[PHP] Unique ID on 5.0.4 (Windows)
Hi, Pretty soon I'm going to be needing to generate a unique identifier in a script. I'm looking into how to go about doing it now. It has to work on Apache 2 / PHP 5.0.4 (Module) / Windows 2000 Server. Any suggestions on how I might be able to do it? I've noticed the uuid PECL package, which looks like it could work for me, but I'm not sure it will work on Windows, or where to get it if it did. I have access to Visual Studio 6, so I think I could compile it if I had to. Anyone have experience compiling PECL extensions on Windows? I'd appreciate any help you could give me. Thanks, Chris -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID on 5.0.4 (Windows)
Hi Chris, Chris wrote: Pretty soon I'm going to be needing to generate a unique identifier in a script. I'm looking into how to go about doing it now. It has to work on Apache 2 / PHP 5.0.4 (Module) / Windows 2000 Server. Any suggestions on how I might be able to do it? I'd appreciate any help you could give me. What you need is called hash function (function for creating keys in hash tables). In PHP manual they are categorized under String functions... The way they work is that for any given string they return a (index) number (it can be shown a string of hex numbers) which is unique or almost unique (depending on the algorithm). The ones you may be interested are: md5, sha1, crc32 http://www.php.net/md5 http://www.php.net/sha1 http://www.php.net/crc32 Failing those, you may use mhash functions: http://www.php.net/manual/en/ref.mhash.php Hope this will help you a little, N:: smime.p7s Description: S/MIME Cryptographic Signature
Re: [PHP] Unique ID on 5.0.4 (Windows)
Dragan Stanojevic - Nevidljivi wrote: Hi Chris, Chris wrote: Pretty soon I'm going to be needing to generate a unique identifier in a script. I'm looking into how to go about doing it now. It has to work on Apache 2 / PHP 5.0.4 (Module) / Windows 2000 Server. Any suggestions on how I might be able to do it? I'd appreciate any help you could give me. What you need is called hash function (function for creating keys in hash tables). In PHP manual they are categorized under String functions... The way they work is that for any given string they return a (index) number (it can be shown a string of hex numbers) which is unique or almost unique (depending on the algorithm). The ones you may be interested are: md5, sha1, crc32 http://www.php.net/md5 http://www.php.net/sha1 http://www.php.net/crc32 Failing those, you may use mhash functions: http://www.php.net/manual/en/ref.mhash.php Hope this will help you a little, N:: I think you may have misunderstood me, sorry. I need to generate a unique string, to use as an identifier, in a script, not something to identify the script itself. A hash won't work for that. Thanks anyway though. Chris -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Unique ID system - need help/ideas
I was given a database that has uniques IDs such as: AL1 AL2 AR1 AR2 MI1 etc... The first two letters are based on the state they live in, then it would need to look at the next 5 characters to determine the next ID. My problem is how do I recreate this in PHP to make sure any newly added records follow this same unique ID? So if I insert another Alabama record the id would be: AL3 Or insert a record from a new state would be: NY1 (so NY0 is the default starting point) Any ideas? Thank you in advance! __ Do you Yahoo!? Yahoo! Search - Find what youre looking for faster http://search.yahoo.com -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
J J mailto:[EMAIL PROTECTED] on Wednesday, March 10, 2004 3:46 PM said: Any ideas? Thank you in advance! does it all have to be in the same field? can you not just combine two fields? one field would be called 'id' and it would be auto-incrementing. second field would be 'state' and it would contain the two letter abbreviation. i can think of a few other ways to get it all into one field, but they're a bit less efficient. then again i'm no sql expert. chris. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
I'd like to keep it in one field if possible since that's how their database is now and data will be going back and forth. Plus there should be allowed two of the same numbers, so like: AL3 KY3 So any state with more than one record can be sequential and not jump around like KY3, AL4, KY5, etc. Make sense? --- Chris W. Parker [EMAIL PROTECTED] wrote: J J mailto:[EMAIL PROTECTED] on Wednesday, March 10, 2004 3:46 PM said: Any ideas? Thank you in advance! does it all have to be in the same field? can you not just combine two fields? one field would be called 'id' and it would be auto-incrementing. second field would be 'state' and it would contain the two letter abbreviation. i can think of a few other ways to get it all into one field, but they're a bit less efficient. then again i'm no sql expert. chris. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php __ Do you Yahoo!? Yahoo! Search - Find what youre looking for faster http://search.yahoo.com -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
