Practical Notes, Code Samples, and Musings
For forcing ucs4 when installing Python:
./configure --enable-unicode=ucs4 make make install
For forcing ucs2 when installing Python:
./configure --enable-unicode=ucs2 make make install
Once you have it installed, a ucs4 version will produce the following:
bash$ python Python 2.5.2 (r252:60911, Mar 4 2008, 10:58:18) [GCC 3.4.5] on linux >>>import sys >>> print sys.maxunicode 1114111
And, a ucs2 version will produce the following:
>>>import sys >>> print sys.maxunicode 65535
./configure --prefix=/your/install/dir/ --enable-shared make make install
I run a version of the Windows batch script below to create a regular backup of one of the folders that contains my daily work; for this illustration, we will call the source folder c:source. Every night, this batch file is triggered by Windows Scheduled Tasks to create a folder in my backup drive with the timestamp in the folder name; let us call this target folder D:target. Before creating this daily backup, the script also checks to see how many backup sets are already existing in the backup folder, and if above the threshhold, the oldest one(s) are deleted; this feature helps making sure the space usage of redundant backup sets do not get out of control. When scheduling this job with Windows Scheduled Tasks, you may wish to output to a file as well to create a log file of what old folders were removed and what files were copied (eg. “c:scriptsbackup.bat > logfile.log”).
@echo off
:: Start Variables
set NumberToKeep=5
set BackupCmd=xcopy /s /c /d /e /h /i /r /y
set BackupSource=C:\source\
set BackupTarget=D:\target\
:: End Variables
:: Actual Script Starts Here!
echo+
echo STARTING BACKUP
echo %date% %time%
:: 1. Delete older backup set(s) beyond the NumberToKeep
for /F "tokens=* skip=%NumberToKeep%" %%I In ('dir "%BackupTarget%" /AD /B /O-D /TW') do (
echo+
echo DELETING OLD BACKUP SET %BackupTarget%%%~I
rd /s /q "%BackupTarget%%%~I"
)
:: 2. Create new backup set
set bkuphour=%time:~0,2%
if "%bkuphour:~0,1%"==" " set bkuphour=0%time:~1,1%
set bkupfldr=%date:~10,4%_%date:~4,2%_%date:~7,2%_%bkuphour%_%time:~3,2%
echo+
echo CREATING FOLDER %BackupTarget%%bkupfldr%\
if not exist "%BackupTarget%%bkupfldr%\" mkdir "%BackupTarget%%bkupfldr%\"
echo+
echo BACKING UP FILES...
%BackupCmd% "%BackupSource%*.*" "%BackupTarget%%bkupfldr%\"
echo+
echo BACKUP COMPLETED!
echo %date% %time%
Hope you will find this useful!
Let’s suppose we have a simple array containing 5 strings. The .sort() method sorts them in ascending order and the .reverse() method sorts them in descending order, as demonstrated below.
var simpleArray1 = ["Paul", "Peter", "Mary"]; // .sort() will sort the array to become "Mary", "Paul", and then "Peter". simpleArray1.sort(); // .reverse() will sort the array to become "Peter", "Paul", and then "Mary". simpleArray1.reverse();
In the case of numbers, we should be aware that they are sorted in a similar fashion as strings. In the following example, 100 appears before 32 when using .sort() because 1 comes before 3.
var simpleArray2 = [100, 999, 32]; // .sort() will sort the array to become 100, 32, and then 999. simpleArray2.sort(); // .reverse() will sort the array to become 999, 32, and then 100. simpleArray2.reverse();
To sort an array of numbers by their numeric values, we should pass in a function into the .sort() method as follows.
// .sort(function(a, b) {return a - b}) will sort the array to become 32, 100, and then 999.
simpleArray2.sort(function(a, b) {return a - b});
// .sort(function(a, b) {return b - a}) will sort the array to become 999, 100, and then 32.
simpleArray2.sort(function(a, b) {return b - a});
We can sort an array of objects, ie. each element of the array has its own set of properties, on any of the properties as well. In the following example, each object in the array has a property called “id”, which we are using to sort.
// Ascending order
complexArray.sort(
function(a, b) {
return (a.id > b.id ? 1 : (a.id < b.id ? -1 : 0));
}
);
// Descending order
complexArray.sort(
function(a, b) {
return (a.id > b.id ? -1 : (a.id < b.id ? 1 : 0));
}
);
I often use CSV format to transport data, largely because its text nature makes the data extremely portable across different platforms and applications. Yesterday when a colleague opened a CSV file from me using Excel, one of the fields I provided was displayed incorrectly. Two examples of the troubles Excel are listed below.
| Value Provided in CSV | Shown by Excel |
| 0E36 | 0 |
| 0055 | 55 |
In the first case, Excel thought the value was a scientific notion; in the second case, Excel treated it as a number. In both cases, they were both wrong.
