Archive for the ‘Python’ Category
First of all, Happy New Year!
IEEE Spectrum published a ranking of the most popular programming languages. Computational journalist Nick Diakopoulos wrote the article. While it may surprise some, I wasn’t surprised to find SQL in the top ten.
- Compiled programming languages (Java [#1], C [#2], C++ [#3], C# [#4], Objective-C [#16])
- Data languages (SQL [#9], MATLAB [#10], R [#13])
I couldn’t resist including Objective-C because it shows how the iPhone, iPad, and Mac OS impact our daily lives. At the same time, Assembly [#15] is actually more popular than Objective-C. Shell [#17] follows Objective-C. While the Visual Basic [#14] programming language still remains very popular.
There are many “why” questions raised by this list of popular programming languages. The “why” from my perspective deals with what are the market drivers for their popularity. The money drivers I see are as follows:
- Business software: Java, C++, C#, and AIDE – Android IDE (works with Java and C++ source code)
- OS X and iOS Development: Objective-C
- Development Tools: Java, C, C++, and Python
- System Admin/Utilities Tools: C, Perl, and Shell
- Web Development: Python, PHP, Ruby, and Perl
- Data Analysis: SQL, MATLAB, and R
Business Intelligence (BI) software manages most high-level data analysis tools and they’ll continue to get better over time. However, if SQL has shown us anything over 30 years it’s that ultimately we revert to it to solve problems. The conclusion from the reality of BI probably means the programming languages that develop those tools will continue to rise and so will the underlying data languages.
In prior years a daily update from Open World was possible, but this year my schedule was too full to support it. This is my compendium of thoughts about MySQL Connect, JavaOne, and Open World 2012.
MySQL Connect was great – good sessions re-enforcing the positive investments Oracle is making in the product. I’ll leave to others to qualify changes in what elements of technology are opened or closed along the road to a better MySQL. The announcement of Connector/Python 1.0 GA on Saturday was great news and as a community we owe a lot to Greet Vanderkelen.
NoSQL is a hot topic along with using JSON objects and it was interesting hearing of some unequal testing paradigms to position non-Oracle solutions to be “better” than Oracle solutions. Naturally, the MongoDB was the elephant in the room during those conversations. Some of the discussions seemed more like political rants than technical dialog. A great spot to start with NoSQL and JSON would be downloading Oracle’s MySQL 5.6 Release Candidate.
There were also more PostgreSQL conversations this year and fairly accurate comparisons between it and Oracle or MySQL from folks. It certainly looks like it may gain more ground.
Java 7 is awesome, and my favorite feature is clearly NIO2, reinforced at JavaOne. NIO2 brings static methods to interactions with external directory and file sources. It removes directories from the files class, which is long overdue. The nature of those static methods also happen to fit within the definition of Java code that lives inside the Oracle database and gives me a whole host of thoughts about potential in Oracle Database 12c.
Larry Ellison’s keynote was impressive because it gives us a clear vision of Oracle’s direction and Duncan Davies captured the keynote well in his blog. The continued presence of Red Hat and VMWare offers interesting reality checks to their key contributions to world wide implementation of the Oracle technical stack.
Issues that seem most critical to those I’ve chatted with are storage, security, tools, and development languages. A nice update on security can be found in the new edition of Hacking Exposed 7: Network Security Secrets & Solutions (7th Edition).
On the forthcoming Oracle 12c release, Information Week just released a good summary view. The introduction of the R programming language on the Exadata Server leads me to wonder about what uses may magically appears in Oracle Enterprise Manager down the road. The TIOBE Index for September 2012 doesn’t list the R language in the top 20 programming languages but there’s always the future. No mention of Erlang programming language at any of the conferences that I caught but it’s inevitably on the horizon as application servers evolve.
Now we wait for the Oracle Database 12c release, which looks like something in the very short term. Perhaps right after the holidays …
It’s always interesting when somebody asks why they got an error message, and especially sweet when you’re working on something related that lets you answer the question. They were using MySQL Workbench and wanted to know why they couldn’t open a SQL script file by clicking on the Scripting menu option.
