Let's assume I have two Entities in my Symfony2 bundle, User and Group. Associated by a many-to-many relationship.
┌────────────────┐ ┌────────────────┐ ┌────────────────┐
| USER | | USER_GROUP_REL | | GROUP |
├────────────────┤ ├────────────────┤ ├────────────────┤
| id# ├---------┤ user_id# | ┌----┤ id# |
| username | | group_id# ├----┘ | groupname |
| email | | created_date | | |
└────────────────┘ └────────────────┘ └────────────────┘
What would be a good practice or a good approach to add additional columns to the join table, like a created date which represents the date when User joined Group?
I know that I could use the QueryBuilder to write an INSERT statement.
But as far as I have not seen any INSERT example of QueryBuilder or native SQL which makes me believe that ORM/Doctrine try to avoid direct INSERT statements (e.g. for security reasons). Plus as far as I have understood Symfony and Doctrine I would be taken aback if such a common requirement wouldn't be covered by the framework.
You want to set a property of the relation. This is how it's done in doctrine:
doctrine 2 many to many (Products - Categories)
I answered that question with a use case (like yours).
This is an additional question / answer which considers the benefits and use cases: Doctrine 2 : Best way to manage many-to-many associations
Related
How does one return a list of unique users from a dynamodb table with the following (simplified) schema? Does it require a GSI? This is for an app with small number of users, and I can think of ways that will work for my needs without creating a GSI (like scanning and filtering on SK, or creating a new item with list of user ids inside). But what is the scalable solution?
------------------------------------------------------
| pk | sk | amount | balance
------------------------------------------------------
| "user1" | "2021-01-01T12:00:00Z" | 7 |
| "user1" | "2021-01-03T12:00:00Z" | 5 |
| "user2" | "2021-01-01T12:00:00Z" | 3 |
| "user2" | "2021-01-03T12:00:00Z" | 2 |
| "user1" | "user1" | | 12
| "user2" | "user2" | | 5
Your data model isn't designed to fetch all unique users efficiently.
You certainly could use a scan operation and filter with your current data model, but that is inefficient.
If you want to fetch all users in a single query, you'll need to get all user information into a single partition. As you've identified, you could do this with a GSI. You could also re-organize your data model to accommodate this access pattern.
For example, you mentioned that the application has a small number of users. If the number of users is small enough, you could create a partition that stores a list of all users (e.g. PK=USERS). If you could do this under 400kb, that may be a viable solution.
The idiomatic solution is to create a global secondary index.
Dynamodb can make the simplest of database operations difficult. I have the following table and all I want to do is simply sort by the due column. How is this achieved in DynamoDb? I read everything I could find online and there doesn't seem to be a straightforward walk through anywhere.
payor | amount | due | paid
----------------------------------
Ally | 200.00 | 13 | 1
Chase | 80.00 | 2 | 0
Wells | 30.00 | 17 | 1
Directv | 150.00 | 5 | 0
So without considering the payor, amount or paid columns, how can I simply sort on the due column.
Simply, this can't be achieved in DynamoDB if the due attribute is not defined as sort key. Even if you define the due attribute as sort key, the ordering can be done only within the particular partition key. The ordering can't be done across the partition key.
Assume, you have defined the due as sort key of the table. You can use ScanIndexForward to true/false to order the items in ascending / descending order.
Data modeling in dynamo db involves designing the partition key and then determining the sort key for a use case. Partition key is compulsory for any query. This is a basic design premise of a key value nosql store which is completely different than a relational store
I am a web developer who's working on an Exam Generator project. Now I am stuck at one point.
I have one database with different tables. Four of these tables are somehow similar with their columns. I want to know what is the best practice for such thing.
The similar tables I have are:
Exam (Used to store the exam name and the number of questions
included in the exam).
ID | ExamName | NumberofQuestions
UserExam (Used to store the the exams availble for a user with his
grade in each exam he took).
ID | MemberID | ExamID | Grade
QuestionExam (Used to store Question IDs included in each exam).
ID | ExamID | QuestionID
UserSolution (Used to store the user's answers for each exam he
took).
