PHP 8.3.0 RC 6 available for testing

mysql_fetch_array

(PHP 4, PHP 5)

mysql_fetch_array 从结果集中取得一行作为关联数组

说明

mysql_fetch_array(resource $result, int $result_type = ?): array

mysql_fetch_array()mysql_fetch_row() 的扩展版本。除了将数据以数字索引方式储存在数组中之外,还可以将数据作为关联索引储存,用字段名作为键名。

有一点很重要必须指出,用 mysql_fetch_array()不明显 比用 mysql_fetch_row() 慢,而且还提供了明显更多的值。

参数

result

resource 型的结果集。此结果集来自对 msql_query() 的调用。

result_type

接受以下常量值: MSQL_ASSOCMSQL_NUMMSQL_BOTH,默认为 MSQL_BOTH。如果用了 MYSQL_BOTH,将得到一个同时包含关联和数字索引的数组。用 MYSQL_ASSOC 只得到关联索引(如同 mysql_fetch_assoc() 那样),用 MYSQL_NUM 只得到数字索引(如同 mysql_fetch_row() 那样)。

返回值

返回根据从结果集取得的行生成的数组,如果没有更多行则返回 false

示例

示例 #1 msql_fetch_array() 示例

<?php
$con
= msql_connect();
if (!
$con) {
die(
'Server connection problem: ' . msql_error());
}

if (!
msql_select_db('test', $con)) {
die(
'Database connection problem: ' . msql_error());
}

$result = msql_query('SELECT id, name FROM people', $con);
if (!
$result) {
die(
'Query execution problem: ' . msql_error());
}

while (
$row = msql_fetch_array($result, MSQL_ASSOC)) {
echo
$row['id'] . ': ' . $row['name'] . "\n";
}

msql_free_result($result);
?>

参见

  • msql_fetch_row()
  • msql_fetch_object()
  • msql_data_seek()
  • msql_result()

add a note

User Contributed Notes 31 notes

up
14
robjohnson at black-hole dot com
21 years ago
Benchmark on a table with 38567 rows:

mysql_fetch_array
MYSQL_BOTH: 6.01940000057 secs
MYSQL_NUM: 3.22173595428 secs
MYSQL_ASSOC: 3.92950594425 secs

mysql_fetch_row: 2.35096800327 secs
mysql_fetch_assoc: 2.92349803448 secs

As you can see, it's twice as effecient to fetch either an array or a hash, rather than getting both. it's even faster to use fetch_row rather than passing fetch_array MYSQL_NUM, or fetch_assoc rather than fetch_array MYSQL_ASSOC. Don't fetch BOTH unless you really need them, and most of the time you don't.
up
10
KingIsulgard
14 years ago
I have found a way to put all results from the select query in an array in one line.

// Read records
$result = mysql_query("SELECT * FROM table;") or die(mysql_error());

// Put them in array
for($i = 0; $array[$i] = mysql_fetch_assoc($result); $i++) ;

// Delete last empty one
array_pop($array);

You need to delete the last one because this will always be empty.

By this you can easily read the entire table to an array and preserve the keys of the table columns. Very handy.
up
-2
puzbie at facebookanswers dot co dot uk
12 years ago
<?php
while($r[]=mysql_fetch_array($sql));
?>

Yes, that will generate a dummy array element containing the false of the final mysql_fetch_array. You should either truncate the array or (more sensibly in my mind) check that the result of mysql_fetch_array is not false before adding it to the array.
up
-3
mehdi dot haresi at gmail dot com
14 years ago
For all of you having problems accessing duplicated field names in queries with their table alias i have implemented the following quick solution:

<?php
function mysql_fetch_alias_array($result)
{
if (!(
$row = mysql_fetch_array($result)))
{
return
null;
}

$assoc = Array();
$rowCount = mysql_num_fields($result);

for (
$idx = 0; $idx < $rowCount; $idx++)
{
$table = mysql_field_table($result, $idx);
$field = mysql_field_name($result, $idx);
$assoc["$table.$field"] = $row[$idx];
}

return
$assoc;
}
?>

Lets asume we have 2 tables student and contact each having fID as the index field and want to access both fID fields in php.

