UWECProgrammer/ConsolidatedMentionResolutionBenchmark

sentence stringlengths 1 1.38k	label stringclasses 3 values
I would suggest creating a thin logging facade that can write to any of the logging frameworks, at which point the choice of backing engine become pretty much a moot point.	o
That is what commons logging does so why reinvent the wheel?	o
Also there are some issues that your facade would have to handle that commons logging already does.	o
+1 To counter commons logging argument, Some enterprise apps use custom logger.	o
It always a good idea to hide underlying implementation.	p
The 'thin' wrapper eliminates the need for another jar being packaged and is not as much as reinventing.	o
This approach is valid in many cases, especially when your code must run on an arbitrary platform.	o
The underlying logging API tends to be different on application servers compared to Eclipse, for example.	o
About hiding the underlying implementation: All the logging APIs mentioned can do just that: They are interfaces to pluggable implementations.	o
This saved me recently when we ported our framework to the Compact Framework (in .Net).	o
If we had hardcoded dependencies on nLog, we would have been screwed.	n
Because we used this approach, we were able to drop in a null logger and be happy.Someone vote me back up to 0 Please :-).	o
One already exists - take a look at slf4j.	o
It's an accepted thin wrapper that allows you to switch logging frameworks on application startup by switching jars on the runtime classpath.	o
clogging has its own way of discovering what framework it will use - its not nice and rather convuluted.	o
How many logging frameworks has Ceki created.	o
One should always strive to slot a level of interdirection and hide an implementation even if it is clogging with your own logging.	o
Behind the scenes you can then plugin whatever f/w you desire as mentioned by Travis.	o
Java - Create a new String instance with specified length and filled with specific character.	o
Best solution?	o
I did check the other questions; this question has its focus on solving this particular question the most efficient way.	o
Sometimes you want to create a new string with a specified length, and with a default character filling the entire string.	o
ie, it would be cool if you could do CODETERM1 and create a new String from there, with a length of 10 characters all having a .	p
Because such a constructor does not exist, and you cannot extend from String, you have either to create a wrapper class or a method to do this for you.	o
At this moment I am using this: CODESNIPPET_JAVA1 .	o
It still lacks any checking (ie, when length is 0 it will not work).	n
I am constructing the array first because I believe it is faster than using string concatination or using a StringBuffer to do so.	p
Anyone else has a better sollution?	o
Simply use the StringUtils class from URL_http://commons.apache.org/lang [apache-commons-lang] project.	o
You have a URL_http://commons.apache.org/lang// api-2.4/org/apache/commons/lang/StringUtils.html#leftPad%28java.lang.String,%2 0int,%20char%29 [leftPad] method: CODESNIPPET_JAVA1 .	o
+1 - that's exactly what he was looking for (aka 'cool').	o
For the exact same behaviour, but that's obvious, do: StringUtils.leftPad("", 10, '*');.	o
+1 nice solution, though I wouldn't add a jar to my project just for that.	p
Thanks.	o
Yes that looks like a more suitable way to use it.	o
Luckily I am already using this jar, so i could use this straight away.	p
Or use repeat(), which doesn't require the empty string at the start and is arguably clearer in intent (see my answer).	p
@abyx> Yes, you will need to add a JAR, but commons lang offers so nice utils methods, so I don't think it is really a drawback!.	p
Over 300kb for this jar??	o
Yuck, no thanks.	o
Not for a mobile app.	o
No need for bloatware.	o
No need to do the loop, and using just standard Java library classes: CODESNIPPET_JAVA1 .	p
As you can see, I also added suitable code for the CODETERM1 case.	o
won't the code in Arrays.fill() contain a loop...	o
Maybe, but 1) it's likely to be faster and 2) that's code you don't need to write.	p
Good find.	p
Yes Arrays.fill does a for loop itself.	o
But fundementally this is better.	o
You should throw an IllegalArgumentException if (length < 0).	o
(Even java.lang.NegativeArraySizeException is better.	o
).	o