J J mailto:[EMAIL PROTECTED] on Wednesday, March 10, 2004 4:03 PM said: I'd like to keep it in one field if possible since that's how their database is now and data will be going back and forth. [snip] So any state with more than one record can be sequential and not jump around like KY3, AL4, KY5, etc. Make sense? yeah that makes sense. hmm.. well i'll give you my idea.. but it's very possible there's a better way to do it (than what i'm going to suggest). ok. one field will be used. it'll be a text (varchar) field. before you do an insert you'll need to find out what's in there ahead of time and then adjust accordingly. i'd select the last record that contained the two letter abbreviation, find out what the number was, increment it, then reconstruct the id and insert the new one. this of course won't prevent the exact same process from happening at the exact same time. i mean, two sessions could pull the same id and then reconstruct the (new, but) same id. make sense? chris. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
Will it be able to figure out what the last number is if it's a character type and the number is something like 3? So it'll add one to make it 4 and not something like 10003. ?? --- Chris W. Parker [EMAIL PROTECTED] wrote: yeah that makes sense. hmm.. well i'll give you my idea.. but it's very possible there's a better way to do it (than what i'm going to suggest). ok. one field will be used. it'll be a text (varchar) field. before you do an insert you'll need to find out what's in there ahead of time and then adjust accordingly. i'd select the last record that contained the two letter abbreviation, find out what the number was, increment it, then reconstruct the id and insert the new one. this of course won't prevent the exact same process from happening at the exact same time. i mean, two sessions could pull the same id and then reconstruct the (new, but) same id. make sense? chris. __ Do you Yahoo!? Yahoo! Search - Find what youre looking for faster http://search.yahoo.com -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
J J mailto:[EMAIL PROTECTED] on Wednesday, March 10, 2004 4:22 PM said: Will it be able to figure out what the last number is if it's a character type and the number is something like 3? So it'll add one to make it 4 and not something like 10003. php knows. you'll just have to pad the string with zeros. ?php $number = 01; ++$number; echo $number; ? that will output 2 not 101. chris. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
At 04:10 PM 3/10/2004 -0800, Chris W. Parker wrote: J J mailto:[EMAIL PROTECTED] on Wednesday, March 10, 2004 4:03 PM said: I'd like to keep it in one field if possible since that's how their database is now and data will be going back and forth. [snip] So any state with more than one record can be sequential and not jump around like KY3, AL4, KY5, etc. Make sense? yeah that makes sense. hmm.. well i'll give you my idea.. but it's very possible there's a better way to do it (than what i'm going to suggest). ok. one field will be used. it'll be a text (varchar) field. before you do an insert you'll need to find out what's in there ahead of time and then adjust accordingly. i'd select the last record that contained the two letter abbreviation, find out what the number was, increment it, then reconstruct the id and insert the new one. this of course won't prevent the exact same process from happening at the exact same time. i mean, two sessions could pull the same id and then reconstruct the (new, but) same id. make sense? chris. Yes Chris, that's the right idea. The problem is that this is being done across the 'Net. Ideally you try and get a lock on the row for a given state, fetch the value and increment/update, then release the lock. In a web environment locks are a little tricky - what happens if a connection is lost or times out? For the original poster: Having said all this, make certain that there is a unique, system-generated, primary key for each table. As these keys *never* have to be seen by the public, don't get tampered with, etc., they can be safely relied on for inter-table relationships. Down the road they will save your bacon. If the purpose of having separate numbering sequences for each state is to keep track of a count, why bother? Just select for a count on each state. If it's a matter of ego, in that if I have a lower registration number I have higher status, well fill your boots with whatever scheme will work. Really look at this v. closely. Quite often clients insist on wacko numbering schemes which they are convinced are important, but frequently result only because that's the way it's always been done. Also remember there should be no data encoding within a field - that's why so many columns are possible. Example, membership numbers like MABT082003BM2, where the first four characters are my initials, the next six the date I joined, and each of the next the colour of my eyes, marital status and number of children, ARE FORBIDDEN. All that information belongs in separate fields. This is what you are tending towards with AL0003 and KY00107. Bad practice. As I'm Canadian, if I stepped on any toes I'll apologize in advance. g Cheers - Miles -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID system - need help/ideas
