает во внимание настройки локали. К примеру, если LANG установлен в en_US.UTF-8, то файлы в однобайтовой кодировке будут неправильно прочитаны этой функцией.
Замечание: Если у вас возникают проблемы с распознаванием PHP окончания строк при чтении файлов на Macintosh-совместимом компьютере или при чтении файлов, созданных на Macintosh-совместимом компьютере, необходимо включить опцию auto_detect_line_endings.
См.также описание функций explode(), file(), pack() и fputcsv().
- PHP Руководство
- Функции по категориям
- Индекс функций
- Справочник функций
- Расширения для работы с файловой системой
- Функции для работы с файловой системой
- basename
- chgrp
- chmod
- chown
- clearstatcache
- copy
- delete
- dirname
- disk_free_space
- disk_total_space
- diskfreespace
- fclose
- feof
- fflush
- fgetc
- fgetcsv
- fgets
- fgetss
- file_exists
- file_get_contents
- file_put_contents
- file
- fileatime
- filectime
- filegroup
- fileinode
- filemtime
- fileowner
- fileperms
- filesize
- filetype
- flock
- fnmatch
- fopen
- fpassthru
- fputcsv
- fputs
- fread
- fscanf
- fseek
- fstat
- ftell
- ftruncate
- fwrite
- glob
- is_dir
- is_executable
- is_file
- is_link
- is_readable
- is_uploaded_file
- is_writable
- is_writeable
- lchgrp
- lchown
- link
- linkinfo
- lstat
- mkdir
- move_uploaded_file
- parse_ini_file
- parse_ini_string
- pathinfo
- pclose
- popen
- readfile
- readlink
- realpath_cache_get
- realpath_cache_size
- realpath
- rename
- rewind
- rmdir
- set_file_buffer
- stat
- symlink
- tempnam
- tmpfile
- touch
- umask
- unlink
Коментарии
You should pay attention to the fact that "fgetcsv" does remove leading TAB-chars "chr(9)" while reading the file.
This means if you have a chr(9) as the first char in the file and you use fgetcsv this char is automaticaly deleted.
Example:
file content:
chr(9)first#second#third#fourth
source:
<?php $line = fgetcsv($handle,500,"#"); ?>
The array $line looks like:
$line[0] = first
$line[1] = second
$line[2] = third
$line[3] = fourth
and not
$line[0] = chr(9)first
$line[1] = second
$line[2] = third
$line[3] = fourth
All chr(9) after another char is not deleted!
Example:
file content:
Achr(9)first#second#third#fourth
source:
<?php $line = fgetcsv($handle,500,"#"); ?>
The array $line looks like:
$line[0] = Achr(9)first
$line[1] = second
$line[2] = third
$line[3] = fourth
Another version [modified michael from mediaconcepts]
<?php
function arrayFromCSV($file, $hasFieldNames = false, $delimiter = ',', $enclosure='') {
$result = Array();
$size = filesize($file) +1;
$file = fopen($file, 'r');
#TO DO: There must be a better way of finding out the size of the longest row... until then
if ($hasFieldNames) $keys = fgetcsv($file, $size, $delimiter, $enclosure);
while ($row = fgetcsv($file, $size, $delimiter, $enclosure)) {
$n = count($row); $res=array();
for($i = 0; $i < $n; $i++) {
$idx = ($hasFieldNames) ? $keys[$i] : $i;
$res[$idx] = $row[i];
}
$result[] = $res;
}
fclose($file);
return $result;
}
?>
I've had alot of projects recently dealing with csv files, so I created the following class to read a csv file and return an array of arrays with the column names as keys. The only requirement is that the 1st row contain the column headings.
I only wrote it today, so I'll probably expand on it in the near future.
<?php
class CSVparse
{
var $mappings = array();
function parse_file($filename)
{
$id = fopen($filename, "r"); //open the file
$data = fgetcsv($id, filesize($filename)); /*This will get us the */
/*main column names */
if(!$this->mappings)
$this->mappings = $data;
while($data = fgetcsv($id, filesize($filename)))
{
if($data[0])
{
foreach($data as $key => $value)
$converted_data[$this->mappings[$key]] = addslashes($value);
$table[] = $converted_data; /* put each line into */
} /* its own entry in */
} /* the $table array */
fclose($id); //close file
return $table;
}
}
?>
Hier is an example for a CSV Iterator.
<?php
class CsvIterator implements Iterator
{
const ROW_SIZE = 4096;
/**
* The pointer to the cvs file.
