OK, I need to get some data from multiple tables, all of which have a key.
Projects (key:id)
Projects_users (key:project_id,user_id)
Users (key:id)
Basically each project can have N users on it. I need to get all the users that are in each project in a formatted response.
For instance:
projects = {[
1: {
'title':'Project 1',
'users': [
1: { id: 23, name:'john' },
2: { id: 55, name:'sally' }
]
}
2: {
'title':'Project 2',
'users': [
1: { id: 41, name:'jeff' },
2: { id: 55, name:'sally' }
]
}
]}
Also, I understand I can do this in PHP with multiple MYSQL statements. Would that be faster and do the concatenation in PHP, or letting MYSQL do this be more efficient / quick.
You can use following statement to get the desired data:
SELECT `Projects`.*, `Users`.* FROM
`Projects` JOIN (
`Projects_users` JOIN
`Users`
ON `Projects_users`.`user_id` = `Users`.`id`
)
ON `Projects`.`id` = `Projects_users`.`project_id`
ORDER BY `Projects`.`id`
After that, you can traverse the resulting table in single pass and build the array you want.
I haven't tried this, but it should be close to what you're looking for:
select p.title, u.name
from projects p inner join users u
on p.id = u.id
order by p.title
You need the "join" to get the data .. but use PHP to format it.
PS:
"projects" is one table, "users" is a second table ... and the "projects" table needs to hold "foreign keys" for user IDs 23, 55 and 41.
MySql won't be able to return data in a tree-like structure like you want. You can do this in two ways:
Get all projects in one data set and all users associated to those projects on a second data set. You would have to write the code that generates the JSON objects you need.
Just get one data set doing an inner join of the project's table with the user's table. You will get repeated information on many columns but you will get everything in one sql statement. You could then iterate through the result set and generate the structure you need.
If you need help with the inner join, it would be something like:
select p.title, u.id, u.name
from projects p
inner join project_users pu on pu.project_id=p.id
inner join users u on u.id=pu.user_id
Related
My sql:
SELECT * FROM ex_pair
LEFT JOIN ex_currency as client
ON client_curr = currency_id
LEFT JOIN ex_currency as company
ON co_curr = currency_id
I need to get data for two currency_id but I have an error
ambiguous column name: 'currency_id'
Is there any way to do it right or i need to use two queries?
You need to include your alias in your join, like this:
SELECT *
FROM ex_pair
LEFT JOIN ex_currency AS client
ON client_curr = client.currency_id
LEFT JOIN ex_currency as company
ON co_curr = company.currency_id
You may also want to do something other than select * as you will have two tables with the same columns - something like
SELECT pair.*, company.currency_id AS company_currency_id, client.currency_id as client_currency_id, [...]
FROM ex_pair AS pair
[...]
This way when you explicitly declare the columns you intend to use from ex_currency with an alias, you can know on the other end more easily which are client and which are company. You will need to do this for each column in the currency table that you want in your result, though that can be done if you have your table structure in your code easily enough by looping over the list of columns and appending the alias.
$array = [
1=> "currency_id",
2=> "currency_name"
];
$columns = ""
foreach($array as $column){
$columns.= "company.".$column." AS company_".$column;
$columns.= ",client.".$column." AS client_".$column.",";
}
$columns = rtrim($columns,',');
Gives you
company.currency_id AS company_currency_id,client.currency_id AS client_currency_id,company.currency_name AS company_currency_name,client.currency_name AS client_currency_name
Add that after your SELECT pair.* and you get your columns from the currency table, aliased so you know which is which.