Yeah thanks for the feedback, I'm already including a primary key auto_increment field in place to tie the tables together with an ID I *know* will be unique. This other ID scheme is just to work within their existing system that uses this. I'm just trying to mirror/maintain that scheme for their purpose. --- Miles Thompson [EMAIL PROTECTED] wrote: For the original poster: Having said all this, make certain that there is a unique, system-generated, primary key for each table. As these keys *never* have to be seen by the public, don't get tampered with, etc., they can be safely relied on for inter-table relationships. Down the road they will save your bacon. If the purpose of having separate numbering sequences for each state is to keep track of a count, why bother? Just select for a count on each state. If it's a matter of ego, in that if I have a lower registration number I have higher status, well fill your boots with whatever scheme will work. Really look at this v. closely. Quite often clients insist on wacko numbering schemes which they are convinced are important, but frequently result only because that's the way it's always been done. Also remember there should be no data encoding within a field - that's why so many columns are possible. Example, membership numbers like MABT082003BM2, where the first four characters are my initials, the next six the date I joined, and each of the next the colour of my eyes, marital status and number of children, ARE FORBIDDEN. All that information belongs in separate fields. This is what you are tending towards with AL0003 and KY00107. Bad practice. As I'm Canadian, if I stepped on any toes I'll apologize in advance. g Cheers - Miles -- __ Do you Yahoo!? Yahoo! Search - Find what youre looking for faster http://search.yahoo.com -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID system - need help/ideas
On 10 Mar 2004 J J wrote:
My problem is how do I recreate this in PHP to make
sure any newly added records follow this same unique
ID?
I haven't played much with these functions but I think for MySQL you'd
want something like this, for the state XX (excuse the wrapped
lines):
if ($result =
mysql_query(select max(id) from table where left(id, 2) = 'XX')) {
$row = mysql_fetch_row($result);
$newid = strpad($row[0] + 1, 5, 0, STR_PAD_LEFT);
mysql_query(insert into table set id = 'XX' . $newid);
$newrecnum = mysql_insert_id();
}
That is crude and uuntested, needs some error checking and most likely
table locking, but it's one possibility for a base.
--
Tom
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID system - need help/ideas
Wow I think you are on to something!!! I'll give it a
whirl!
Thanks!
--- [EMAIL PROTECTED] wrote:
On 10 Mar 2004 J J wrote:
My problem is how do I recreate this in PHP to
make
sure any newly added records follow this same
unique
ID?
I haven't played much with these functions but I
think for MySQL you'd
want something like this, for the state XX (excuse
the wrapped
lines):
if ($result =
mysql_query(select max(id) from table where
left(id, 2) = 'XX')) {
$row = mysql_fetch_row($result);
$newid = strpad($row[0] + 1, 5, 0,
STR_PAD_LEFT);
mysql_query(insert into table set id = 'XX' .
$newid);
$newrecnum = mysql_insert_id();
}
That is crude and uuntested, needs some error
checking and most likely
table locking, but it's one possibility for a base.
--
Tom
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
__
Do you Yahoo!?
Yahoo! Search - Find what youre looking for faster
http://search.yahoo.com
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Unique ID - again
Hi folks, I'm using usual md5(microtime()); to create IDs, but now I've encountered a problem with that. I need to create explicitly 6 digit unique number(decimal). Yes, I know I can use for/while loop to fill a string with digits, but is it ABSOLUTELY sure that the random will never return same number when seed is microtime()? It is very important as the number will identify a bank transfer... Is there any other way than checking with all previous IDs? Or some MySQL function to do this for me? thank you Alex -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID - again
Hello Alex, Thursday, February 12, 2004, 11:00:41 AM, you wrote: A I'm using usual md5(microtime()); to create IDs, but now I've encountered a A problem with that. I need to create explicitly 6 digit unique A number(decimal). Yes, I know I can use for/while loop to fill a string with A digits, but is it ABSOLUTELY sure that the random will never return same A number when seed is microtime()? It is very important as the number will A identify a bank transfer... If the data is so important - why are you restricting it to a 6 digit key? You should ideally use the uniqid() function perhaps with a combined lcg entropy at the end. That will give you a 13 character long unique id - or md5 it for a 32 character one (see the manual for examples). A Is there any other way than checking with all previous IDs? Or some MySQL A function to do this for me? For something so important I would use (a) a longer unique ID and (b) I'd still run a SQL check to see if the ID has been already used or not. A simple MySQL count will bring that result back very quickly. -- Best regards, Richardmailto:[EMAIL PROTECTED] -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID - again
