Bug 481272 - another case of "hidden" hardcoding.
Change-Id: Ia117a2119bdde91909674ddaabfe147bc04bbc1f
diff --git a/bundles/org.eclipse.test.performance/src/org/eclipse/test/internal/performance/db/DB.java b/bundles/org.eclipse.test.performance/src/org/eclipse/test/internal/performance/db/DB.java
index 954200a..a51e501 100644
--- a/bundles/org.eclipse.test.performance/src/org/eclipse/test/internal/performance/db/DB.java
+++ b/bundles/org.eclipse.test.performance/src/org/eclipse/test/internal/performance/db/DB.java
@@ -54,8 +54,13 @@
return name;
}
}
+ // dw 7/16/2016: assuming this pattern is supposed to be JUST the date part of build id
+ // yyyymmdd
+ // There are probably much better patterns than even my "quick and dirty" method?
+ final Pattern pattern = Pattern.compile("20[0-9][0-9][01][0-9][0-3][0-9]"); //$NON-NLS-1$
+// This poorly hardcoded pattern could have one of the big limitations in finding the "right" version to use in plots?
+// final Pattern pattern = Pattern.compile("200[3-9][01][0-9][0-3][0-9]"); //$NON-NLS-1$
- final Pattern pattern = Pattern.compile("200[3-9][01][0-9][0-3][0-9]"); //$NON-NLS-1$
final Matcher matcher = pattern.matcher(name);
if (!matcher.find()) {