* @var resource
* @access private
*/
private $filePointer = null;
/**
* The current element, which will
* be returned on each iteration.
* @var array
* @access private
*/
private $currentElement = null;
/**
* The row counter.
* @var int
* @access private
*/
private $rowCounter = null;
/**
* The delimiter for the csv file.
* @var str
* @access private
*/
private $delimiter = null;
/**
* This is the constructor.It try to open the csv file.The method throws an exception
* on failure.
*
* @access public
* @param str $file The csv file.
* @param str $delimiter The delimiter.
*
* @throws Exception
*/
public function __construct($file, $delimiter=',')
{
try {
$this->filePointer = fopen($file, 'r');
$this->delimiter = $delimiter;
}
catch (Exception $e) {
throw new Exception('The file "'.$file.'" cannot be read.');
}
}
/**
* This method resets the file pointer.
*
* @access public
*/
public function rewind() {
$this->rowCounter = 0;
rewind($this->filePointer);
}
/**
* This method returns the current csv row as a 2 dimensional array
*
* @access public
* @return array The current csv row as a 2 dimensional array
*/
public function current() {
$this->currentElement = fgetcsv($this->filePointer, self::ROW_SIZE, $this->delimiter);
$this->rowCounter++;
return $this->currentElement;
}
/**
* This method returns the current row number.
*
* @access public
* @return int The current row number
*/
public function key() {
return $this->rowCounter;
}
/**
* This method checks if the end of file is reached.
*
* @access public
* @return boolean Returns true on EOF reached, false otherwise.
*/
public function next() {
return !feof($this->filePointer);
}
/**
* This method checks if the next row is a valid row.
*
* @access public
* @return boolean If the next row is a valid row.
*/
public function valid() {
if (!$this->next()) {
fclose($this->filePointer);
return false;
}
return true;
}
}
?>
Usage :
<?php
$csvIterator = new CsvIterator('/path/to/csvfile.csv');
foreach ($csvIterator as $row => $data) {
// do somthing with $data
}
?>
The fgetcsv function seems to follow the MS excel conventions, which means:
- The quoting character is escaped by itself and not the back slash.
(i.e.Let's use the double quote (") as the quoting character:
Two double quotes "" will give a single " once parsed, if they are inside a quoted field (otherwise neither of them will be removed).
\" will give \" whether it is in a quoted field or not (same for \\) , and
if a single double quote is inside a quoted field it will be removed. If it is not inside a quoted field it will stay).
- leading and trailing spaces (\s or \t) are never removed, regardless of whether they are in quoted fields or not.
- Line breaks within fields are dealt with correctly if they are in quoted fields. (So previous comments stating the opposite are wrong, unless they are using a different PHP version.... I am using 4.4.0.)
So fgetcsv if actually very complete and can deal with every possible situation. (It does need help for macintosh line breaks though, as mentioned in the help files.)
I wish I knew all this from the start. From my own benchmarks fgetcsv strikes a very good compromise between memory consumption and speed.
-------------------------
Note: If back slashes are used to escape quotes they can easily be removed afterwards. Same for leading and trailing spaces.
beware of characters of binary value == 0, as they seem to make fgetcsv ignore the remaining part of a line where they appear.
Maybe this is normal under some convention I don't know, but a file exported from Excel had those as values for some cells *sometimes*, thus fgetcsv return variable cell counts for different lines.
i'm using php 4.3
Note that fgetcsv() uses the system locale setting to make assumptions about character encoding.