MySQL (or another SQL database) is the answer. You can check for an existing OR use it's auto_increment or whatever else you like. Create a field to track all this stuff. There is no other way to guarantee that an ID is unique than to check it against existing. Random functions yield random results, so each time, there is ALWAYS a chance of a number being repeated. Microtime works pretty well, but there is STILL a chance two people could have a transaction at the same time. Yes, the odds can be insanely, insanely small for various functions (i.e. MD5) but as you shorten the length of your number, the odds get higher. And yes, it's important to realize that a 6 digit unique identifier is limited to 1 million possible combinations, not nearly enough for most banks. I might suggest you do some more PHP learning before you go and develop an online bank application - this stuff is pretty fundamental programming and database concepts. If your overall skill level isn't very high, you're likely to make mistakes, and either have errors, problems in functionality, or, worse, security problems. -Galen On Feb 12, 2004, at 3:00 AM, Alex wrote: Hi folks, I'm using usual md5(microtime()); to create IDs, but now I've encountered a problem with that. I need to create explicitly 6 digit unique number(decimal). Yes, I know I can use for/while loop to fill a string with digits, but is it ABSOLUTELY sure that the random will never return same number when seed is microtime()? It is very important as the number will identify a bank transfer... Is there any other way than checking with all previous IDs? Or some MySQL function to do this for me? thank you Alex -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID
Javier Muniz wrote: Neither auto-increment nor sequences allow you to do what you're asking for though, which is like database garbage collection. If you really need something like this you should implement it in a cron job that runs during non-peak hours as it will be quite an expensive operation on a large table (the best way I can think of would be to iterate over all table rows and look for gaps in the table, and for each gap take the record with the highest ID and change the ID to fill the gap, then update all replys to match the new ID, after everything is finished reset the sequence to the highest id in the modified table and you're done). I would not care about removing old posts. Good archive is always good :) I don't know what you mean that your database could get huge, in my understanding that means millions of post per week. Note from the MySQL manual for you auto_increment fans out there: Posted by Jim Martin on Tuesday October 1 2002, @11:57am Just in case there's any question, the AUTO_INCREMENT field /DOES NOT WRAP/. Once you hit the limit for the field size, INSERTs generate an error. (As per Jeremy Cole) That is I will be able to insert only 4294967295 rows if the auto_increment is UNSIGNED INT. :( I guess I'll start to use UNSIGNED BIGINT, that will give me 18446744073709551615 rows, that ought to be enough for anybody. ;) Marek -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Unique ID
Hi there, Here's my situation: I'm making a message board and I've decided the best way to go about the structure is to have 3 tables, two of them will be Posts and Replys. Now, in order for this ti work, each post has to have a UniqueID - same with the replys. Looking at another program, I can see that one way to do this is to do it by rows (ie. count how many rows, add 1, then that is the ID). It would be unique because no two rows would be 1 for example. The problem I can see is that the database would become incredibly huge (size wise I mean). I want to delete the posts after 30 days, and if I delete the row, then that would mess up the row system. Any suggestions? Dimitri Marshall -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID
Dimitri Marshall wrote: Hi there, Here's my situation: I'm making a message board and I've decided the best way to go about the structure is to have 3 tables, two of them will be Posts and Replys. Now, in order for this ti work, each post has to have a UniqueID - same with the replys. Looking at another program, I can see that one way to do this is to do it by rows (ie. count how many rows, add 1, then that is the ID). It would be unique because no two rows would be 1 for example. The problem I can see is that the database would become incredibly huge (size wise I mean). I want to delete the posts after 30 days, and if I delete the row, then that would mess up the row system. Any suggestions? Have you heard about auto_increment? Read on: http://www.mysql.com/doc/en/example-AUTO_INCREMENT.html -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] Unique ID
