This is not the problem.
The query fails when there is, as examples, an invalid syntax in the query
or the table is not found.
However, if the query executes but updates nothing. How can I know when the
update has changed a row or not? Not changing a row does not return a
failure.
Matt's solutions works well. Since the application is hosted and I have
limited choices... the mysql_info() does the trick.
- Glenn
On 11/26/05, Rich Hutchins <rhutchin@xxxxxxxxxxxxxxxx> wrote:
>
> Please forgive me as I've been working with C# recently so my PHP is a
> little rusty. But if I'm reading this correctly...
>
> If your query fails outright, your function is going to return false:
>
> "if (!(@ mysql_query($query, $connection))) {
> $this->errors = array(mysql_errno(), mysql_error());
> $error = new Error();
> $error->logError($this->errors);
> return false;
> }"
>
> At that point, the script will exit and you'll see the error information
> in
> your error log. Because the function/script will exit when it encounters a
> return statement, you wouldn't ever get to the line of code with the
> mysql_affected_rows() function in it if your SQL did, indeed, error out.
> It
> would write to the error log, return false and exit.
>
> If the query was successful, but no data was changed, all you're returning
> is false. That, as you said, doesn't help much. Plus, it's the same result
> you're returning for a failed query. Pretty ambiguous. Instead of
> returning
> false from these functions, maybe returning some sort of status instead
> would help. For example, if the query fails due to an error write to the
> error log as you're currently doing, but return -1 instead of false. If
> the
> query succeeds, but no data was changed, return a 0 instead of false since
> false isn't exactly what the result would mean, but 0 is a little closer
> to
> what you really mean. If the query was successful, you could still return
> true or you could return the number of rows affected if you care to have
> that information.
>
> In the calling code, you would have a function to receive the statuses
> above
> and display more meaningful information. For example,
>
> (pseudocode, of course)
>
> if($queryresult == -1){
> $msg = "The query failed. Check the error log.";
> }elseif($queryresult == 0){
> $msg = "The query was successful, but no rows were altered."
> }else{
> $msg = "The query was successful. ".$queryresult." rows were
> altered.";
> }
> echo $msg;
>
> Hope this helps.
>
> Rich
>
> -----Original Message-----
> From: maestro [mailto:maestroQC@xxxxxxxxx]
> Sent: Friday, November 25, 2005 1:11 PM
> To: php-db@xxxxxxxxxxxxx
> Subject: MySQL Update Quandry
>
>
> Greetings,
>
> I am doing an update to a table as such:
>
> class DB {
> ....
>
> function updateMember($email, $password, $postalCode,
> $language, $id, $word) {
>
> ...
> if (!(@ mysql_query($query, $connection))) {
> $this->errors = array(mysql_errno(), mysql_error());
> $error = new Error();
> $error->logError($this->errors);
> return false;
> }
>
> if (mysql_affected_rows() == 0) {
> return false;
> }
>
> return true;
> }
> }
>
>
>
> If the user supplies a wrong $word the mysql_affected_rows will be 0 and
> the method will return false. This is fine.
> My problem arises when the user supplies data that has not changed. The
> mysql_affected_rows will be 0 as well.
> How can I know the difference between an update that fails due to a
> failed where clause and one that updates nothing?
>
> TIA,
> Glenn
>
> --
> PHP Database Mailing List (http://www.php.net/)
> To unsubscribe, visit: http://www.php.net/unsub.php
>
>