So if you are trying to process a UTF-8 CSV file on an EUC-JP server (for example),
you will need to do something like this before you call fgetcsv():
setlocale(LC_ALL, 'ja_JP.UTF8');
[Also not that setlocale() doesn't *permanently* affect the system locale setting]
Here is a OOP based importer similar to the one posted earlier. However, this is slightly more flexible in that you can import huge files without running out of memory, you just have to use a limit on the get() method
Sample usage for small files:-
-------------------------------------
<?php
$importer = new CsvImporter("small.txt",true);
$data = $importer->get();
print_r($data);
?>
Sample usage for large files:-
-------------------------------------
<?php
$importer = new CsvImporter("large.txt",true);
while($data = $importer->get(2000))
{
print_r($data);
}
?>
And heres the class:-
-------------------------------------
<?php
class CsvImporter
{
private $fp;
private $parse_header;
private $header;
private $delimiter;
private $length;
//--------------------------------------------------------------------
function __construct($file_name, $parse_header=false, $delimiter="\t", $length=8000)
{
$this->fp = fopen($file_name, "r");
$this->parse_header = $parse_header;
$this->delimiter = $delimiter;
$this->length = $length;
$this->lines = $lines;
if ($this->parse_header)
{
$this->header = fgetcsv($this->fp, $this->length, $this->delimiter);
}
}
//--------------------------------------------------------------------
function __destruct()
{
if ($this->fp)
{
fclose($this->fp);
}
}
//--------------------------------------------------------------------
function get($max_lines=0)
{
//if $max_lines is set to 0, then get all the data
$data = array();
if ($max_lines > 0)
$line_count = 0;
else
$line_count = -1; // so loop limit is ignored
while ($line_count < $max_lines && ($row = fgetcsv($this->fp, $this->length, $this->delimiter)) !== FALSE)
{
if ($this->parse_header)
{
foreach ($this->header as $i => $heading_i)
{
$row_new[$heading_i] = $row[$i];
}
$data[] = $row_new;
}
else
{
$data[] = $row;
}
if ($max_lines > 0)
$line_count++;
}
return $data;
}
//--------------------------------------------------------------------
}
?>
Here's something I put together this morning. It allows you to read rows from your CSV and get values based on the name of the column. This works great when your header columns are not always in the same order; like when you're processing many feeds from different customers. Also makes for cleaner, easier to manage code.
So if your feed looks like this:
product_id,category_name,price,brand_name, sku_isbn_upc,image_url,landing_url,title,description
123,Test Category,12.50,No Brand,0,http://www.example.com, http://www.example.com/landing.php, Some Title,Some Description
You can do:
<?php
while ($o->getNext())
{
$dPrice = $o->getPrice();
$nProductID = $o->getProductID();
$sBrandName = $o->getBrandName();
}
?>
If you have any questions or comments regarding this class, they can be directed to michael.martinek@gmail.com as I probably won't be checking back here.
<?php
define('C_PPCSV_HEADER_RAW', 0);
define('C_PPCSV_HEADER_NICE', 1);
class PaperPear_CSVParser
{
private $m_saHeader = array();
private $m_sFileName = '';
private $m_fp = false;
private $m_naHeaderMap = array();
private $m_saValues = array();
function __construct($sFileName)
{
//quick and dirty opening and processing.. you may wish to clean this up
if ($this->m_fp = fopen($sFileName, 'r'))
{
$this->processHeader();
}
}
function __call($sMethodName, $saArgs)
{
//check to see if this is a set() or get() request, and extract the name
if (preg_match("/[sg]et(.*)/", $sMethodName, $saFound))
{
//convert the name portion of the [gs]et to uppercase for header checking
$sName = strtoupper($saFound[1]);
//see if the entry exists in our named header-> index mapping
if (array_key_exists($sName, $this->m_naHeaderMap))
{
//it does.. so consult the header map for which index this header controls
$nIndex = $this->m_naHeaderMap[$sName];
if ($sMethodName{0} == 'g')
{
//return the value stored in the index associated with this name
return $this->m_saValues[$nIndex];
}
else
{
//set the valuw
$this->m_saValues[$nIndex] = $saArgs[0];
return true;
}
}
}
//nothing we control so bail out with a false
return false;
}
//get a nicely formatted header name. This will take product_id and make
//it PRODUCTID in the header map. So now you won't need to worry about whether you need
//to do a getProductID, or getproductid, or getProductId.. all will work.
public static function GetNiceHeaderName($sName)
{
return strtoupper(preg_replace('/[^A-Za-z0-9]/', '', $sName));
}
//process the header entry so we can map our named header fields to a numerical index, which
//we'll use when we use fgetcsv().
private function processHeader()
{
$sLine = fgets($this->m_fp);
//you'll want to make this configurable
$saFields = split(",", $sLine);
$nIndex = 0;
foreach ($saFields as $sField)
{
//get the nice name to use for "get" and "set".