As I explained to the individual who asked, you should always click the Edit SQL Script link in the SQL Development section of the MySQL Workbench home page to work on SQL scripts. The Scripting menu option supports Python and Lua plug-ins development and scripts.
They did the following initially, which led down the rabbit warren and left them stumped because they don’t know anything about Python or Lua. This is provided to those who choose to experiment with this advanced feature of MySQL Workbench.
That presents you with a chooser dialog and it lets you pick any type of file. (You may wonder, as I did, why they didn’t restrict it to
.lua file extensions, which would preclude opening a
.sql file. I actually logged an enhancement request to see if the development team may agree with me.) You get the following message when you choose something other than a Python or Lua script. You can click on any of the reduced size screen shots to enlarge them and make them readable.
As you may note, the dialog says the activity is unsupported by provides no cancellation button. Click the OK button and the unsupported file is loaded into a tab that is useless. All you can do is click to close the tab and dismiss the window.
After you dismiss (by clicking the x) the non-editable
.sql file, you need to click on the Open Script file icon shown below.
This chooser really should open where the default is for the MySQL Workbench application script files but it doesn’t. It opens in the last accessed directory. You need to navigate to where your Python or Lua scripts are stored, which is the following directory on Windows:
Please note that on a Windows system you can’t chose this directory option because it’s protected. You must enter the navigation bar and type it. Then, you should see any scripts that you saved from within MySQL Workbench.
ReadFile.py below contains a rather simplistic and static program that reads a file and prints it to console (it’s small and fits in the screen). Obviously, it dispenses with a bunch to keep it small but check a Python website or book for the right way to manage a try block and handle exceptions.
ReadFile.py file shown in the preceding and next screen shots. For those new to Python, watch out because tabs aren’t equivalent to spaces. I made a change in the script below to display the trailing semicolon because one of my students asked about it.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
# -*- coding: utf-8 -*- # MySQL Workbench Python script # ReadFile.py # Written in MySQL Workbench 5.2.41 import os f = open("c:\\Data\\MySQL\\query.sql",'rU') while True: line = f.readline() # Parse string to avoid reading line return. if not line[len(line) - 1:len(line)] == ";": print(line[0:len(line) - 1]) else: print(line) if not line: break
Life’s funny, and you can never please everyone. The latest question, “Why did I choose to use substrings when suppressing line returns from the
print() function is easier?” Simple answer because the approach differs between Python 2.7 and 3.0 and I didn’t want this post to have a lot of Python nuance.
Python 2.7 (compatible with MySQL Workbench 5.2):
1 2 3 4 5 6 7 8 9
import os f = open("c:\\Data\\MySQL\\query.sql",'rU') while True: line = f.readline() # Suppress line return. print(line), if not line: print break
Python 3.0 (not-compatible with MySQL Workbench 5.2)
You should take note that both version require a print statement on line #8. Line #6 above shows that Python 2.7 uses a comma to suppress the line return, and below line #6 shows Python 3 requires you set
end equal to an empty string. Line #8 below also has a set of empty parentheses, which works in Python 3.x but not in Python 2.7. Python 2.7 would print the parentheses unless you put an empty string inside of them, like a
1 2 3 4 5 6 7 8 9
import os f = open("c:\\Data\\MySQL\\query.sql",'rU') while True: line = f.readline() # Suppress line return. print(line, end = '') if not line: print() break
Hopefully, everyone concurs the parsing was simpler than explaining all these Python nuances. Although, it’s nice somebody was so curious.
If your script complies with the Python 2.7 rules (that’s what is deployed in MySQL Workbench), click the lighting bolt and your code will run and display the results. That’s shown in the last screen shot.
Naturally, I hope this helps those experimenting but personally it’s a cool advanced feature of the MySQL Workbench.