Apache Commons Lang (probably useful enough to be on the classpath of any non- trivial project) has URL_http://commons.apache.org/lang//api-2.4/org/apache/co mmons/lang/StringUtils.html#repeat%28java.lang.String,%20int%29 [StringUtils.repeat()] : CODESNIPPET_JAVA1 .	n
Easy!	p
CODESNIPPET_JAVA1 .	o
What is wrong?	o
The above is fine.	o
Do you mind if I ask you a question - Is this causing you a problem?	o
It seams to me you are optimizing before you know if you need to.	o
Now for my over engineered solution.	o
In many (thou not all) cases you can use CharSequence instead of a String.	o
CODESNIPPET_JAVA1 .	o
CODESNIPPET_JAVA1 .	o
One extra note: it seems that all public ways of creating a new CODETERM1 instance involves necessarily the copy of whatever buffer you are working with, be it a CODETERM2 , a CODETERM3 or a CODETERM4 .	p
From the CODETERM5 javadoc (and is repeated in the respective CODETERM6 methods from the other classes): _The contents of the character array are copied; subsequent modification of the character array does not affect the newly created string._ So you'll end up having a possibly big memory copy operation after the "fast filling" of the array.	n
The only solution that may avoid this issue is the one from @mlk, if you can manage working directly with the proposed CODETERM7 implementation (what may be the case).	p
PS: I would post this as a comment but I don't have enough reputation to do that yet.	p
using URL_http://bitbucket.org/dfa/dollar [Dollar] is simple: CODESNIPPET_JAVA1 .	p
There are 9 in the comment.	o
To improve performance you could have a single predefined sting if you know the max length like: String template "####################################"; And then simply perform a substring once you know the length.	n
Karl .	o
+1: I just read on another post that substring creates a String that just uses the same char[] of the original String.	o
So it's an O(1) operation (no loop) and it saves memory (if that is a real issue).	n
Try this Using the substring(int start, int end); method CODESNIPPET_JAVA1 .	o
this will return abcde.	o
C# equivalent to Java's charAt()?	o
I know we can use the CODETERM1 method in Java get an individual character in a string by specifying its position.	o
Is there an equivalent method in C#?	o
You can index into a string in C# like an array, and you get the character at that index.	o
Example: In Java, you would say str.charAt(8);In C#, you would say str[8]; .	o
CODESNIPPET_JAVA1 .	o
And <strike> CODETERM1 Reference: URL_http://msdn.microsoft.com/en- us/library/system.string.chars%28v=VS.71%29.aspx [ URL_http://msdn.microsoft.com /en-us/library/system.string.chars%28v=VS.71%29.aspx] </strike> The above is same as using indexers in c#.	o
You beat me to it -- you get the upvote.	p
I really don't know if it helps anything at all but I'll be damned if I don't try.	n
I had a similar problem when I required a specific position/index, and LINQ Saved the day actually, I just wish I remembered where I put that blasted code.	o
Sorry for the crappy answer, but it might be a good starting point if ALL ELSE fails.	n
Console.WriteLine allows the user to specify a position in a string.	o
See sample: string str "Tigger";Console.WriteLine( str[0] ); //returns "T";Console.WriteLine( str[2] ); //returns "g"; There you go!	o
this topic, and your answer, have nothing to do with WriteLine...	o
String, StringBuffer, and StringBuilder.	o
Please tell me a real time situation to compare CODETERM1 , CODETERM2 , and CODETERM3 ?	o
Difference:** CODETERM1 is if you try to alter their values, another object gets created, whereas CODETERM2 and CODETERM3 are so they can change their values.	o
Difference:** The difference between CODETERM4 and CODETERM5 is that CODETERM6 is thread-safe.	p
So when the application needs to be run only in a single thread then it is better to use CODETERM7 .	o
CODETERM8 is more efficient than CODETERM9 .	p
If your string is not going to change use a String class because a CODETERM10 object is immutable.	o
If your string can change (example: lots of logic and operations in the construction of the string) and will only be accessed from a single thread, using a CODETERM11 is good enough.	p
If your string can change, and will be accessed from multiple threads, use a CODETERM12 because CODETERM13 is synchronous so you have thread-safety.	p
Also, using String for logic operations is rather slow, and is not advised at all, since the JVM converts the String to a StringBuffer in the bytecode.	n
A lot of overhead is wasted converting from String to StringBuffer and then back to String again.	o

