Getting Values out of a String with patterns - string

Question

I have textfiles, which have attributes saved in Strings. Those Strings have a pattern like this:
[attributeName]:[value]
I can't generalize the [value], because it could be of any primitive datatype.
Saving the effectively values is not my concern, because it's depending on the user which attribute has to be loaded. The same file won't be loaded very often.
Now I have 2 problems:
1) For some reason the program which creates those files sometimes adds spaces around the : at some attributes and [value] could also contain spaces, so I have to get rid of those
2) Making the reading of those attributes more performant:
I've come up with this method:
public String getAttribute(File file, String attribute)
{
try
{
BufferedReader reader = new BufferedReader(new FileReader(file), 1024);
String line;
Pattern p = Pattern.compile(Pattern.quote(attribute), Pattern.CASE_INSENSITIVE);
while ((line = reader.readLine()) != null)
{
int i = line.indexOf(":");
if(line.charAt(i-1) == ' ')
line = line.substring(0,i-2) + line.substring(i);
if(line.charAt(i+1) == ' ')
line = line.substring(0,i) + line.substring(i+2);
if (p.matcher(line).find())
{
return line.replace(attribute, "").trim();
}
}
} catch (IOException e)
{
e.printStackTrace();
}
return null;
}
However, this method will probably be one of the most called by my application, so I can't leave it so unperformant as it is right now,
Thanks for any help!

Answer 1

I modified code to find appropriate line. Check example code below.
If you have a lot of files and attributes in these files you could think about saving somewhere pair attribute=value in code. In example code I provided very primitive cache by using Table interface from guava library.
Example code:
# guava library
import com.google.common.collect.Table;
import com.google.common.collect.HashBasedTable;
# apache commons lang
import static org.apache.commons.lang.StringUtils.startsWithIgnoreCase;
# apache commons io
import static org.apache.commons.io.IOUtils.closeQuietly;
[...]
# very primitive cache implementation. To find a value in table you have to
# pass row and column keys. In this implementation row is equal to file
# absolute path (because you can have 2 files with the same name) and column
# is equal to attribute name.
# If you have a lot of files and attributes probably you will have to clear
# from time to time the cache otherwise you will get out of memory
private static final Table<String, String, String> CACHE = HashBasedTable.create();
[...]
public String getAttribute(File file, String attribute) {
# get value for the given attribute from the given file
String value = CACHE.get(file.getAbsolutePath(), attribute);
# if cache does not contain value, method will read value from file
if (null == value) {
BufferedReader reader = null;
String value = null;
try {
reader = new BufferedReader(new FileReader(file), 1024);
String line;
while ((line = reader.readLine()) != null) {
# From your description I understood that each line in file
# starts with attribute name
if (startsWithIgnoreCase(line, attribute) {
# if we found correct line we simple split it by ':'
String[] array = line.split(":");
# this is to ensure that line contains attribute name
# and value
if (array.length >= 2) {
# we found value for attribute and we remove spaces
value = array[1].trim();
# we put value to the cache to speed up finding
# value for the same attribute in the future
CACHE.put(file.getAbsolutePath(), attribute, value);
break;
}
}
}
} catch (IOException e) {
e.printStackTrace();
} finally {
# you should always close
closeQuietly(reader);
}
}
return value;
}