you can use the alias that you give to the tables:
SELECT client.currency_id as client_currency, company.currency_id as company_currency
FROM ex_pair
LEFT JOIN ex_currency as client ON client_curr = client.currency_id
LEFT JOIN ex_currency as company ON co_curr = company.currency_id
I am trying to learn about SQL joins and trying to apply them to an application I am building. I am doing a query to find a "game record" on a schedule based on a specific game id. But on this game record; for the "h_team" and the "v_team"; only the ids of the teams are on the game record. And so what I want to do is join the "teams" table and look up the two different team_names of the "h_team" and "v_team". I have it also pull in a "division name" as well using a join since only the division id is stored on the game record. I have gotten this all to work fine; except I do not know how to get the results separately for the "team_name" for h_team and v_team. Basically the key for each one is just "team_name"; I will paste in my code and then explain further:
$array_game_id6=32;
$sql = "SELECT * FROM playoff_schedule LEFT OUTER JOIN teams on playoff_schedule.h_team = teams.team_id || playoff_schedule.v_team = teams.team_id LEFT OUTER JOIN playoff_divisions on playoff_schedule.po_div_id = playoff_divisions.po_div_id WHERE tom_game_id=$array_game_id6";
foreach ($dbh->query($sql) as $resultsg39)
{
$h_team=$resultsg39[h_team];
$v_team=$resultsg39[v_team];
$po_div_id=$resultsg39[po_div_id];
$round=$resultsg39[round];
$game_id=$resultsg39[game_id];
$date=$resultsg39[date];
$timestamp=$resultsg39[timestamp];
$h_score=$resultsg39[h_score];
$v_score=$resultsg39[v_score];
$tom_game_id=$resultsg39[tom_game_id];
$h_name=$resultsg39[team_name];
$div_name=$resultsg39[playoff_name];
}
the problem comes in when i am trying to get the results of the query and store them all in the different variables…
the last two "$h_name" and "$div_name" are being pulled from the JOINs all the prior ones are on the game record itself…
what I want to do is store both the names from "v_team" and "h_team" in the respective variables $h_name and $v_name;
I have it storing the $h_name no problem; but i do not know how to make it store both $h_name and $v_name separately as they are both values in the column "team_name" from "teams" table. So I just need to somehow make it so when i get my results it can tell the difference between the two different "team_names" and I can store them in the two different variables…
If this is not clear please let me know.
Thanks!
***** UPDATE 10:49pm EST 2/5/2015
have made some progress on this but my query is not working; I think it is a problem with the aliases and such are not right; here is my non-working query as it is right now:
$sth = $dbh->prepare("SELECT home_team.team_name as home_team_name, visiting_team.team_name as visiting_team_name,
h_team, v_team, po_div_id, round, game_id, date, timestamp, h_score, v_score, tom_game_id, playoff_name FROM playoff_schedule
LEFT OUTER JOIN teams as home_team on playoff_schedule.h_team = teams.team_id
LEFT OUTER JOIN teams as visiting_team on playoff_schedule.v_team = teams.team_id
LEFT OUTER JOIN playoff_divisions on playoff_schedule.po_div_id = playoff_divisions.po_div_id
WHERE tom_game_id=$array_game_id6");
$sth->execute();
$article_list = $sth->fetchAll(PDO::FETCH_ASSOC);
foreach ($article_list as $row => $link) {
$h_team=$link['h_team'];
$v_team=$link['v_team'];
$po_div_id=$link['po_div_id'];
$round=$link['round'];
}
if anyone can spot a problem with my new query I would really appreciate it!
I think what you are trying to do is:
select home_team.team_name as home_team_name,
visiting_team.team_name as visiting_team_name
from playoff_schedule
join team as home_team on playoff_schedule.h_team = teams.team_id
join team as visiting_team on playoff_schedule.v_team = teams.team_id
You can join to the same table as many times as you want to. In this case, it makes sense, because you really are trying to get two different bits of information.
Based on your last edit, the following query appears to work:
SELECT home_team.team_name AS home_team_name,
visiting_team.team_name AS visiting_team_name,
h_team,
v_team,
playoff_schedule.po_div_id,
round,
game_id,
date,
timestamp,
h_score,
v_score,
tom_game_id,
playoff_name
FROM playoff_schedule
LEFT OUTER JOIN teams AS home_team
ON playoff_schedule.h_team = home_team.team_id
LEFT OUTER JOIN teams AS visiting_team
ON playoff_schedule.v_team = visiting_team.team_id
LEFT OUTER JOIN playoff_divisions
ON playoff_schedule.po_div_id = playoff_divisions.po_div_id
WHERE tom_game_id=$array_game_id6
You can check the query and the schema at: SQLFiddle
A couple of thing that might be happening:
Is the query itself running?
What happens if you run the query in a mySQL client?
Are there any PHP errors in your log?
Could you post the schema itself?
Is $array_game_id6 actually an array of values? In that case, you need to use "in" as opposed to "=" in your where clause.
With regard to your updated query, I think the main thing you are missing is using the aliases in your JOIN conditions. You should keep your table aliases consistent throughout your query. Also, IMO its better to keep table aliases short so they are easier to read:
So applying those things to your query:
SELECT h.team_name as h_team_name, v.team_name as v_team_name, s.h_team, s.v_team, s.po_div_id, s.round, s.game_id, s.date, s.timestamp, s.h_score, s.v_score, s.tom_game_id, s.playoff_name
FROM playoff_schedule s
LEFT OUTER JOIN teams h ON (
s.h_team = h.team_id
)
LEFT OUTER JOIN teams as v ON (
s.v_team = v.team_id
)
LEFT OUTER JOIN playoff_divisions d ON (
s.po_div_id = d.po_div_id
)
WHERE s.tom_game_id = ?