I would suggest creating a thin logging facade that can write to any of the logging frameworks, at which point the choice of backing engine become pretty much a moot point.

o

That is what commons logging does so why reinvent the wheel?

o

Also there are some issues that your facade would have to handle that commons logging already does.

o

+1 To counter commons logging argument, Some enterprise apps use custom logger.

o

It always a good idea to hide underlying implementation.

p

The 'thin' wrapper eliminates the need for another jar being packaged and is not as much as reinventing.

o

This approach is valid in many cases, especially when your code must run on an arbitrary platform.

o

The underlying logging API tends to be different on application servers compared to Eclipse, for example.

o

About hiding the underlying implementation: All the logging APIs mentioned can do just that: They are interfaces to pluggable implementations.

o

This saved me recently when we ported our framework to the Compact Framework (in .Net).

o

If we had hardcoded dependencies on nLog, we would have been screwed.

n

Because we used this approach, we were able to drop in a null logger and be happy.Someone vote me back up to 0 Please :-).

o

One already exists - take a look at slf4j.

o

It's an accepted thin wrapper that allows you to switch logging frameworks on application startup by switching jars on the runtime classpath.

o

clogging has its own way of discovering what framework it will use - its not nice and rather convuluted.

o

How many logging frameworks has Ceki created.

o

One should always strive to slot a level of interdirection and hide an implementation even if it is clogging with your own logging.

o

Behind the scenes you can then plugin whatever f/w you desire as mentioned by Travis.

o

Java - Create a new String instance with specified length and filled with specific character.

o

Best solution?

o

I did check the other questions; this question has its focus on solving this particular question the most efficient way.

o

Sometimes you want to create a new string with a specified length, and with a default character filling the entire string.

o

ie, it would be cool if you could do CODETERM1 and create a new String from there, with a length of 10 characters all having a .

p

Because such a constructor does not exist, and you cannot extend from String, you have either to create a wrapper class or a method to do this for you.

o

At this moment I am using this: CODESNIPPET_JAVA1 .

o

It still lacks any checking (ie, when length is 0 it will not work).

n

I am constructing the array first because I believe it is faster than using string concatination or using a StringBuffer to do so.

p

Anyone else has a better sollution?

o

Simply use the StringUtils class from URL_http://commons.apache.org/lang [apache-commons-lang] project.

o

You have a URL_http://commons.apache.org/lang// api-2.4/org/apache/commons/lang/StringUtils.html#leftPad%28java.lang.String,%2 0int,%20char%29 [leftPad] method: CODESNIPPET_JAVA1 .

o

+1 - that's exactly what he was looking for (aka 'cool').

o

For the exact same behaviour, but that's obvious, do: StringUtils.leftPad("", 10, '*');.

o

+1 nice solution, though I wouldn't add a jar to my project just for that.

p

Thanks.

o

Yes that looks like a more suitable way to use it.

o

Luckily I am already using this jar, so i could use this straight away.

p

Or use repeat(), which doesn't require the empty string at the start and is arguably clearer in intent (see my answer).

p

@abyx> Yes, you will need to add a JAR, but commons lang offers so nice utils methods, so I don't think it is really a drawback!.

p

Over 300kb for this jar??

o

Yuck, no thanks.

o

Not for a mobile app.

o

No need for bloatware.

o

No need to do the loop, and using just standard Java library classes: CODESNIPPET_JAVA1 .

p

As you can see, I also added suitable code for the CODETERM1 case.

o

won't the code in Arrays.fill() contain a loop...

o

Maybe, but 1) it's likely to be faster and 2) that's code you don't need to write.

p

Good find.

p

Yes Arrays.fill does a for loop itself.

o

But fundementally this is better.

o

You should throw an IllegalArgumentException if (length < 0).

o

(Even java.lang.NegativeArraySizeException is better.

o

).

o

Apache Commons Lang (probably useful enough to be on the classpath of any non- trivial project) has URL_http://commons.apache.org/lang//api-2.4/org/apache/co mmons/lang/StringUtils.html#repeat%28java.lang.String,%20int%29 [StringUtils.repeat()] : CODESNIPPET_JAVA1 .

