[patch] Prevent NoSuchMethodError when compiling in IntelliJ IDEA

[Aleksey Shipilev]

Looks good. I shall reword some of the existing comments before the
integration. Will push as soon as OCA is processed.

Cheers,
-Aleksey

On 10/20/2015 12:25 PM, Evgeny Mandrikov wrote:
> Pff, looks like ML removes attachments sometimes. Inlined version:
> 
> diff -r 9b3dc57e8d41
> jmh-core/src/main/java/org/openjdk/jmh/generators/core/BenchmarkGenerator.java
> ---
> a/jmh-core/src/main/java/org/openjdk/jmh/generators/core/BenchmarkGenerator.java
>    Mon Oct 19 16:31:14 2015 +0300
> +++
> b/jmh-core/src/main/java/org/openjdk/jmh/generators/core/BenchmarkGenerator.java
>    Tue Oct 20 10:24:29 2015 +0200
> @@ -134,12 +134,14 @@
>              Set<BenchmarkListEntry> entries = new
> HashSet<BenchmarkListEntry>();
> 
>              // Try to read the benchmark entries from previous
> generator sessions.
> +            Multimap<String, BenchmarkListEntry>
> previousEntriesByUserClassQName = new HashMultimap<String,
> BenchmarkListEntry>();
>              try {
>                  Reader reader =
> destination.getResource(BenchmarkList.BENCHMARK_LIST.substring(1));
>                  Collection<String> existingLines =
> FileUtils.readAllLines(reader);
>                  for (String line : existingLines) {
>                      BenchmarkListEntry br = new BenchmarkListEntry(line);
>                      entries.add(br);
> +                  
>  previousEntriesByUserClassQName.put(br.getUserClassQName(), br);
>                  }
>              } catch (IOException e) {
>                  // Expected in most cases, move on.
> @@ -147,7 +149,7 @@
>                  destination.printWarning("Unable to read the existing
> benchmark list, because of UnsupportedOperationException. Run on JDK 7
> or higher.");
>              }
> 
> -            // Generate new benchmark entries, potentially overwriting
> the previous lines
> +            // Generate new benchmark entries, potentially overwriting
> all the previous lines belonging to the classes of the newly generated
> entries
>              for (BenchmarkInfo info : benchmarkInfos) {
>                  try {
>                      MethodGroup group = info.methodGroup;
> @@ -177,9 +179,10 @@
>                                  group.getTimeout()
>                          );
> 
> -                        if (entries.contains(br)) {
> -                            destination.printNote("Benchmark entry " +
> br + " already exists, overwriting");
> -                            entries.remove(br);
> +                        if
> (previousEntriesByUserClassQName.keys().contains(info.userClassQName)) {
> +                            destination.printNote("Benchmark entries
> for " + info.userClassQName + " already exist, overwriting");
> +                          
>  entries.removeAll(previousEntriesByUserClassQName.get(info.userClassQName));
> +                          
>  previousEntriesByUserClassQName.remove(info.userClassQName);
>                          }
> 
>                          entries.add(br);
> 
> On Tue, Oct 20, 2015 at 11:19 AM, Evgeny Mandrikov <mandrikov at gmail.com
> <mailto:mandrikov at gmail.com>> wrote:
> 
>     CLA has been signed and sent. Updated patch is in attachment.
>     However it lacks test, because I don't see how such scenario can be
>     easily tested using current infrastructure of tests for generator. I
>     thought about mocked ClassInfo, but maybe you have a better idea?
> 
> 
> -- 
> Best regards,
> Evgeny Mandrikov aka Godin <http://godin.net.ru>
> http://twitter.com/_godin_