auto_increment is not in the SQL standard, and not everyone uses MySQL. To be cross-compatible between databases use a Sequence: http://pear.php.net/manual/en/package.database.db.intro-sequences.php provides a brief introduction, the PEAR::DB class allows you to use sequences easily. Neither auto-increment nor sequences allow you to do what you're asking for though, which is like database garbage collection. If you really need something like this you should implement it in a cron job that runs during non-peak hours as it will be quite an expensive operation on a large table (the best way I can think of would be to iterate over all table rows and look for gaps in the table, and for each gap take the record with the highest ID and change the ID to fill the gap, then update all replys to match the new ID, after everything is finished reset the sequence to the highest id in the modified table and you're done). Note from the MySQL manual for you auto_increment fans out there: Posted by Jim Martin on Tuesday October 1 2002, @11:57am Just in case there's any question, the AUTO_INCREMENT field /DOES NOT WRAP/. Once you hit the limit for the field size, INSERTs generate an error. (As per Jeremy Cole) And the comment a few beyond that regarding being able to drop the index and renumber doesn't help here, as there is an exterior resource (the reply table) that will be broken as soon as the ids change. If anyone has a better suggestion on how to do this I'd love to hear it :) -Javier -Original Message- From: Marek Kilimajer [mailto:[EMAIL PROTECTED] Sent: Tuesday, November 11, 2003 3:30 PM To: Dimitri Marshall Cc: [EMAIL PROTECTED] Subject: Re: [PHP] Unique ID Dimitri Marshall wrote: Hi there, Here's my situation: I'm making a message board and I've decided the best way to go about the structure is to have 3 tables, two of them will be Posts and Replys. Now, in order for this ti work, each post has to have a UniqueID - same with the replys. Looking at another program, I can see that one way to do this is to do it by rows (ie. count how many rows, add 1, then that is the ID). It would be unique because no two rows would be 1 for example. The problem I can see is that the database would become incredibly huge (size wise I mean). I want to delete the posts after 30 days, and if I delete the row, then that would mess up the row system. Any suggestions? Have you heard about auto_increment? Read on: http://www.mysql.com/doc/en/example- AUTO_INCREMENT.html -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID
Dimitri Marshall wrote: Hi there, Here's my situation: I'm making a message board and I've decided the best way to go about the structure is to have 3 tables, two of them will be Posts and Replys. Now, in order for this ti work, each post has to have a UniqueID - same with the replys. Looking at another program, I can see that one way to do this is to do it by rows (ie. count how many rows, add 1, then that is the ID). It would be unique because no two rows would be 1 for example. The problem I can see is that the database would become incredibly huge (size wise I mean). I want to delete the posts after 30 days, and if I delete the row, then that would mess up the row system. Any suggestions? Dimitri Marshall A standard solution in this case is to use one common table for all messages, both 'posts' and 'replays', distinguishing them by 'type' (TypeID, say 1 for post, 2 for replay), and then build a parent/child relationship between both type of messages in that table. In this way each message will have unique id. When you need all or some 'post' messages, use a query for selecting TypeID = 1, resp. TypeID = 2 for 'replay' messages. Concerning the deletion of a message (post or replay), you should write a query which will recursively delete the message selected as well as all child messages assigned to that (parent) message. The easiest way to do this is to use triggers, but that depends on the database you work with. To get closer to this approach and to find some nice examples, check out the following resources: http://www.sqlmag.com/Articles/Index.cfm?ArticleID=8826 http://www.sqlteam.com/item.asp?ItemID=8866 http://msdn.microsoft.com/library/default.asp?url=/library/en-us/acdata/ac_8_qd_14_5yk3.asp http://www.yafla.com/papers/sqlhierarchies/sqlhierarchies.htm HTH, Boyan -- -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] Unique ID
Marek Kilimajer wrote: Dimitri Marshall wrote: I'm making a message board and I've decided the best way to go about the structure is to have 3 tables, two of them will be Posts and Replys. Now, in order for this ti work, each post has to have a UniqueID - same with the replys. Looking at another program, I can see that one way to do this is to do it by rows (ie. count how many rows, add 1, then that is the ID). It would be unique because no two rows would be 1 for example. The problem I can see is that the database would become incredibly huge (size wise I mean). I want to delete the posts after 30 days, and if I delete the row, then that would mess up the row system. Any suggestions? Have you heard about auto_increment? Read on: http://www.mysql.com/doc/en/example-AUTO_INCREMENT.html Keep reading about database design, too. You'll only need one table and a parent-child relationship. The initial post will have a parent of zero and child posts under it will have a parent column relating back to the original post. I may not have explained that the best way, but there are plenty of articles out there about it. Have fun. :) -- ---John Holmes... Amazon Wishlist: www.amazon.com/o/registry/3BEXC84AB3A5E/ php|architect: The Magazine for PHP Professionals www.phparch.com -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