The usage of this function will be pretty similar to calling mysql_fetch_array:

<?php
$result
= mysql_query("select * from student s inner join contact c on c.fID = s.frContactID");

while (
$row = mysql_fetch_alias_array($result))
{
echo
"StudenID: {$row['s.fID']}, ContactID: {$row['c.fID']}";
}
?>

Voila, that's it :)

Please be aware that by using this function, you have to access all fields with their alias name (e.g. s.Name, s.Birhtday) even if they are not duplicated.

If you have questions, just send me a mail.

Best regards,
Mehdi Haresi
die-webdesigner.at
up
-8
maileremi at gmail dot com
12 years ago
Simple way to put table in an array...

<?php
//= Query ========================//
$sql=mysql_query("select * from table1");

//= Closed while ====================//
/*everytime it fetches the row, adds it to array...*/
while($r[]=mysql_fetch_array($sql));

echo
"<pre>";
//= Prints $r as array =================//
print_r ($r);
//=============================//
echo "</pre>";
?>
up
-4
john at skem9 dot com
17 years ago
my main purpose was to show the fetched array into a table, showing the results side by side instead of underneath each other, and heres what I've come up with.

just change the $display number to however many columns you would like to have, just dont change the $cols number or you might run into some problems.

<?php
$display
= 4;
$cols = 0;
echo
"<table>";
while(
$fetched = mysql_fetch_array($result)){
if(
$cols == 0){
echo
"<tr>\n";
}
// put what you would like to display within each cell here
echo "<td>".$fetched['id']."<br />".$fetched['name']."</td>\n";
$cols++;
if(
$cols == $display){
echo
"</tr>\n";
$cols = 0;
}
}
// added the following so it would display the correct html
if($cols != $display && $cols != 0){
$neededtds = $display - $cols;
for(
$i=0;$i<$neededtds;$i++){
echo
"<td></td>\n";
}
echo
"</tr></table>";
} else {
echo
"</table>";
}
?>

Hopefully this will save some of you a lot of searching.

any kind of improvements on this would be awesome!
up
-5
joey at clean dot q7 dot com
19 years ago
The issue of NULL fields seems to not be an issue anymore (as of 4.2.2 at least). mysql_fetch_* now seems to fully populate the array and put in entries with values of NULL when that is what the database returned. This is certainly the behaviour I expected, so I was concerned when i saw the notes here, but testing shows it does work the way I expected.
up
-4
Anonymous
12 years ago
If I use

<?php
while($r[]=mysql_fetch_array($sql));
?>

so in array $r is one more entry then rows returned from the database.
up
-5
noagbodjivictor at gmail dot com
16 years ago
Little improvement to the previous function.

function mysql_fetch_rowsarr($result, $numass=MYSQL_BOTH) {
$got = array();

if(mysql_num_rows($result) == 0)
return $got;

mysql_data_seek($result, 0);

while ($row = mysql_fetch_array($result, $numass)) {
array_push($got, $row);
}

return $got;
}
up
-4
info at o08 dot com
14 years ago
As opposite of mysql_fetch_array:

<?php
function mysql_insert_array ($my_table, $my_array) {
$keys = array_keys($my_array);
$values = array_values($my_array);
$sql = 'INSERT INTO ' . $my_table . '(' . implode(',', $keys) . ') VALUES ("' . implode('","', $values) . '")';
return(
mysql_query($sql));
}
#http://www.weberdev.com/get_example-4493.html
?>
up
-6
final at skilled dot ch
12 years ago
I ran into troubles with MySQL NULL values when I generated dynamic queries and then had to figure out whether my resultset contained a specific field.

First instict was to use isset() and is_null(), but these function will not behave as you probably expect.

I ended up using array_key_exists, as it was the only function that could tell me whether the key actually existed or not.