$sField = trim($sField);
$sNiceName = PaperPear_CSVParser::GetNiceHeaderName($sField);
//track correlation of raw -> nice name so we don't have to do on-the-fly nice name checks
$this->m_saHeader[$nIndex] = array(C_PPCSV_HEADER_RAW => $sField, C_PPCSV_HEADER_NICE => $sNiceName);
$this->m_naHeaderMap[$sNiceName] = $nIndex;
$nIndex++;
}
}
//read the next CSV entry
public function getNext()
{
//this is a basic read, you will likely want to change this to accomodate what
//you are using for CSV parameters (tabs, encapsulation, etc).
if (($saValues = fgetcsv($this->m_fp)) !== false)
{
$this->m_saValues = $saValues;
return true;
}
return false;
}
}
//quick example of usage
$o = new PaperPear_CSVParser('F:\foo.csv');
while ($o->getNext())
{
echo "Price=" . $o->getPrice() . "\r\n";
}
?>
If you need to set auto_detect_line_endings to deal with Mac line endings, it may seem obvious but remember it should be set before fopen, not after:
This will work:
<?php
ini_set('auto_detect_line_endings',TRUE);
$handle = fopen('/path/to/file','r');
while ( ($data = fgetcsv($handle) ) !== FALSE ) {
//process
}
ini_set('auto_detect_line_endings',FALSE);
?>
This won't, you will still get concatenated fields at the new line position:
<?php
$handle = fopen('/path/to/file','r');
ini_set('auto_detect_line_endings',TRUE);
while ( ($data = fgetcsv($handle) ) !== FALSE ) {
//process
}
ini_set('auto_detect_line_endings',FALSE);
?>
I used fgetcsv to read pipe-delimited data files, and ran into the following quirk.
The data file contained data similar to this:
RECNUM|TEXT|COMMENT
1|hi!|some comment
2|"error!|another comment
3|where does this go?|yet another comment
4|the end!"|last comment
I read the file like this:
<?php
$row = fgetcsv( $fi, $length, '|' );
?>
This causes a problem on record 2: the quote immediately after the pipe causes the file to be read up to the following quote --in this case, in record 4. Everything in between was stored in a single element of $row.
In this particular case it is easy to spot, but my script was processing thousands of records and it took me some time to figure out what went wrong.
The annoying thing is, that there doesn't seem to be an elegant fix. You can't tell PHP not to use an enclosure --for example, like this:
<?php
$row = fgetcsv( $fi, $length, '|', '' );
?>
(Well, you can tell PHP that, but it doesn't work.)
So you'd have to resort to a solution where you use an extremely unlikely enclosure, but since the enclosure can only be one character long, it may be hard to find.
Alternatively (and IMNSHO: more elegantly), you can choose to read these files like this, instead:
<?php
$line = fgets( $fi, $length );
$row = explode( '|', $line );
?>
As it's more intuitive and resilient, I've decided to favor this 'construct' over fgetcsv from now on.
Note that fgetcsv, at least in PHP 5.3 or previous, will NOT work with UTF-16 encoded files. Your options are to convert the entire file to ISO-8859-1 (or latin1), or convert line by line and convert each line into ISO-8859-1 encoding, then use str_getcsv (or compatible backwards-compatible implementation). If you need to read non-latin alphabets, probably best to convert to UTF-8.
See str_getcsv for a backwards-compatible version of it with PHP < 5.3, and see utf8_decode for a function written by Rasmus Andersson which provides utf16_decode. The modification I added was that the BOP appears at the top of the file, then not on subsequent lines. So you need to store the endian-ness, and then re-send it upon each subsequent line decoding. This modified version returns the endianness, if it's not available:
<?php
/**
* Decode UTF-16 encoded strings.
*
* Can handle both BOM'ed data and un-BOM'ed data.
* Assumes Big-Endian byte order if no BOM is available.
* From: function.utf8-decode
*
* @param string $str UTF-16 encoded data to decode.
* @return string UTF-8 / ISO encoded data.
* @access public
* @version 0.1 / 2005-01-19
* @author Rasmus Andersson {@link http://rasmusandersson.se/}
* @package Groupies
*/
function utf16_decode($str, &$be=null) {
if (strlen($str) < 2) {
return $str;
}
$c0 = ord($str{0});
$c1 = ord($str{1});
$start = 0;
if ($c0 == 0xFE && $c1 == 0xFF) {
$be = true;
$start = 2;
} else if ($c0 == 0xFF && $c1 == 0xFE) {
$start = 2;
$be = false;
}
if ($be === null) {
$be = true;
}
$len = strlen($str);
$newstr = '';
for ($i = $start; $i < $len; $i += 2) {
if ($be) {
$val = ord($str{$i}) << 4;
$val += ord($str{$i+1});
} else {
$val = ord($str{$i+1}) << 4;
$val += ord($str{$i});
}
$newstr .= ($val == 0x228) ? "\n" : chr($val);
}
return $newstr;
}
?>
Trying the "setlocale" trick did not work for me, e.g.