n

Easy!

p

CODESNIPPET_JAVA1 .

o

What is wrong?

o

The above is fine.

o

Do you mind if I ask you a question - Is this causing you a problem?

o

It seams to me you are optimizing before you know if you need to.

o

Now for my over engineered solution.

o

In many (thou not all) cases you can use CharSequence instead of a String.

o

CODESNIPPET_JAVA1 .

o

CODESNIPPET_JAVA1 .

o

One extra note: it seems that all public ways of creating a new CODETERM1 instance involves necessarily the copy of whatever buffer you are working with, be it a CODETERM2 , a CODETERM3 or a CODETERM4 .

p

From the CODETERM5 javadoc (and is repeated in the respective CODETERM6 methods from the other classes): _The contents of the character array are copied; subsequent modification of the character array does not affect the newly created string._ So you'll end up having a possibly big memory copy operation after the "fast filling" of the array.

n

The only solution that may avoid this issue is the one from @mlk, if you can manage working directly with the proposed CODETERM7 implementation (what may be the case).

p

PS: I would post this as a comment but I don't have enough reputation to do that yet.

p

using URL_http://bitbucket.org/dfa/dollar [Dollar] is simple: CODESNIPPET_JAVA1 .

p

There are 9 in the comment.

o

To improve performance you could have a single predefined sting if you know the max length like: String template "####################################"; And then simply perform a substring once you know the length.

n

Karl .

o

+1: I just read on another post that substring creates a String that just uses the same char[] of the original String.

o

So it's an O(1) operation (no loop) and it saves memory (if that is a real issue).

n

Try this Using the substring(int start, int end); method CODESNIPPET_JAVA1 .

o

this will return abcde.

o

C# equivalent to Java's charAt()?

o

I know we can use the CODETERM1 method in Java get an individual character in a string by specifying its position.

o

Is there an equivalent method in C#?

o

You can index into a string in C# like an array, and you get the character at that index.

o

Example: In Java, you would say str.charAt(8);In C#, you would say str[8]; .

o

CODESNIPPET_JAVA1 .

o

And <strike> CODETERM1 Reference: URL_http://msdn.microsoft.com/en- us/library/system.string.chars%28v=VS.71%29.aspx [ URL_http://msdn.microsoft.com /en-us/library/system.string.chars%28v=VS.71%29.aspx] </strike> The above is same as using indexers in c#.

o

You beat me to it -- you get the upvote.

p

I really don't know if it helps anything at all but I'll be damned if I don't try.

n

I had a similar problem when I required a specific position/index, and LINQ Saved the day actually, I just wish I remembered where I put that blasted code.

o

Sorry for the crappy answer, but it might be a good starting point if ALL ELSE fails.

n

Console.WriteLine allows the user to specify a position in a string.

o

See sample: string str "Tigger";Console.WriteLine( str[0] ); //returns "T";Console.WriteLine( str[2] ); //returns "g"; There you go!

o

this topic, and your answer, have nothing to do with WriteLine...

o

String, StringBuffer, and StringBuilder.

o

Please tell me a real time situation to compare CODETERM1 , CODETERM2 , and CODETERM3 ?

o

Difference:** CODETERM1 is if you try to alter their values, another object gets created, whereas CODETERM2 and CODETERM3 are so they can change their values.

o

Difference:** The difference between CODETERM4 and CODETERM5 is that CODETERM6 is thread-safe.

p

So when the application needs to be run only in a single thread then it is better to use CODETERM7 .

o

CODETERM8 is more efficient than CODETERM9 .

p

If your string is not going to change use a String class because a CODETERM10 object is immutable.

o

If your string can change (example: lots of logic and operations in the construction of the string) and will only be accessed from a single thread, using a CODETERM11 is good enough.

p

If your string can change, and will be accessed from multiple threads, use a CODETERM12 because CODETERM13 is synchronous so you have thread-safety.

p

Also, using String for logic operations is rather slow, and is not advised at all, since the JVM converts the String to a StringBuffer in the bytecode.

n

A lot of overhead is wasted converting from String to StringBuffer and then back to String again.

o