<?php
$row
= mysql_fetch_assoc(mysql_query("SELECT null as a"));
var_dump($row); //array(1) { ["a"]=> NULL }
var_dump(isset($row['a'])); //false
var_dump(isset($row['b'])); //false
var_dump(is_null($row['a'])); //true
var_dump(is_null($row['b'])); //true + throws undefined index notice
var_dump(array_key_exists('a', $row)); // true
var_dump(array_key_exists('b', $row)); // false
?>
up
-4
kunky at mail dot berlios dot de
18 years ago
This is very useful when the following query is used:

`SHOW TABLE STATUS`

Different versions of MySQL give different responses to this.

Therefore, it is better to use mysql_fetch_array() because the numeric references given my mysql_fetch_row() give very different results.
up
-4
joelwan at gmail dot com
18 years ago
Try Php Object Generator: http://www.phpobjectgenerator.com

It's kind of similar to Daogen, which was suggested in one of the comments above, but simpler and easier to use.

Php Object Generator generates the Php Classes for your Php Objects. It also provides the database class so you can focus on more important aspects of your project. Hope this helps.
up
-7
Ben
19 years ago
One of the most common mistakes that people make with this function, when using it multiple times in one script, is that they forget to use the mysql_data_seek() function to reset the internal data pointer.

When iterating through an array of MySQL results, e.g.

<?php
while ($line = mysql_fetch_array($result, MYSQL_ASSOC)) {
foreach (
$line as $col_value) {
echo
$col_value . '<br />';
}
}
?>

the internal data pointer for the array is advanced, incrementally, until there are no more elements left in the array. So, basically, if you copy/pasted the above code into a script TWICE, the second copy would not create any output. The reason is because the data pointer has been advanced to the end of the $line array and returned FALSE upon doing so.

If, for some reason, you wanted to interate through the array a second time, perhaps grabbing a different piece of data from the same result set, you would have to make sure you call

<?php
mysql_data_seek
($result, 0);
?>

This function resets the pointer and you can re-iterate through the $line array, again!
up
-6
hanskrentel at yahoo dot de
20 years ago
for the problem with fields containing null values in an associated array, feel free to use this function. i've got no more problems with it, just drop it in your script:

/*
* mysql_fetch_array_nullsafe
*
*
* get a result row as an enumerated and associated array
* ! nullsafe !
*
* parameter: $result
* $result: valid db result id
*
* returns: array | false (mysql:if there are any more rows)
*
*/
function mysql_fetch_array_nullsafe($result) {
$ret=array();

$num = mysql_num_fields($result);
if ($num==0) return $ret;

$fval = mysql_fetch_row ($result);
if ($fval === false) return false;

$i=0;
while($i<$num)
{
$fname[$i] = mysql_field_name($result,$i);
$ret[$i] = $fval[$i]; // enum
$ret[''.$fname[$i].''] = $fval[$i]; // assoc
$i++;
}

return $ret;
}
up
-7
romans at servidor dot unam dot mx
18 years ago
Regarding duplicated field names in queries, I wanted some way to retrieve rows without having to use alias, so I wrote this class that returns rows as 2d-arrays

<?
$field = $drow['table']['column'];
?>

Here is the code:

<?
class mysql_resultset
{
var $results, $map;

function mysql_resultset($results)
{
$this->results = $results;
$this->map = array();

$index = 0;
while ($column = mysql_fetch_field($results))
{
$this->map[$index++] = array($column->table, $column->name);
}
}

function fetch()
{
if ($row = mysql_fetch_row($this->results))
{
$drow = array();

foreach ($row as $index => $field)
{
list($table, $column) = $this->map[$index];
$drow[$table][$column] = $row[$index];
}

return $drow;
}
else
return false;
}
}
?>

The class is initialized with a mysql_query result:

<?
$resultset = new mysql_resultset(mysql_query($sql));
?>

The constructor builds an array that maps each field index to a ($table, $column) array so we can use mysql_fetch_row and access field values by index in the fetch() method. This method then uses the map to build up the 2d-array.