<?php
setlocale(LC_CTYPE, "en.UTF16");
$line = fgetcsv($file, ...)
?>
But that's perhaps because my platform didn't support it. However, fgetcsv only supports single characters for the delimiter, etc. and complains if you pass in a UTF-16 version of said character, so I gave up on that rather quickly.
Hope this is helpful to someone out there.
If you want to load some translations for your application, don't use csv files for that, even if it's easier to handle.
The following code snippet:
<?php
$lang = array();
$handle = fopen('en.csv', 'r');
while($row = fgetcsv($handle, 500, ';'))
{
$lang[$row[0]] = $row[1];
}
fclose($handle);
?>
is about 400% slower than this code:
<?php
$lang = array();
$values = parse_ini_file('de.ini');
foreach($values as $key => $val)
{
$lang[$key] = $val;
}
?>
That's the reason why you should allways use .ini files for translations...
http://php.net/parse_ini_file
This is how to read a csv file into a multidimensional array.
<?php
# Open the File.
if (($handle = fopen("file.csv", "r")) !== FALSE) {
# Set the parent multidimensional array key to 0.
$nn = 0;
while (($data = fgetcsv($handle, 1000, ",")) !== FALSE) {
# Count the total keys in the row.
$c = count($data);
# Populate the multidimensional array.
for ($x=0;$x<$c;$x++)
{
$csvarray[$nn][$x] = $data[$x];
}
$nn++;
}
# Close the File.
fclose($handle);
}
# Print the contents of the multidimensional array.
print_r($csvarray);
?>
I had a problem with multibytes. File was windows-1250, script was UTF-8 and set_locale wasn't work so I made a simple and safe workaround:
<?php
$fc = iconv('windows-1250', 'utf-8', file_get_contents($_FILES['csv']['tmp_name']));
file_put_contents('tmp/import.tmp', $fc);
$handle = fopen('tmp/import.tmp', "r");
$rows = array();
while (($data = fgetcsv($handle, 0, ";")) !== FALSE) {
$rows[] = $data;
}
fclose($handle);
unlink('tmp/import.tmp');
?>
I hope You will find it out usefull.
Sorry for my english.
I needed a function to analyse a file for delimiters and line endings prior to importing the file into MySQL using LOAD DATA LOCAL INFILE
I wrote this function to do the job, the results are (mostly) very accurate and it works nicely with large files too.
<?php
function analyse_file($file, $capture_limit_in_kb = 10) {
// capture starting memory usage
$output['peak_mem']['start'] = memory_get_peak_usage(true);
// log the limit how much of the file was sampled (in Kb)
$output['read_kb'] = $capture_limit_in_kb;
// read in file
$fh = fopen($file, 'r');
$contents = fread($fh, ($capture_limit_in_kb * 1024)); // in KB
fclose($fh);
// specify allowed field delimiters
$delimiters = array(
'comma' => ',',
'semicolon' => ';',
'tab' => "\t",
'pipe' => '|',
'colon' => ':'
);
// specify allowed line endings
$line_endings = array(
'rn' => "\r\n",
'n' => "\n",
'r' => "\r",
'nr' => "\n\r"
);
// loop and count each line ending instance
foreach ($line_endings as $key => $value) {
$line_result[$key] = substr_count($contents, $value);
}
// sort by largest array value
asort($line_result);
// log to output array
$output['line_ending']['results'] = $line_result;
$output['line_ending']['count'] = end($line_result);
$output['line_ending']['key'] = key($line_result);
$output['line_ending']['value'] = $line_endings[$output['line_ending']['key']];
$lines = explode($output['line_ending']['value'], $contents);
// remove last line of array, as this maybe incomplete?