On Sat, 2003-07-26 at 07:11, Curt Zirzow wrote: * Thus wrote jan ([EMAIL PROTECTED]): Hi there! What's the best way to create uids (unique ids) even when runnig at exactly same time (microseconds)? is this enough ??? $r = mt_rand(); $uid = uniqid(getmypid() . $r); If you're running a cluster of machines I would include information on the machine that is generating the uid. $uid = uniqid(getmypid() . $r . '_MACHINE_NAME_'); Curt this is similar to what we do as well: hexadecimal timestamp + hexadecimal pid + hexadecimal randomvalue + hex ip address of server (only used if using multiple machines, otherwise this part is left off). I've also considered using semaphores in my id generation function to insure only one counter can be created at a time, but am worried that this might cause a bit of a performance hit. Anyone tried this? gabe. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
* Thus wrote Gabriel Guzman ([EMAIL PROTECTED]): On Sat, 2003-07-26 at 07:11, Curt Zirzow wrote: * Thus wrote jan ([EMAIL PROTECTED]): Hi there! What's the best way to create uids (unique ids) even when runnig at exactly same time (microseconds)? is this enough ??? $r = mt_rand(); $uid = uniqid(getmypid() . $r); If you're running a cluster of machines I would include information on the machine that is generating the uid. $uid = uniqid(getmypid() . $r . '_MACHINE_NAME_'); Curt this is similar to what we do as well: hexadecimal timestamp + hexadecimal pid + hexadecimal randomvalue + hex ip address of server (only used if using multiple machines, otherwise this part is left off). I've also considered using semaphores in my id generation function to insure only one counter can be created at a time, but am worried that this might cause a bit of a performance hit. Anyone tried this? The only other way I have done it was to make a central database with a table containing a auto incremented id. All scripts would add a record and get back the auto incremented id number. Since db inserts are atomic, it is guaranteed to be unique. I think the biggest weak point in this approach is if the connection goes down between the web server and db server, the script is left clueless as what to do. Curt -- I used to think I was indecisive, but now I'm not so sure. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] unique id
Hi there! What's the best way to create uids (unique ids) even when runnig at exactly same time (microseconds)? is this enough ??? $r = mt_rand(); $uid = uniqid(getmypid() . $r); thanks Jan -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
Hi, Saturday, July 26, 2003, 5:58:41 PM, you wrote: j Hi there! j What's the best way to create uids (unique ids) even when runnig at j exactly same time (microseconds)? j is this enough ??? j $r = mt_rand(); j $uid = uniqid(getmypid() . $r); j thanks j Jan I use: $id = md5(uniqid(rand(),1)); -- regards, Tom -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
* Thus wrote jan ([EMAIL PROTECTED]): Hi there! What's the best way to create uids (unique ids) even when runnig at exactly same time (microseconds)? is this enough ??? $r = mt_rand(); $uid = uniqid(getmypid() . $r); If you're running a cluster of machines I would include information on the machine that is generating the uid. $uid = uniqid(getmypid() . $r . '_MACHINE_NAME_'); Curt -- I used to think I was indecisive, but now I'm not so sure. -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] unique id
How do I generate a uniqueid in php? Thanks, Eddie -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
How do I generate a uniqueid in php? www.php.net/uniqid ---John Holmes... -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] unique id
*WHY* not use auto_increment?
Would the builtin http://php.net/uniqid function not work better?
*NEVER* seed the random number generator more than once per script. Call
srand() once, and only once.
The return() statement will return the MySQL resource ID, which is probably
not useful, but all the other variables will be forgotten so I think
whatever pointer and loop position 0 you mean will be reset-- assuming all
this is in a function.
You are limiting yourself to 9 - 1 records, which is pretty small.
Basically, you almost for sure are making your life difficult and should
just use auto_increment.
- Original Message -
From: AJDIN BRANDIC [EMAIL PROTECTED]
Newsgroups: php.general
Sent: Tuesday, January 09, 2001 11:14 AM
Subject: [PHP] unique id
Hi again,
I am generating a randum number but before inserting it into a table I
want to make sure that onother number with the same value doesn't exist
in the table. Here is what I have:
srand((double)microtime()*100);
$ran_num = rand(1, 9);
$query01 = "SELECT thisnum FROM mytable";
$results = mysql_query($query01) or die("Failed on select
mytable!");
while ($result = mysql_fetch_array($results)) {
$thisnum=$result['thisnum'];
if($ran_num==$thisnum) {
srand((double)microtime()*100);
$ran_num = rand(1, 9);
reset($results)
// will this reset the pointer and will the
while loop start from position 0??? If not how do I force the check to
start from position 0?
}
}
I donot want to use autoincrement???
Thanks
Ajdin
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
