how to get all the values of a column (not row) and put them in an array

marcnyc

Hello, I was hoping to get some help with this thing that got me stalled...
I am trying to get ALL the values contained in a column of a database (meaning I am trying to get all the values contained in a specific column of all the rows present in a table).
There is a mysql_fetch_row but there doesn't seem to be a mysql_fetch_col, so to speak.
My table rows have IDs but they are non-sequential (the IDs are pretty random).
How can I do this?

Thanks

To give you a little bit more background in case you need it to understand my question:
my table has a column called 'id' in which I store the id of an entry and one called 'addl_ids' in which I store the ids (separated by comma) of all the other entries that this one entry relates to (its a billing system with invoices tied to separate projects).
I want to grab all these additional ids from all the 'addl_ids' fields of all the rows of all the 'ids' so I can do a check on them. Hope this helps.
Thanks

Piranha

I think that you have to do this yourself. Just select the column, then step through the results and build an array from it.

marcnyc

is that really the only way?

the problem with that is that, unless I do two separate cycles (one before the main one to build the array), once I get to the row whose id contains additional ids, if those additional ids are lower in number then I am already past the part in the loop where I can do something with the entry whose id is the same of one of the additional ids... I hope i explained myself clearly...

laserlight

one called 'addl_ids' in which I store the ids (separated by comma) of all the other entries that this one entry relates to

This is your first mistake: the use of a multi-valued field. You need to change your database design. For example, you might create another table that represents the relationships between an entry and another entry.

For more information, read the article on Database Normalization and Design Techniques.

Piranha

What laserlight says is extremely important. Always noamalize to at least third normal form to avoid problems such as the one you currently have. Actually you should normalize to BCNF and 4NF as well, but normally it is in those forms when it is in 3NF.

marcnyc

I have no idea what you guys are talking about... I'll have to read that article...
I am a self-taught guy and programming is not my main job/activity so I can only put so much time into it...
are there any valuable examples of how this would be done?

if I wanted to avoid having to learn this whole new thing (because I want to finish this script and be done with it), is there a way I can cycle through all the rows in a table without knowing the IDs of the rows?
Thanks

laserlight

if I wanted to avoid having to learn this whole new thing (because I want to finish this script and be done with it), is there a way I can cycle through all the rows in a table without knowing the IDs of the rows?

I do not think so, since as soon as you know the ID (and its associated additional IDs) of a row, you must have retrieved it from the database. It seems like the two pass approach that you described is the only way, but if I read it right, this means that you have to have one more query than there are rows in the table. This is rather expensive, and very susceptible to race conditions.

marcnyc

are there any examples to do this table that represents the relationships between an entry and another entry?