This issue also happens frequently when displaying ZIP codes for addresses in the United States, eg. Excel showing the intended “04001” as “4001”.
Because this is an Excel-specific behavior, I consider the following fix to be a work-around rather than a permanent solution; nevertheless, it works well for me. Instead of a simple CSV format as I had originally:
“aaa”,”bbb”,”ccc”,“0055”,”ddd”
I inserted an equal sign in front of it. This forces Excel to think that I am writing a formula that outputs a string, thus solving the Excel issue.
“aaa”,”bbb”,”ccc”,=”0055″,”ddd”
For this particular example, let us assume that the information being between application takes on the form of two fields, row_id and row_status. Given that, we will create this following object type, which will be used by the queue shortly.
create or replace type obj_stuff as object ( row_id number, row_status varchar2(10) );
With the object type created, we will first create a queue table based on this object type, followed by a queue.
declare begin dbms_aqadm.create_queue_table ( queue_table => 'myschema.stuff_queue_table', queue_payload_type => 'myschema.obj_stuff' ); dbms_aqadm.create_queue ( queue_name => 'myschema.stuff_queue', queue_table => 'myschema.stuff_queue_table' ); end;
With the script above, we now have the queue table and the queue set up. To start the queue, use the .start_queue script, as below.
declare begin dbms_aqadm.start_queue ( queue_name => 'myschema.stuff_queue', enqueue => true ); end;
To do so, we can utilize the timestampdiff() function provided by DB2. This function takes in two parameters. The first parameter is a numeric value indicating the unit in which you wish to receive the results in; the values must be one of the following.
The second parameter should contain the subtraction formula between two timestamps, with the results converted to character format. Below is one example of usage.
select
timestampdiff(
16,
char(timestamp('2010-01-11-15.01.33.453312') - current timestamp))
from sysibm.sysdummy1;
The result from this statement is 210 at the time of the writing. Notice that even though the first timestamp is set to be prior than the current timestamp, the outcome is still positive — This function returns the absolute value (ie. always positive) reflecting the difference in time between two timestamps. Also, take note that the result will always be an integer, thus it can only be considered an estimation of the date/time difference rather than an exact one.
In Oracle, the dummy table is “dual”. It is present in all Oracle database installations, thus no special setup is needed to use it. Below are some samples on how to use this table.
-- Getting the current system date:
PROMPT> select sysdate from dual;
SYSDATE
---------
06-AUG-10
-- Getting the current logged-in user:
PROMPT> select user from dual;
USER
------------------------------
YTCDBA
-- Display the result of some string manipulation
PROMPT> select substr('hello world',1,7) from dual;
SUBSTR(
-------
hello w
The DB2 equivalent of Oracle’s “dual” is “sysibm.sysdummy1”. Below is a sample of the usage.
-- Getting the current system date:
PROMPT> select current date from sysibm.sysdummy1;
00001
----------
2010-08-06
-- Getting the current logged-in user:
PROMPT> select current user from sysibm.sysdummy1;
00001
-----
USER1
-- Display the result of some string manipulation
PROMPT> select substr('hello world',1,7) from sysibm.sysdummy1;
00001
-------
hello w
It is also worthy of note that MySQL and SQL Server can select pseudocolumn data without the use of a dummy table. For example, in SQL Server, you can get the current date by running “select getdate();” without a “from” clause in your SQL statement.
Example: toProperCase(“hello john smith”) will return “Hello John Smith”.
public static String toProperCase(String inputString) {
String ret = "";
StringBuffer sb = new StringBuffer();
Matcher match = Pattern.compile("([a-z])([a-z]*)", Pattern.CASE_INSENSITIVE).matcher(inputString);
while (match.find()) {
match.appendReplacement(sb, match.group(1).toUpperCase() + match.group(2).toLowerCase()) ;
}
ret = match.appendTail(sb).toString();
return ret;
}
Example: roundToDecimals(1.23456, 2) will return 1.23.
public static float roundToDecimals(float fl, int roundToThisManyDecimalPoints) {
float ret = 0;
float pow = (float) Math.pow(10, roundToThisManyDecimalPoints);
fl = fl * pow;
float tmpFl = Math.round(fl);
ret = (float) tmpFl / pow;
return ret;
}