array_pop($lines);
// create a string from the legal lines
$complete_lines = implode(' ', $lines);
// log statistics to output array
$output['lines']['count'] = count($lines);
$output['lines']['length'] = strlen($complete_lines);
// loop and count each delimiter instance
foreach ($delimiters as $delimiter_key => $delimiter) {
$delimiter_result[$delimiter_key] = substr_count($complete_lines, $delimiter);
}
// sort by largest array value
asort($delimiter_result);
// log statistics to output array with largest counts as the value
$output['delimiter']['results'] = $delimiter_result;
$output['delimiter']['count'] = end($delimiter_result);
$output['delimiter']['key'] = key($delimiter_result);
$output['delimiter']['value'] = $delimiters[$output['delimiter']['key']];
// capture ending memory usage
$output['peak_mem']['end'] = memory_get_peak_usage(true);
return $output;
}
?>
Example Usage:
<?php
$Array = analyse_file('/www/files/file.csv', 10);
// example usable parts
// $Array['delimiter']['value'] => ,
// $Array['line_ending']['value'] => \r\n
?>
Full function output:
Array
(
[peak_mem] => Array
(
[start] => 786432
[end] => 786432
)
[line_ending] => Array
(
[results] => Array
(
[nr] => 0
[r] => 4
[n] => 4
[rn] => 4
)
[count] => 4
[key] => rn
[value] =>
)
[lines] => Array
(
[count] => 4
[length] => 94
)
[delimiter] => Array
(
[results] => Array
(
[colon] => 0
[semicolon] => 0
[pipe] => 0
[tab] => 1
[comma] => 17
)
[count] => 17
[key] => comma
[value] => ,
)
[read_kb] => 10
)
Enjoy!
Ashley
Parse from Microsoft Excel "Unicode Text (*.txt)" format:
<?php
function parse($file) {
if (($handle = fopen($file, "r")) === FALSE) return;
while (($cols = fgetcsv($handle, 1000, "\t")) !== FALSE) {
foreach( $cols as $key => $val ) {
$cols[$key] = trim( $cols[$key] );
$cols[$key] = iconv('UCS-2', 'UTF-8', $cols[$key]."\0") ;
$cols[$key] = str_replace('""', '"', $cols[$key]);
$cols[$key] = preg_replace("/^\"(.*)\"$/sim", "$1", $cols[$key]);
}
echo print_r($cols, 1);
}
}
?>
I had a csv file whose fields included data with line endings (CRLF created by hitting the carriage returns in html textarea). Of course, the LF in these fields was escaped by MySQL during the creation of the csv. Problem is I could NOT get fgetcsv to work correctly here, since each and every LF was regarded as the end of a line of the csv file, even when it was escaped!
Since what I wanted was to get THE FIRST LINE of the csv file, then count the number of fields by exploding on all unescaped commas, I had to resort to this:
<?php
/*
First five lines of csv: the 4th row has a line-break within a data field. The LFs represent line-feeds or \n
1,okonkwo joseph,nil,2010-01-12 17:41:40LF
2,okafor john,cq and sulphonamides,2010-01-12 17:58:03LF
3,okoye andrew,lives with hubby in abuja,2011-03-30 13:39:19LF
4,okeke peter,In 2001\, had appendicectomy in AbaCR
\LF
In 2004\, had ELCS at a private hoapital in Lagos,2011-03-30 13:39:19LF
5,adewale chris,cq and sulphonamides,2010-01-12 17:58:03LF
*/
$fp = fopen('file.csv', 'r');
$i = 1;
$str='';
$srch='';
while (false !== ($char = fgetc($fp))) {
$str .= $char;//use this to collect the string for outputting
$srch .= $char;//use this to search for LF, possible preceded by \'
if(strlen($srch) > 2){
$srch = substr($srch, 1);//ie trim off the first char
}
if($i > 1 && $srch[1] == chr(10) && $srch[0] != '\\'){//chr(10) is LF, ie \n
break;//if you get to the \n NOT preceded by \, that's the real line-ending, stop collecting the string;
}
$i++;
}
echo $str;//should contain the first line as string
?>
Perhaps there exists a more elegant solution to this issue, in which case I'd be glad to know!
I was getting a bytes exhausted error when iterating through a CSV file. ini_set('auto_detect_line_endings', 1); fixed it.fgetcsv seems to handle newlines within fields fine. So in fact it is not reading a line, but keeps reading untill it finds a \n-character that's not quoted as a field.
Example:
<?php
/* test.csv contains:
"col 1","col2","col3"
"this
is
having
multiple
lines","this not","this also not"
"normal record","nothing to see here","no data"
*/
$handle = fopen("test.csv", "r");
while (($data = fgetcsv($handle)) !== FALSE) {
var_dump($data);
}
?>
Returns:
array(3) {
[0]=>
string(5) "col 1"
[1]=>
string(4) "col2"
[2]=>
string(4) "col3"
}
array(3) {
[0]=>
string(29) "this
is
having
multiple
lines"
[1]=>
string(8) "this not"
[2]=>
string(13) "this also not"
}
array(3) {
[0]=>
string(13) "normal record"
[1]=>
string(19) "nothing to see here"
[2]=>
string(7) "no data"
}
This means that you can expect fgetcsv to handle newlines within fields fine. This was not clear from the documentation.
