By running dbt docs generate and dbt docs serve, I can generate a documentation for my models that contains all column names automatically. If I do something like this:
models/.yml
version: 2
models:
- name: events
description: This table contains clickstream events from the marketing website
the documentation that is generated already contains a list of columns + their type. However, the column description column is empty. If I want to add a description, I can do something like this:
version: 2
models:
- name: events
description: This table contains clickstream events from the marketing website
columns:
- name: event_id
description: This is a unique identifier for the event
tests:
- unique
- not_null
However, in this way, I will have to manually type all column names. Is there any way I can automate this? Such that the column names can be automatically extracted in a list etc and I only have to write the descriptions? I will have to do this for multiple files.
The canonical way to do this is with the codegen package, which generates the files exactly as you have described.
Related
Our organization hundreds of Periscope daashboard that is generated from a database, named "Animal".
Now, let's suppose there is a table named "Puppy" with column name "is_spotted". Is there an easy way to find out whether "is_spotted" has been used to create a dashboard, without going through every single dashboards?
I would like to add descriptions to each field in a table. My issue is that we are using dbt and this tool recreates the table each time you run a job causing the descriptions to be deleted if they are there. I am able to control the datatypes casting the fields in the last SELECT statement but I am not sure if I can add the description using SQL.
I have been googling for a while and I am not able to see whether descriptions can be added using SQL this way.
I've thought a workaround which would be to create the table and then insert but this in theory is bad practice using dbt.
Thanks!
just want to post the solution in case someone has the same problem. dbt didn't update the descriptions in BQ itself. However, they released this new feature last month: https://github.com/fishtown-analytics/dbt/releases/tag/v0.17.0
Docs can be generated as usual and BQ will show the descriptions of tables and columns. You will only need to add the below to your dbt_project.yml file:
+persist_docs:
relation: true
columns: true
You can insert decriptions two ways.
Using the schema.yml file
{version: 2
models:
- name: events
description: This table contains clickstream events from the marketing website
columns:
- name: event_id
description: This is a unique identifier for the event
tests:
- unique
- not_null
- name: user-id
quote: true
description: The user who performed the event
tests:
- not_null
}
You can also use jinga templating within the SQL.
{% docs table_events %}
This table contains clickstream events from the marketing website.
The events in this table are recorded by [Snowplow](http://github.com/snowplow/snowplow) and piped into the warehouse on an hourly basis. The following pages of the marketing site are tracked:
- /
- /about
- /team
- /contact-us
{% enddocs %}
Several documentation examples mention that you can customize friendly_name
this is different from the name of the table.
Example: friendly_name="my_table". This property is equivalent to the
friendlyName table resource property
I am wondering what is the practical use to have a different table name, and a different friendly name? Does the UI or console display something different when these are used?
In Classic UI when you set friendly name - you get an extra field in Table Info called "Name" with value of whatever you set as Friendly Name - like for example "My Favorite Table"
In New UI same value can be found in second row of Table Info and is called - surprisingly - "Friendly name"
Rather than this I don't see any use of it so far in both UIs
You can't use the friendly_name to query a table.
I want to add a Column on the existing UIBB Schema i have
(SOLMAN_WORKCENTER - > TEST SUITE ANALYTICS - > Defect Analysis)
I tried to Enhance the Component Configuration.
So I thought i could simply add additional Columns of choice by entering the (+)Column wizard.
But the Column I need is not pre-included.
What is the correct way of adding extra columns to a UIBB Schema?
Basically, say I have a report A and I want to add a link column to it, and I set its target to a page in the application...the page its referring to has another report (lets call it B)
Is it possible for me to somehow make it so that table A's link column opens the page with Report B with rows that have the same column value for one of their columns?
Example here:
A
Name Num1 Num2
--------------
A 5 3
B 3 3
C 4 2
B
Name Quantity Serial
--------------------
D 2 3
E 1 8
F 4 6
So if I click the link column for row A, I want it to open report B and only shows rows where its Num2 = Serial, so only row D would show since it is the only one that equals 3
Using Report Linking to Filter Other APEX Report Outputs
I had some fun with this one although. The "A" and "B" stuff was pretty dry, so I decided to create a data set that was more engaging, and perhaps clearer to understand for the rest of us... :) This is how the data-ecosystem was broken down, and the way I fulfilled the OP requirements.
The Test Schema:
Welcome to the manufacturing facilities of the "Recipe Stack" Food Works. The schema design and ERD (Entity Relation Diagram) is below, with the sample data used for this demonstration:
The data relations are as follows:
The staff at the Stack Food-Works keeps an inventory of all ingredients for the types of meals and prepared foods they manufacture. Each ingredient has a unique ID (INGREDIENT_ID) and the staff tracks the amounts of each item in their pantry.