An example:

<?
$sql =
"select orders.*, clients.*, productos.* ".
"from orders, clients, products ".
"where join conditions";

$resultset = new mysql_resultset(mysql_query($sql));

while ($drow = $resultset->fetch())
{
echo 'No.: '.$drow['orders']['number'].'<br>';
echo 'Client: '.$drow['clients']['name'].'<br>';
echo 'Product: '.$drow['products']['name'].'<br>';
}
?>

I hope others find this useful as it has been to me.
up
-8
Tristan
14 years ago
Here's a quicker way to clone a record. Only 3 lines of code instead of 4. But the table must have an auto-incremented id.
I took the code from Tim and altered it. Props to Tim.

<?php
// copy content of the record you wish to clone
$entity = mysql_fetch_array(mysql_query("SELECT * FROM table_name WHERE id='$id_to_be cloned'"), MYSQL_ASSOC) or die("Could not select original record");

// set the auto-incremented id's value to blank. If you forget this step, nothing will work because we can't have two records with the same id
$entity["id"] = "";

// insert cloned copy of the original record
mysql_query("INSERT INTO table_name (".implode(", ",array_keys($entity)).") VALUES ('".implode("', '",array_values($entity))."')");

// if you want the auto-generated id of the new cloned record, do the following
$newid = mysql_insert_id();
?>

There you go.
up
-7
dkantha at yahoo dot com
21 years ago
I did find 'jb at stormvision's' code useful above, but instead of the number of rows you need the number of fields; otherwise you get an error.

So, it should read like the following:

$result=mysql_query("select * from mydata order by 'id'")or die('died');
$num_fields = mysql_num_fields($result);
$j=0;
$x=1;
while($row=mysql_fetch_array($result)){
for($j=0;$j<$num_fields;$j++){
$name = mysql_field_name($result, $j);
$object[$x][$name]=$row[$name];
}$x++;
}

For Later in the script you may use the below array to gain access to your data

$i=1;
$ii=count($object); //quick access function
for($i=1;$i<=$ii;$i++){
echo $object[$i]['your_field_name'];
}

I have tested this in my apps and it works great! :-)
up
-7
buddy at databoysoftware dot com
14 years ago
In the note entered by Typer85, concerning the use of mysql_data_seek(), it should be noted that there are two parameters, both of which are required.

If you have already iterated through a result array (for instance, using mysql_fetch_array()), and have a need to start from the top, the proper syntax is:

mysql_data_seek({result set},{record#})

EG:
mysql_data_seek($result,0)
("0" represents the first record in the array.)

This will reset your result to the top of the array so that you can then re-process with
while($row = mysql_fetch_array($result)) or other array processing.
up
-6
eddie at nailchipper dot com
18 years ago
mob AT stag DOT ru has a nice function for getting simple arrays from MySQL but it has a serious bug. The MySQL link being set as an argument is NULL when no link is supplied meaning that you're passing NULL to the mysql funcctions as a link, which is wrong. I am not using multitple connections so I removed the link and using the global link. If you want to support multiple links check to see if its set first.

/*
* to support multiple links add the $link argument to function then
* test it before you use the link
*
* if(isset($link))
* if($err=mysql_errno($link))return $err;
* else
* if($err=mysql_errno())return $err;
*/

function mysql_fetch_all($query){
$r=@mysql_query($query);
if($err=mysql_errno())return $err;
if(@mysql_num_rows($r))
while($row=mysql_fetch_array($r))$result[]=$row;
return $result;
}
function mysql_fetch_one($query){
$r=@mysql_query($query);
if($err=mysql_errno())return $err;
if(@mysql_num_rows($r))
return mysql_fetch_array($r);
}
up
-5
barbieri at NOSPAMzero dot it
21 years ago
Here is a suggestion to workaround the problem of NULL values:

// get associative array, with NULL values set
$record = mysql_fetch_array($queryID,MYSQL_ASSOC);

// set number indices
if(is_array($record))
{
$i = 0;
foreach($record as $element)
$record[$i++] = $element;
}

This way you can access $result array as usual, having NULL fields set.
up
-5
tslukka at cc dot hut dot fi
21 years ago
If you think MySQL (or other) database
handling is difficult and requires lot's of
code, I recommend that you try http://titaniclinux.net/daogen/

DaoGen is a program source code generator
that supports PHP and Java. It makes database
programming quick and easy. Generated sources
are released under GPL.
up
-7
glenn dot hoeppner at yakhair dot com
21 years ago
Just another workaround for columns with duplicate names...