Now I'm not 100% sure of your schema so I may have referenced some of the columns to the wrong table but you should be able to sort that out.
I have a table structure with three tables: profiles, profile_subrubriek and rubrieken. I query the data with the following query:
SELECT profiles.hoofdrubriek, profiles.plaats
, profiles.bedrijfsnaam, profiles.gemeente, profiles.bedrijfsslogan
, profiles.straatnaam, profiles.huisnummer, profiles.postcode
, profiles.telefoonnummer, profiles.fax, profiles.email
, profiles.website, profiles.bedrijfslogo
FROM profiles INNER JOIN profile_subrubriek ON profiles.ID=profile_subrubriek.profile_id
INNER JOIN rubrieken ON profile_subrubriek.subrubriek_id=rubrieken.ID
where (
rubrieken.rubriek = 'Aannemersbedrijven'
OR
profiles.hoofdrubriek = 'Aannemersbedrijven')
AND profiles.gemeente = 'Dongen'
The result, 0 rows. That is not correct. If I take out the Inner Join and only incorporate the 'hoofdrubriek' column in the WHERE clausule I get about 25 rows as a result, that is correct. So this query (modified version of the above) does actually work:
SELECT profiles.hoofdrubriek, profiles.plaats, profiles.bedrijfsnaam
, profiles.gemeente, profiles.bedrijfsslogan, profiles.straatnaam
, profiles.huisnummer, profiles.postcode, profiles.telefoonnummer
, profiles.fax, profiles.email, profiles.website, profiles.bedrijfslogo
FROM profiles where (profiles.hoofdrubriek = 'Aannemersbedrijven')
AND profiles.gemeente = 'Dongen'
What am I doing wrong?
Thanks!
Probably the joined tables don't contain referenced values. Try LEFT JOIN instead of INNER JOIN.
Start troubleshooting with this query.
select count(*) records
FROM profiles INNER JOIN profile_subrubriek ON profiles.ID=profile_subrubriek.profile_id
If records is greater than 1, add this line and run it again:
INNER JOIN rubrieken ON profile_subrubriek.subrubriek_id=rubrieken.ID
Keep adding bits of your original query, one by one, until records is zero. The last thing you added will be the reason.
I have 5 mysql tables as described below.
clinics table
id
name
d_location_subscription table
id
clinic_id
t_id //t_id will contain a foreign key of d_cities, d_states or d_countries table
type "country" "state" "city"
d_countries table
id
name
code
d_states table
id
d_country_id
name
code
d_city table
id
d_state_id
name
code
d_location_subscription table is used to record clinic's subscription for a location(it may be a city, state or country). I'm expecting to get all subscribed cities for a specific
clinic using d_location_subscription table.
For example, if clinic A is subscribed to Texas state, I should be able to get all city ids for clinic A.
I created following sql query, it looks ugly but generate a close result what i want to achieve.
select
`d`.`id` AS `clinic_id`,
if((`dct`.`id` is not null),`dct`.`id`,if((`dct1`.`id` is not null),`dct1`.`id`,`dct2`.`id`)) AS `d_city_id`
from ((((((((
`d_location_subscriptions` `dls`
join `clinics` `d`
on((`d`.`id` = `dls`.`clinic_id`)))
left join `d_countries` `dc`
on(((`dc`.`id` = `dls`.`t_id`) and (`dls`.`type` = 'country'))))
left join `d_states` `ds`
on((`ds`.`d_country_id` = `dc`.`id`)))
left join `d_cities` `dct2`
on((`dct2`.`d_state_id` = `ds`.`id`)))
left join `d_states` `ds1`
on(((`ds1`.`id` = `dls`.`t_id`) and (`dls`.`type` = 'state'))))
left join `d_cities` `dct`
on((`dct`.`d_state_id` = `ds1`.`id`)))
left join `d_cities` `dct1`
on(((`dct1`.`id` = `dls`.`t_id`) and (`dls`.`type` = 'city'))))
)
when there is record with type "country" in d_location_subscription table, I receive following result. total number of records returned are equal to the number of d_states table records.
How should I get rid of those Null values by changing above query?
And please advice me if this is the correct way to acheive similar functionality. Thanks in advance :)
The quickest, dirtiest way to achieve what you want is just to append this where condition to your query:
WHERE d_city_id is not null
but you might prefer to rework your query and decide where you really need LEFT joins and not INNER joins
the IF() computed column is in essence what STT LCU was trying to offer, but you can't use that directly in the where for some reason.