If you don't want to define an enclosure charachter you can do the following:
<?php
$row = fgetcsv($handle, 0, $delimiter, 0x00);
?>
I needed this to detect the enclosure used for csv files.
For anyone else struggling with disappearing non-latin characters in one-byte encodings - setting LANG env var (as the manual states) does not help at all. Look at LC_ALL instead.
In my case it was set to "pl_PL.utf8" but since my input file was in CP1250 most of polish characters (but not all of them!) had gone missing and city of "Łódź" had become just "dź". I've "fixed" it with "pl_PL".
Here is an example how to use this function with generators
https://github.com/luchaninov/csv-file-loader (composer require "luchaninov/csv-file-loader:1.*")
$loader = new CsvFileLoader();
$loader->setFilename('/path/to/your_data.csv');
foreach ($loader->getItems() as $item) {
var_dump($item); // do something here
}
If you have CSV-file like
id,name,surname
1,Jack,Black
2,John,Doe
you'll get 2 items
['id' => '1', 'name' => 'Jack', 'surname' => 'Black']
['id' => '2', 'name' => 'John', 'surname' => 'Doe']
Setting the $escape parameter dosn't return unescaped strings, but just avoid splitting on a $delimiter that have an escpae-char infront of it:
<?php
$tmp_file = "/tmp/test.csv";
file_put_contents($tmp_file, "\"first\\\";\\\"secound\"");
echo "raw:" . PHP_EOL . file_get_contents($tmp_file) . PHP_EOL . PHP_EOL;
echo "fgetcsv escaped bs:" . PHP_EOL;
$f = fopen($tmp_file, 'r');
while($r = fgetcsv($f, 1024, ';', '"', "\\"))
{
print_r($r);
}
fclose($f);
echo PHP_EOL;
echo "fgetcsv escaped #:" . PHP_EOL;
$f = fopen($tmp_file, 'r');
while($r = fgetcsv($f, 1024, ';', '"', "#"))
{
print_r($r);
}
fclose($f);
echo PHP_EOL;
?>
The $escape parameter is completely unintuitive, but it is not broken. Here is a breakdown of fgetcsv()'s behaviour. In the examples I've used underscores (_) to show spaces and brackets ([]) to show individual fields:
- Leading whitespace in each field will be stripped if it comes immediately before an enclosure: ___"foo" -> [foo]
- There can only be one enclosure per field, although it will be concatenated with any data that appears between the end enclosure and the next delimiter/new line, including any trailing whitespaces ___"foo"_"bar"__ -> [foo_"bar"__]
- If the field does not start with (leading whitespace +) an enclosure, the whole field is interpreted as raw data, even if enclosure characters appear elsewhere within the field: _foo"bar"_ -> [_foo"bar"_]
- Delimiters cannot be escaped outside enclosures, they have to be enclosed instead. Delimiters don't need to be escaped inside enclosures: "foo,bar","baz,qux" -> [foo,bar][baz,qux]; foo\,bar -> [foo\][bar]; "foo\,bar" -> [foo\,bar]
- Double enclosures inside single enclosures are converted to single enclosures: "foobar" -> [foobar]; "foo""bar" -> [foo"bar]; """foo""" -> ["foo"]; ""foo"" -> [foo""] (empty enclosure followed by raw data)
- The $escape parameter works as expected, but unlike enclosures DOES NOT get unescaped. It is necessary to unescape the data elsewhere in the code: "\"foo\"" -> [\"foo\"]; "foo\"bar" -> [foo\"bar]
Note: the following data (which is a very common problem) is invalid: "\". Its structure is equivalent to "@ or in other words, an open enclosure, some data and no closing enclosure.