Each ingredient can be used in multiple recipes, but they will be used exactly once for a given recipe.
The ENTREE_RECIPE table has a COMPOSITE KEY which means it is the combination of the two pieces of this composite key (ENTREE_NAME and INGREDIENT_SEQ) that should be unique.
There is a FOREIGN KEY relation between the INGREDIENT_ID values of both tables.
Report Display Requirements (APEX and SQL Design Elements)
Selecting an item from the FIRST report is used as the input and the restriction/filter criteria of the second report.
User Case #1:
User Selects a Ingredient ID from the list of available ingredients in the pantry.
Input from (1) filters output of the RECIPE REPORT. This is a list of all the recipes that have the chosen ingredient in their formula.
User Case #2:
User Selects an ENTREE_NAME from the RECIPE REPORT. The ENTREE_NAME is used to deliver a third report: the RECIPE FORMULA which is the full recipe for the entree item that was selected from the previous report.
Testing Tools
(You usually need these these for the more complex pages, so it's a good start to use or develop them for the easier ones...)
I made my own, but you can also invoke the SESSION link on the developer's tool bar at the bottom of the APEX page on your running application (when it is displayed).
Here's my idea; it's a header region that also has a button to RESET input values so that I can clear the cache and retest or try other examples. I'll show later how you can use this link to see what is going on. You can see it in the discussion of testing at the bottom of this guide.
APEX Report/Page Design and SQL Parametrization
My columnn linking scheme looks similar to the previous post such as the one from FTaveras. This is how my linking works. What's different is that I do not go to another page, I simply go BACK to the same page I came from. Redirects and Branches apparently don't care if they are simply returning to the same location.
What is different on the return trip is that page parameters that were originally null or unpopulated NOW have a value. That value now brings life to the reports on the page that were empty.
Step 1: The PANTRY REPORT
Output: Query all items from the FOOD_SUPPLIES table.
Inputs: Supply links by INGREDIENT_ID to filter the RECIPE_REPORT output.
How to do it: (hint) To accomplish this, define the report column/field value in your report layout design page as a "linked" column and assign its value as a page item. The page item will be referenced in the SQL query of the next report...
Step 2: The RECIPE REPORT
Output: Query all records from the ENTREE_RECIPE table which have the INGREDIENT_ID from Step 1 within their formula.
Input: Supply links by ENTREE_NAME to filter the RECIPE_FORMULA output.
How to do it: (hint) Include the page item defined from Step 1 within the SQL query of this report:
SELECT * FROM entree_recipe
WHERE ingredient_id = :P3_INGREDIENT_ID
Step 3: The FORMULA REPORT
Output: Query all records from the ENTREE_RECIPE table which have the ENTREE_NAME selected from Step 2.
Debug and Test Run
Most will be able to get this far without any problems. If not, here are a couple of examples of how you can debug and test your work. There may be some built-in tools and packages that already exist within Apex, so any suggestions on alternate approaches are welcome in the comments...!
Using the APEX Developer SESSION Output
After selecting the inputs for the trial run, click on the SESSION link on the developer toolbar at the bottom of the page. This is an example output:
Note that the page items that were set for that session are displayed. The inputs I used for this test were:
INGREDIENT_ID: 6432
ENTREE_NAME: peach cobbler (fresh)
Extra Credit: This one is an alternate approach. It may be useful to design something like this as an add-on to any app you design. You do not need to remove it from your app when you push it to production because there is a "conditional display/suppress" feature for page regions. (check it out)
Simply set a global parameter as a "mode" on your Apex app. Set the value to "DEBUG" or "TEST" or whatever and key all your instances of this page region to display only when the global parameter is set to it.
Wrap Up and Discussion
Hopefully, you've enjoyed your visit to the "Stack Food Works" (no tasting or sampling from the line, please).
This has been more of a holistic approach to Apex app design. It helps to have a methodology to map out each step, and a way to check your work at both the beginning and the end of your development process. Using smaller examples like this demo to apply these methods provides a chance to understand Apex development as a Software Creation PROCESS.
Yes it's possible.
On page B, add and hidden Item name for example P(#)_SERIAL where (#) is your current page number.
Modify your query and add one line like AND SERIAL=:P(#)_SERIAL.
On Page A go to "Report Attributes" tab on the report region, click edit.
On the column link section configure your link to page B and set the hidden item on page B to the value on report column of page A.
Name: Item 1 [P(#)_SERIAL] Value: #Num2#
Demo