ID | MemberID | ExamID | QuestionID | UserAnswer
In the beginning, I wanted to merge the "Exam" table with the "QuestionExam" table, but then I asked myself if I merged them how would I have one ID for each exam? So I kept it as it is.
Everything is correct with the tables. IF you want to be more desciriptive some practices say to value like Exam_ID, UserExam_ID, or UserSolution_ID that way you can distinguish between the two in a join. It all depends on personal preference. It is a little more writing but saves you from a headache in the long run.
I am a software engineer, but I am very new to databases and I am trying to hack up a tool to show some demo.
I have an Apache server which serves a simple web page full of tables. Each row in the table has a proposal id and a link to a web page where the proposal is explained. So just two columns.
----------------------
| id | proposal |
|--------------------
| 1 | foo.html |
| 2 | bar.html |
----------------------
Now, I want to add a third column titled Comments where a user can leave comments.
------------------------------------------------
| id | proposal | Comments |
|-----------------------------------------------
| 1 | foo.html | x: great idea ! |
| | | y: +1 |
| 2 | bar.html | z: not for this release |
------------------------------------------------
I just want to quickly hack up something to show this as a demo and get feedback. I am planning to use SQLite to create a table per id and store the userid, comments in the table. People can add comment at the same time. I am planning to use lock to perform operations on the SQLite database. I am not worried about scaling just want to show and get feedback. Are there any major flaw in this implementation?
There are similar questions. But I am looking for a simplest possible implementation.
Table per ID; why would you want to do that? If you get a large number of proposals, the number of tables can get out of hand very quickly. You just need to keep an id column in the table to keep track of things and keep the number of tables in a sane figure.
The other drawback of using a table for each proposal is that you will not be able to use prepared statements for those, because table names cannot be bound as a parameter.
SQLite assumes the table name is 'a'
Add column
alter table a add column Comments text;
Insert comment
insert into a values (4,"hello.html","New Comment");
You need to provide values for the other two columns along with the new comment.
I want to create a page in Drupal to report some basic forum information. I thought I'd use Views, but Views only lets you set one "entity" type per view but forum topics are made up of nodes and comments (aka, topics and replies).
Ideally, I'd like a single view that lists all forum nodes and comments together in a single table (sorted by date), along with a total number of both combined, if possible. Is there a way to do that with Views?
Update: What I'm looking for is something like this:
-------------------------------------------------------
| User | Post | Type | Date |
-------------------------------------------------------
| amy | post text appears here | post | 1/5/01 |
| bob | comment text appears here | comment | 1/5/01 |
| amy | another comment here | comment | 1/5/01 |
| cid | another post appears here | post | 1/4/01 |
| dave | yet another comment here | comment | 1/4/01 |
-------------------------------------------------------
total posts + comments: 5
Not sure what you really want. Either you can display nodes + number of comments or nodes and comments at the same level but then they don't have a total number because they are all separate? Or do you want to show each comment separate together with the number of comments in that thread?
If the latter, that might not be trivial.
Basically, you could create a UNION Select query and query both the node and the comment table. could look like this:
(SELECT 'node' AS type, n.nid as id, n.title as title, nncs.comment_count as comment_count, n.created as timestamp FROM {node} n INNER JOIN {node_comment_statistics} nncs ON n.nid = nncs.nid)
UNION
(SELECT 'comment' AS type, c.cid as id, c.subject as title, cncs.comment_count as comment_count, c.timestamp as timestamp FROM {comments} c INNER JOIN {node_comment_statistics} cncs ON c.nid = cncs.nid)
ORDER BY timestamp DESC LIMIT 10;
That will return a result containing: node/comment | id | title | comment_count | timestamp.
See http://dev.mysql.com/doc/refman/5.1/en/union.html for more information about UNION.
You can then theme that as a table.
Hints:
If you need more data, either extend
the query or use node/comment_load
You could also join {node} in the
second query and use the node title
instead of comment subject
That query is going to be slow
because it will always do a filesort
because you have a union there. It
might actually be faster to execute
two separate queries and then mangle
them together in PHP if you have a
large number of nodes/comments
It turns out the Tracker 2 module provides enough of what I needed.