No matter how many monitors you have or how large they may be, sometimes you just need more real-estate for your coding. You can have a whole lot going on in your IDE, and SQL Developer is no exception. Of course you can always close or hide the offending panels and windows, but then it’s a real pain to get them back.

WordPress has a nice fullscreen editor I can use when writing these posts. It’s just a click away…

The Word Press Full Screen Post Editor

Wouldn’t it be nice to have the same thing in SQL Developer? For example, I’m trying to work on my query in the worksheet, and I’m having problems concentrating…Squirrel!

Oracle SQL Developer in all it's 'glory'

Just Double-Click on the Tab to go full-screen

Double-click to go full-screen

Et voila! SQL Developer closes all the other panels and you get JUST the worksheet (or PLSQL Editor.)

Nothing comes between you and your code!

When you’re done, just double-click to go back to your normal view.

If you happen to close a support panel like the Snippets, remember you can always get those back by going to the View menu. And if you’d prefer to minimize or hide the panels, just click on the little minimize ‘_’ button at the top right of each panel.

Just so you don’t think I’m crazy, you should know that I’m a father to a wonderful two year old daughter. She’s awesome, except when she wants to act her age. When we get desperate, we turn on her favorite show. I have seen about 200 different episodes of Dora the Explorer. My Spanish is getting better, but I’ve got these really bad songs stuck in my head. I’m going to share one of these with you right now.

If there’s a place you got to go
I’m the one you need to know
I’m the Map
I’m the Map
I’m the Map
If there’s a place you got to get
I can get you there I bet
I’m the Map!

Wouldn’t it be awesome if life were like this? Whenever you needed something, there was this magical talking map that told you exactly how to get what you wanted. Unfortunately life can’t be that simple, but sometimes our software can be.

Yes, I’m using Dora as a segue for a SQL Developer Tip and Trick post. Blogging is much cheaper than therapy, and the drugs are better!

SQL Developer Data Modeler Navigator Can Get You There I Bet!

Data models can be quite large. It’s a real challenge to take a workspace and evenly distribute the table and view objects in the diagram in such a way that it’s easy to read. Especially when everyone’s monitor and resolution setups are different. You can just about bet the default presentation of your model is going to be problematic. That’s why we have many different tools for helping re-arranging objects automatically, zooming in and out, and of course the ‘Go To Diagram’ function you can call directly from the object explorer.

But, sometimes in life, what you REALLY need is a map!

Let’s take a quick look at a big model:

Ok, so where are my darn tables and views already?!?

Tucked neatly in the upper-right-hand corner of the modeler is a panel labeled ‘Navigator.’ This will show us the way!

The Navigator is our Map

If you click on the panel and then on the ‘Restore’ button, you’ll get something like this -

The Navigator makes it easy to navigate your models

As you click around the navigator, the ‘red box outliner thingie’ (can you tell that I am college educated?) will auto-zoom/navigate the area in the main model space.

You might be wondering why I have dedicated 500+/- words to such a simple topic, and I’ll be honest with you – I did not notice this feature until just recently. The navigator by default is minimized in the application and it’s very easy to miss. Plus we are often very reticent to ask for help and directions – even when the map is right there for us to consult. I hope this helps save you at least a few seconds while you work with your data models.

Hopefully your code is cleaner than what runs through this every morning

Gutter:
The gutter area can be used to display bookmarks, line numbers and outline sections.
The gutter is the area to the left of the text, the purpose of which is to display miscellaneous indicators for the corresponding lines of text. Among these indicators are bookmark indicators, line wrapping indicators, line styles icons, and line numbers. All the images displayed in the gutter are contained in the gutter image list.

Notice the VERY last thing mentioned in that description. A place to display ‘line numbers.’

Now here is the question of the day -

Should your IDE show these by DEFAULT?

I’ll save you the suspense and tell you that many of the IDEs out there DO NOT, and that includes SQL Developer. There are probably a few reasons, and I can make up more, but it probably comes down to this:

• JDeveloper doesn’t
• Keep the coding area as clean as possible

One of those reasons is a technical reason, and end users could care less about those. The other is more of a personal choice / philosophical decision. When you make these types of decisions, you opt for something with minimal impact to the user base, and aim to ‘do no harm.’ Did we make the right decision here? For many of you, yes we did. For many others of you, no we did not.

In our defense I could publicly mention several other database IDEs that have the exact same behavior. But, I would also argue with myself that those people are no better off and that we should strive to be the best. Yes, I do tend to argue with myself quite a bit.

One of my Twitter friends asked me if SQL Developer was an IDE or was it Notepad. Ouch, being compared to Notepad. Although I will admit to Notepad being my default HTML editor and I have no problems not having line numbers there.

I get his point though, serious coders are working on programs with many thousands of lines of code. They need little helpers to get them through the day.

Lets take a quick look at the following picture:

SQL Developer Worksheet with Line Numbers Shown

So the error message is on line 11. This is where the Oracle parser gets upset that it’s not finding that COMMA that it expects. If it were complaining about line 24,327 it might be a pain to get to that line number, right?