Modify your SQL to use the AS keyword.

Instead of:
$sql = "SELECT t1.cA, t2.cA FROM t1, t2 WHERE t1.cA = t2.cA";

Try:
$sql = "SELECT t1.cA AS foo1, t2.cA AS foo2 FROM t1, t2 WHERE t1.cA = t2.cA";

Then you can reference the results by name in the array:
$row[foo1], $row[foo2]
up
-5
tim at wiltshirewebs dot com
17 years ago
Here's a quick way to duplicate or clone a record to the same table using only 4 lines of code:

// first, get the highest id number, so we can calc the new id number for the dupe
// second, get the original entity
// third, increment the dupe record id to 1 over the max
// finally insert the new record - voila - 4 lines!

$id_max = mysql_result(mysql_query("SELECT MAX(id) FROM table_name"),0,0) or die("Could not execute query");
$entity = mysql_fetch_array(mysql_query("SELECT * FROM table." WHERE id='$id_original'),MYSQL_ASSOC) or die("Could not select original record"); // MYSQL_ASSOC forces a purely associative array and blocks twin key dupes, vitally, it brings the keys out so they can be used in line 4
$entity["id"]=$id_max+1;
mysql_query("INSERT INTO it_pages (".implode(", ",array_keys($Entity)).") VALUES ('".implode("', '",array_values($Entity))."')");

Really struggled in cracking this nut - maybe there's an easier way out there? Thanks to other posters for providing inspiration. Good luck - Tim
up
-7
Amine O
12 years ago
If you use implode() with the return value by mysql_fetch_array, if you use MYSQL_BOTH on parameter 2, the result is not really what you're expecting.
For example :
my sql database contains "Amine, Sarah, Mohamed";

$array = mysql_fetch_array($resource,MYSQL_BOTH);
or $array = mysql_fetch_array($resource);
echo implode(" - ", $array);

the result is : Amine-Amine-Sarah-Sarah-Mohamed-Mohamed
and we expect just : Amine-Sarah-Mohamed

You must use MYSQL_NUM or MYSQL_ASSOC on parameter 2 to resolve the problem.
up
-5
some at gamepoint dot net
21 years ago
I never had so much trouble with null fields but it's to my understanding that extract only works as expected when using an associative array only, which is the case with mysql_fetch_assoc() as used in the previous note.

However a mysql_fetch_array will return field values with both the numerical and associative keys, the numerical ones being those extract() can't handle very well.
You can prevent that by calling mysql_fetch_array($result,MYSQL_ASSOC) which will return the same result as mysql_fetch_assoc and is extract() friendly.
up
-6
mob AT stag DOT ru
18 years ago
I wrote some utility functions to improve usability and readability, and use them everywhere in my code. I suppose they can help.

function mysql_fetch_all($query,$MySQL=NULL){
$r=@mysql_query($query,$MySQL);
if($err=mysql_errno($MySQL))return $err;
if(@mysql_num_rows($r))
while($row=mysql_fetch_array($r))$result[]=$row;
return $result;
}
function mysql_fetch_one($query,$MySQL=NULL){
$r=@mysql_query($query,$MySQL);
if($err=mysql_errno($MySQL))return $err;
if(@mysql_num_rows($r))
return mysql_fetch_array($r);
}

Example use:
if(is_array($rows=mysql_fetch_all("select * from sometable",$MySQL))){
//do something
}else{
if(!is_null($rows)) die("Query failed!");
}
up
-7
juancri at tagnet dot org
21 years ago
An example with mysql_fetch_array():

$result = mysql_query("SELECT name FROM table WHERE id=8");
$array = mysql_fetch_array($result);

$array will be:

array ([0] => "John", ['name'] => "John")