I've rewritten your query, but with different aliases to better follow the origination of the tables / relationships to get the data. In the end, I've added a where to test for ANY ONE of the "ID" values as NOT NULL. If they are ALL Null, the record should be excluded..
select
d.id AS clinic_id,
if(CityViaState.id is not null, CityViaState.id,
if( ByCity.id is not null, ByCity.id, CityViaCountry.id )) AS d_city_id
from
d_location_subscriptions dls
join clinics d
ON dls.clinic_id = d.id
left join d_countries ByCountry
ON dls.t_id = ByCountry.id
and dls.type = 'country'
left join d_states StateViaCountry
ON ByCountry.id = StateViaCountry.d_country_id
left join d_cities CityViaCountry
ON StateViaCountry.id = CityViaCountry.d_state_id
left join d_states ByState
ON dls.t_id = ByState.id
and dls.type = 'state'
left join d_cities CityViaState
ON ByState.id = CityViaState.d_state_id
left join d_cities ByCity
ON dls.t_id = ByCity.id
and dls.type = 'city'
where
CityViaState.id is not null
OR ByCity.id is not null
OR CityViaCountry.id is not null
I've been working on a project that until now has only needed to find 1 row from the joined table. But now I need to grab multiple rows..
So as it stand my sql works something like:
Select rows for each company for this particular project which alone would find company details (name, id, telephone.. blah).
Then I join a table that contains form data submitted for each company (multiple forms - so multiple records)
Until now i have been specifying one formid to look for in the join, but now i need to specify multiple ones.
If I use WHERE form_id = 1 OR form_id = 2 OR form_id = 3 ... I get a result of only the first form match that is found per company..
If I mix up the query so it looks for the forms 1st and returns multiple records for each company with different form data - that works in this sense..
But I am then looping through this array in a view and creating a table row per record (previously each row was a new company) but using the latter would cause multiple records to show for the same company.
Any way I can do this? I tried group by with the latter method but this results in only 1 record again.
SELECT DISTINCT p.project_company_has_user_id, p.project_company_has_user_project_id, p.project_company_has_user_user_id, c.company_id, c.company_hall_no, c.company_company_name, c.company_type, c.company_country, c.company_stand_number, c.company_image_file_1, p2.project_id, p2.project_name, u.user_id, u.user_username, o.orders_id, o2.order_detail_id, o2.order_detail_product_id, f2.form_question_has_answer_id, f2.form_question_has_answer_request, f2.form_question_has_answer_form_id, f2.form_question_has_answer_user_id
FROM project_company_has_user p
INNER JOIN company c ON p.project_company_has_user_company_id = c.company_id
INNER JOIN project p2 ON p.project_company_has_user_project_id = p2.project_id
INNER JOIN user u ON p.project_company_has_user_user_id = u.user_id
INNER JOIN form f ON p.project_company_has_user_project_id = f.form_project_id
LEFT JOIN orders o ON p.project_company_has_user_user_id = o.orders_user_id
LEFT JOIN order_detail o2 ON ((o2.order_detail_orders_id = o.orders_id AND (o2.order_detail_product_id = 65 OR o2.order_detail_product_id = 68 OR o2.order_detail_product_id = 64)))
LEFT JOIN form_question_has_answer f2 ON ((f2.form_question_has_answer_form_id = 297 AND f2.form_question_has_answer_user_id = p.project_company_has_user_user_id))
WHERE (f.form_template_name = "custom" AND p.project_company_has_user_garbage_collection = 0 AND p.project_company_has_user_project_id = 48) AND (LCASE(c.company_country) LIKE "%uk%" OR LCASE(c.company_country) LIKE "%uk%") ORDER BY company_company_name asc
you need another field in order_detail as o2 . this field is row_index(position),etc for positioning record
LEFT JOIN order_detail o2 ON (o2.row_index=1 AND (o2.order_detail_orders_id = o.orders_id AND (o2.order_detail_product_id = 65 OR o2.order_detail_product_id = 68 OR o2.order_detail_product_id = 64)))
Personally I would use an Outer Join for the table of which elements you need to list all matches. Should you them need to clean up that data you can build the logic into the Join Condition (as step 2). Depending on the volume of data you are handling and whether or not you need to reuse it later in the same proc, you may want to post that primary dataset into a temp table and use that as source (primary) for your later logic.
Hope that helps. If you need the code, let me know, but it is pretty straight forward.
Regards
Mac