Except.

Ctrl+G will open the ‘Go to line number’ dialog for easy jumping.

It’s a click-click to get the line numbers on. I suppose you could argue that folks might not know where to click. Right-mouse-click in the gutter space.

Right-mouse in the gutter

If you are in the PL/SQL Editor, you can actually click on the error message in the ‘Compiler Log’ panel and it will take you to the indicated line number for that error message.

Maybe I am too close?

What I see as intuitive, you see as a trick?

If you think we’re doing it wrong, please add your vote in the form of a comment to this blog post. You can also up-vote this request in the Exchange, although it needs re-worded.

Maybe we need a SQL Developer++? Notepad++ after all DOES show line numbers in their gutter by default

Oracle SQL Developer Job Run Log

The job runs can be viewed via the dba_scheduler_job_run_details SYS view. Oracle SQL Developer shows this data in the job details panel of the schema browser even. So, answering the question, ‘How long does this job normally take to run?’ should be relatively easy to answer.

It is, if you can think in terms of INTERVAL. You see, Oracle doesn’t store the time a job takes to run in seconds, it uses the INTERVAL datatype.

desc dba_scheduler_job_run_details
Name              Null Type
----------------- ---- ----------------------------
LOG_ID                 NUMBER
LOG_DATE               TIMESTAMP(6) WITH TIME ZONE
OWNER                  VARCHAR2(30)
JOB_NAME               VARCHAR2(65)
JOB_SUBNAME            VARCHAR2(65)
STATUS                 VARCHAR2(30)
ERROR#                 NUMBER
REQ_START_DATE         TIMESTAMP(6) WITH TIME ZONE
ACTUAL_START_DATE      TIMESTAMP(6) WITH TIME ZONE
RUN_DURATION           INTERVAL DAY(3) TO SECOND(0)
INSTANCE_ID            NUMBER
SESSION_ID             VARCHAR2(30)
SLAVE_PID              VARCHAR2(30)
CPU_USED               INTERVAL DAY(3) TO SECOND(2)
CREDENTIAL_OWNER       VARCHAR2(65)
CREDENTIAL_NAME        VARCHAR2(65)
DESTINATION_OWNER      VARCHAR2(128)
DESTINATION            VARCHAR2(128)
ADDITIONAL_INFO        VARCHAR2(4000)

So what is an INTERVAL?

Technically it’s INTERVAL DAY TO SECOND. From the docs -

INTERVAL DAY [(day_precision)] TO SECOND [(fractional_seconds)]
Stores a period of time in days, hours, minutes, and seconds, where
day_precision is the maximum number of digits in the DAY datetime field. Accepted values are 0 to 9. The default is 2.
fractional_seconds_precision is the number of digits in the fractional part of the SECOND field. Accepted values are 0 to 9. The default is 6.
The size is fixed at 11 bytes.

So if I’m looking at an entry in dba_scheduler_job_run_details, and it shows ’0:0:1.0′, we can read that as ‘Zero days, zero hours, and 1.0 seconds.’ As you can imagine, asking for an average using the avg() function doesn’t turn out so well -

select avg(run_duration) from dba_scheduler_job_run_details;

ORA-00932: inconsistent datatypes: expected NUMBER got INTERVAL DAY TO SECOND
00932. 00000 -  "inconsistent datatypes: expected %s got %s"
*Cause:
*Action:
Error at Line: 3 Column: 11

How do I get a number out of an INTERVAL?

Oracle provides a function called EXTRACT(). It allows you to pull DAYS, HOURS, or SECONDS from an INTERVAL. BONUS FEATURE: It also works for TIMESTAMPS! However, there is no direct conversion function that would take an interval and return the total number of seconds or minutes or hours, or a fraction thereof. So we have to do some MATH. I know, I know – that hurts. But, we can always cheat and ‘borrow’ from the great Tom Kyte

Using Tom’s math – basically taking the number of days and hours in a day and multiplying out to get seconds and then adding to the seconds field gives me total number of seconds. In my case I’m OK with seconds b/c most of my jobs don’t take very long to run. You could adjust the math and get an avg number of hours or days instead. I’m going with seconds.

So here is my borrowed query from Tom:

select job_name,
       job_name,
       avg(extract( day from run_duration )*24*60*60 + extract( hour from run_duration )*60*60 + extract( minute from run_duration )*60 + extract( second from run_duration ))
from dba_scheduler_job_run_details
group by job_name, job_name
having avg(extract( day from run_duration )*24*60*60 + extract( hour from run_duration )*60*60 + extract( minute from run_duration )*60 + extract( second from run_duration )) > 0
order by 3 desc

Why did you put JOB_NAME in there twice?

Because I wanted a chart report!

I like pretty pictures – it’s why my career path looks so bright (this is a joke about managers, they love charts.)