Then you can access to the results:

echo "The name is " . $array[0];
// or
echo "The name is " . $array['name'];

But the array is not referential. $array[0] is not a reference to $array['name'] or $array['name'] to $array[0], they are not relationed between. Because of that, the system will use excesive memory. With large columns, try to use mysql_fetch_assoc() or mysql_fetch_row() only.
up
-8
mjm at porter dot appstate dot edu
21 years ago
If you perform a SELECT query which returns different columns with duplicate names, like this:

--------
$sql_statement = "SELECT tbl1.colA, tbl2.colA FROM tbl1 LEFT JOIN tbl2 ON tbl1.colC = tbl2.colC";

$result = mysql_query($sql_statement, $handle);

$row = mysql_fetch_array($result);
--------

Then

$row[0] is equivalent to $row["colA"]

but

$row[1] is not equivalent to $row["colA"].

Moral of the story: You must use the numerical index on the result row arrays if column names are not unique, even if they come from different tables within a JOIN. This would render mysql_fetch_assoc() useless.

[Ed. note - or you could do the usual 'select tbl1.colA as somename, tbl2.colA as someothername. . .' which would obviate the problem. -- Torben]
up
-10
scott_carney at hotmail dot com
11 years ago
Just thought I'd share these helper functions that I use to simplify processing of query results a bit:
<?php
// For a simple query that should return a single value, this returns just that value (or FALSE) so you can process it immediately
function db_result_single($result) {
return (
$row = mysql_fetch_row($result)) && isset($row[0]) ? $row[0] : false;
}

// Returns the rows as an array of rows
// Providing a key_column gives you access to specific rows (e.g. if (isset($result_array[$user_id])) ...)
function db_result_array($result, $key_column = null) {
for (
$array = array(); $row = mysql_fetch_assoc($result); isset($row[$key_column]) ? $array[$row[$key_column]] = $row : $array[] = $row);
return
$array;
}

// Returns an array of a single column of data that can optionally be keyed from second column (e.g. an array of user names keyed by user id)
function db_result_array_values($result) {
for (
$array = array(); $row = mysql_fetch_row($result); isset($row[1]) ? $array[$row[1]] = $row[0] : $array[] = $row[0]);
return
$array;
}
?>

Naturally, comments [to my email, not here] are welcome.
up
-9
Typer85 at gmail dot com
16 years ago
Please be advised that the resource result that you pass to this function can be thought of as being passed by reference because a resource is simply a pointer to a memory location.

Because of this, you can not loop through a resource result twice in the same script before resetting the pointer back to the start position.

For example:

----------------
<?php

// Assume We Already Queried Our Database.

// Loop Through Result Set.

while( $queryContent = mysql_fetch_row( $queryResult ) {

// Display.

echo $queryContent[ 0 ];
}

// We looped through the resource result already so the
// the pointer is no longer pointing at any rows.

// If we decide to loop through the same resource result
// again, the function will always return false because it
// will assume there are no more rows.

// So the following code, if executed after the previous code
// segment will not work.

while( $queryContent = mysql_fetch_row( $queryResult ) {

// Display.

echo $queryContent[ 0 ];
}

// Because $queryContent is now equal to FALSE, the loop
// will not be entered.

?>
----------------

The only solution to this is to reset the pointer to make it point at the first row again before the second code segment, so now the complete code will look as follows:

----------------
<?php

// Assume We Already Queried Our Database.

// Loop Through Result Set.

while( $queryContent = mysql_fetch_row( $queryResult ) {

// Display.

echo $queryContent[ 0 ];
}

// Reset Our Pointer.

mysql_data_seek( $queryResult );

// Loop Again.

while( $queryContent = mysql_fetch_row( $queryResult ) {

// Display.

echo $queryContent[ 0 ];
}

?>
----------------

Of course you would have to do extra checks to make sure that the number of rows in the result is not 0 or else mysql_data_seek itself will return false and an error will be raised.

Also please note that this applies to all functions that fetch result sets, including mysql_fetch_row, mysql_fetch_assos, and mysql_fetch_array.
To Top