The following functions can be used to get the expected behaviour:
<?php
// Removes escape characters before both enclosures and escapes, but leaves everything else untouched, similiar to single quoting
function fgetcsv_unescape_enclosures_and_escapes($fh, $length = 0, $delimiter = ',', $enclosure = '"', $escape = '\\') {
$fields = fgetcsv($fh, $length, $delimiter, $enclosure, $escape);
if ($fields) {
$regex_enclosure = preg_quote($enclosure);
$regex_escape = preg_quote($escape);
$fields = preg_replace("/{$regex_escape}({$regex_enclosure}|{$regex_escape})/", '$1', $fields);
}
return $fields;
}
// Does NOT remove a lone escape character at the end of a field
function fgetcsv_unescape_all($fh, $length = 0, $delimiter = ',', $enclosure = '"', $escape = '\\') {
$fields = fgetcsv($fh, $length, $delimiter, $enclosure, $escape);
if ($fields) {
$regex_escape = preg_quote($escape);
$fields = preg_replace("/{$regex_escape}(.)/s", '$1', $fields);
}
return $fields;
}
// Removes lone escape characters at the end of fields
function fgetcsv_unescape_all_strip_last($fh, $length = 0, $delimiter = ',', $enclosure = '"', $escape = '\\') {
$fields = fgetcsv($fh, $length, $delimiter, $enclosure, $escape);
if ($fields) {
$regex_escape = preg_quote($escape);
$fields = preg_replace("/{$regex_escape}(.?)/s", '$1', $fields);
}
return $fields;
}
?>
Caution: ideally, there shouldn't be any unescaped escape characters outside enclosures; the field should be enclosed and escaped instead. If there are any, they could end up being removed as well, depending on the function used.
This function has no special BOM handling. The first cell of the first row will inherit the BOM bytes, i.e. will be 3 bytes longer than expected. As the BOM is invisible you may not notice.
Excel on Windows, or text editors like Notepad, may add the BOM.
Forget this while() loop mumbo jumbo! Use this:
$rows = array_map('str_getcsv', file('myfile.csv'));
$header = array_shift($rows);
$csv = array();
foreach ($rows as $row) {
$csv[] = array_combine($header, $row);
}
Source: https://steindom.com/articles/shortest-php-code-convert-csv-associative-array
When a BOM character is suppled, `fgetscsv` may appear to wrap the first element in "double quotation marks". The simplest way to ignore it is to progress the file pointer to the 4th byte before using `fgetcsv`.
<?php
// BOM as a string for comparison.
$bom = "\xef\xbb\xbf";
// Read file from beginning.
$fp = fopen($path, 'r');
// Progress file pointer and get first 3 characters to compare to the BOM string.
if (fgets($fp, 4) !== $bom) {
// BOM not found - rewind pointer to start of file.
rewind($fp);
}
// Read CSV into an array.
$lines = array();
while(!feof($fp) && ($line = fgetcsv($fp)) !== false) {
$lines[] = $line;
}
?>
In-case anyone is having difficulty working around Byte-order-marks, the following should work. As usual no warranty, you should test your code... It's for UTF-8 only
<?php
//...
$fh = fopen('wut.csv', 'r');
$firstThreeBytes = fread($fh , 3);
if($firstThreeBytes !== "\xef\xbb\xbf") {
rewind($fh);
}
while(($row = fgetcsv($fh, 10000, ',')) !== false) {
// Your code here
}
This basically reads 3 bytes and checks if they match
https://en.wikipedia.org/wiki/Byte_order_mark has more information if you are dealing with other code-pages
This part of the length parameter behavior description is tricky, because it's not mentioning that separator is considered as a char and converted into an empty string: "Otherwise the line is split in chunks of length characters (...)".
First, take a look at the example of reading a line which does't contain separators:
<?php
file_put_contents('data.csv', 'foo'); // no separator
$handle = fopen('data.csv', 'c+');
$data = fgetcsv($handle, 2);
var_dump($data);
?>
Example above will output:
array(1) {
[0]=>
string(2) "fo"
}
Now let's add separators:
<?php
file_put_contents('data.csv', 'f,o,o'); // commas used as a separators
$handle = fopen('data.csv', 'c+');
$data = fgetcsv($handle, 2);
var_dump($data);
?>
Second example will output:
array(2) {
[0]=>
string(1) "f"
[1]=>
string(0) ""
}
Now let's alter the length:
<?php
file_put_contents('data.csv', 'f,o,o');
$handle = fopen('data.csv', 'c+');
$data = fgetcsv($handle, 3); // notice updated length
var_dump($data);
?>
Output of the last example is:
array(2) {
[0]=>
string(1) "f"
[1]=>
string(1) "o"
}
The final conclusion is that while splitting line in chunks, separator is considered as a char during the read but then it's being converted into empty string. What's more, if separator is at the very first or last position of a chunk it will be included in the result array, but if it's somewhere between other chars, then it will be just ignored.
There is a special syntax that prevents Excel from automagically convert field contents into dates or floating point numbers: ="fieldcontent" (equation symbol at the beginning). (Mind you, this is not to be used if content has line-end character of field-separator character inside.)
Now this syntax is not supported by fgetcvs, though it can be implemented with some post-processing.