SQL Developer Report Showing Average Job Run Time with Detail Breakouts

So clicking into my …_MAINTAIN_STATS job I can see that it ran in an avg of 5.25 seconds for a total of 4 times in the past 30 days. Averages can be misleading though. Four runs from 0 to 10 seconds is quite the range. You may want to build some logic in your report to restrict any average that doesn’t have at least a statistically significant number of data points.

Thanks to my Twitter friends for giving me something to chew on (and blog about!) this morning

I am just getting my feet wet when it comes to the reports in SQL Developer. The tool itself has quite a lot going on, everything from a database IDE to a migration platform, to a data modeling component. And then throw in the reporting stuff. I’ve already covered how to build a simple bar chart report in an earlier post, but now I want to take this report to the next level!

Basic reports are tied to a single dataset. Adding one or more child reports allow you to take a value from the parent dataset and feed that as an input to a secondary query and report. Maybe my report is showing a breakdown of breweries by country in a bar chart, and I want my report consumer to be able click on a country and then be shows a list of the breweries for that country. SQL Developer can accomplish this using a Child Report.

The Base Report

For the parent report I am going to use my previous beer example. It’s a chart report, and here’s the underlying SQL statement:

SELECT country,country, count(*) FROM beer
group by country, country
having count(*) > 100
order by 3 desc

Running this report gives me a chart that looks like this:

My Breweries by Country Summary Report Chart

Now let’s add the child report

Step 1: Add the Child Report

Right-click on the report in the tree and choose ‘Edit Report.’ We are going to go to the Child Reports tree and ‘Add Child.’ Note that you can add as many child reports as you would like, but I’m going to stick with the walking before the running.

Where do we go?
Where do we go now?
Where do we go?

I am going to choose a style of ‘Table’ for the report. I just want a grid style list of breweries by the selected country. The most important information to provide here will be the query that will be used to populate said grid. To tie the two datasets together, we need to use a bind variable. The ‘trick’ here is to uppercase the bind variable name to match the column name from the master query.

The Bind Variable ties the Master and Child queries together

For this particular report, the bind feeds off the COUNTRY name. So my child report query looks like this:

select brewery, city, state
from beer
where country = :COUNTRY

As I click on a Country in the master chart report, that value is fed to this query and will populate the grid for the child report.

Where do we go?
Where do we go now?
Where do we go?

The Completed Master-Child Report

The report is interactive. The chart labels are active, click and get the details!

A new feature for v3.1 is our PDF reporting. We will have a video tutorial posted to coordinate with the release and will be available on OTN.

Where do we go now?
No, no, no, no, no, no
Sweet child,
Sweet child report of mine.

Data models sure are handy. They tell a story – how does the information get stored as data? Just what is a ‘customer?’ Sometimes though our applications die and only leave legacy scripts behind. Or maybe you are a DBA and someone has given you a set of scripts to create some new objects in your database. You ask them for their model, and they just look at you with that weird blank stare. “We have models, you say?” Of course, geeks have been building models since at least 1985.

The Reverse Engineer, or IMPORT, feature in Oracle SQL Developer Data Modeler will read information from an existing database to build your data model. But did you know it can also read in Data Definition Language (DDL) scripts? That’s right, those directories you have laying around with hundreds of CREATE OR REPLACE… scripts can be used to build a model.

For quick and dirty modeling tasks, I say just run that directly in SQL Developer (accessible under the View menu.) If you are going to get serious and build some new models from scratch, then you should probably run the standalone Oracle SQL Developer Data Modeler program. I also find the user interface much simpler in the standalone – it’s all about modeling. Some of the more interesting modeling features are ‘hidden’ if you’re in the main SQL Developer interface.

For this brief demonstration I’ll just run this in SQL Developer.

Disclaimer: This post was written with Oracle SQL Developer v3.1.

Open your model using the View – Data Modeler – Browser interface

You can now use the IMPORT feature

It says 'Script' but you can import multiple scripts

Select your file(s)

Choose your DLL script files to be parsed and imported

If you don’t happen to have any DDL scripts laying around, feel free to build some new scripts using the new Cart feature.

Tell us what RDBMS mapping scheme to use

It’s easy to think that SQL Developer only works with Oracle, but did you know the Data Modeler works with most RDBMS systems, including DB2 z/OS, DB2 LUW, and SQL Server?

The Data Modeler needs a little help, tell it where your DDL is coming from.

Review the results and proceed to your new model and Entity Relationship Diagram!

If there are any problems parsing the DDL script, the modeler will warn you about it. You can now review your scripts visually using the ERD. A newer feature for v3.1 allows you to automatically step into a SubView based on a Schema. So if your scripts are importing from several sources, you can look at all of it together, or just for user/database ‘x.’ Actually this feature is so new I saw it in a developer demo just today

People like to rag on java and it’s performance issues. But I was able to parse more than 1500 DDL files in less than a minute and have my model created for me. It’s still a fairly small model, only 269 tables, but that’s still pretty good if you ask me.

From DDL Files to Model - why read when you can look at the pretty pictures?