RFR: CODETOOLS-7902217: use try-with-resources where possible

[dmitry.bessonov at oracle.com]

On 13/07/2018 22:44, dmitry.bessonov at oracle.com wrote:
> 
> 
> On 13/07/2018 19:17, Jonathan Gibbons wrote:
>>
>>
>> On 07/13/2018 10:30 AM, dmitry.bessonov at oracle.com wrote:
>>> Please review the following fix related to Automatic Resource 
>>> Management applied to JTHarness code base.
>>>
>>> JBS: https://bugs.openjdk.java.net/browse/CODETOOLS-7902217
>>> Webrev: http://cr.openjdk.java.net/~dbessono/7902217/webrev/
>>>
>>> dmitry
>>
>> Generally good.
>>
>> As always, there's more that can be done:
>>
>> Lots of places where you could use Files.new* API.  E.g. lines like this
>>
>>   489         try (Writer writer = new BufferedWriter(new 
>> OutputStreamWriter(new FileOutputStream(index), 
>> StandardCharsets.UTF_8))) {
>>
>>
>>
>> There are also some potential sites that were not caught by the IDE, 
>> such as in Desktop.java and ConfigManager.java
>>
>>   971             try (FileInputStream fis = new FileInputStream(file);
>>   972                 InputStream in = new BufferedInputStream(fis)) {
>>   973                 Properties props = new Properties();
>>   974                 props.load(in);
>>   975                 in.close();
>>   976                 return props;
>>   977             }
> 
> Applied try-with-resources more extensively and cleaned up some 
> double-closing:
> 
> JBS: https://bugs.openjdk.java.net/browse/CODETOOLS-7902218
> Webrev: http://cr.openjdk.java.net/~dbessono/7902218/webrev/

A corrected patch: http://cr.openjdk.java.net/~dbessono/7902218/01/webrev/

ConfigManager, line 1051 - close() call was deleted without moving 
stream instance inside try-with-resources block.

dmitry

> 
> dmitry
> 
> 
>>
>>
>> -- Jon