Use new path expansion implementation

Remove all uses of old path expansion implementation (expandPathToList)

Fixes #81 #77

Author: PerilousApricot

src/main/java/edu/vanderbilt/accre/laurelin/root_proxy/IOFactory.java

@@ -203,12 +203,4 @@ private static List<Path> resolveGlob(String path) throws IOException {
     private static boolean isGlob(String path) {
         return path.matches(".*[{}\\[\\]*?].*");
     }
-
-    public static List<String> expandPathToList(String path) throws IOException {
-        if (Pattern.matches(hadoopPattern,  path)) {
-            return HadoopFile.expandPathToList(path);
-        } else {
-            return NIOFile.expandPathToList(path);
-        }
-    }
-}
+}

src/main/java/edu/vanderbilt/accre/laurelin/root_proxy/ROOTFileCache.java

@@ -8,6 +8,7 @@
 import java.util.LinkedList;
 import java.util.concurrent.locks.ReentrantLock;
 
+import org.apache.hadoop.fs.Path;
 import org.apache.logging.log4j.LogManager;
 import org.apache.logging.log4j.Logger;
 
@@ -161,6 +162,10 @@ public void forciblyInvalidateOpenFile(String path) {
         }
     }
 
+    public ROOTFile getROOTFile(Path path) throws IOException {
+        return getROOTFile(path.toString());
+    }
+
     /**
      * Gets a ROOTFile pointing to the given URI, ensuring that they're unique
      * within the cache object and with a timeout to prevent immediately
@@ -228,6 +233,8 @@ public long getPhantomReferenceCount() {
         }
     }
 
+
+
     private ROOTFile load(String path) throws IOException {
         assert lock.isHeldByCurrentThread() == true;
         FileInterface fileInterface = timedCache.getIfPresent(path);
@@ -361,4 +368,5 @@ public FileInterface getIfPresent(String path) {
             return fiMap.get(path);
         }
     }
+
 }

src/main/java/edu/vanderbilt/accre/laurelin/spark_ttree/Reader.java

@@ -3,14 +3,15 @@
 import java.io.IOException;
 import java.io.Serializable;
 import java.util.ArrayList;
+import java.util.Arrays;
 import java.util.HashMap;
 import java.util.Iterator;
-import java.util.LinkedList;
 import java.util.List;
 import java.util.Map;
 import java.util.Map.Entry;
 import java.util.function.Function;
 
+import org.apache.hadoop.fs.Path;
 import org.apache.logging.log4j.LogManager;
 import org.apache.logging.log4j.Logger;
 import org.apache.spark.SparkContext;
@@ -47,7 +48,7 @@ public class Reader implements DataSourceReader,
         SupportsPushDownRequiredColumns {
     static final Logger logger = LogManager.getLogger();
 
-    private LinkedList<String> paths;
+    private List<String> paths;
     private String treeName;
     private TTree currTree;
     private TFile currFile;
@@ -62,9 +63,10 @@ public Reader(DataSourceOptions options, SparkContext sparkContext, CollectionAc
         logger.trace("construct ttreedatasourcev2reader");
         this.sparkContext = sparkContext;
         try {
-            this.paths = new LinkedList<String>();
-            for (String path: options.paths()) {
-                this.paths.addAll(IOFactory.expandPathToList(path));
+            List<Path> expanded = IOFactory.resolvePathList(Arrays.asList(options.paths()));
+            this.paths = new ArrayList<String>(expanded.size());
+            for (Path p: expanded) {
+                this.paths.add(p.toString());
             }
             // FIXME - More than one file, please
             currFile = TFile.getFromFile(fileCache.getROOTFile(this.paths.get(0)));

src/test/java/edu/vanderbilt/accre/root_proxy/IOTest.java

@@ -168,15 +168,15 @@ private static ByteBuffer getTestBytes(long off, int len) {
      */
     @Test
     public void searchDirectory_nio() throws IOException {
-        List<String> files = IOFactory.expandPathToList("testdata/recursive");
+        List<org.apache.hadoop.fs.Path> files = resolveHelper("testdata/recursive");
         assertEquals(3, files.size());
     }
 
     @Test
     public void searchDirectory_hadoop() throws IOException {
         Path currentRelativePath = Paths.get("");
         String s = currentRelativePath.toAbsolutePath().toString();
-        List<String> files = IOFactory.expandPathToList("file:///" + s + "/" + "testdata/recursive");
+        List<org.apache.hadoop.fs.Path> files = resolveHelper("file:///" + s + "/" + "testdata/recursive");
         assertEquals(3, files.size());
     }
 

src/test/java/edu/vanderbilt/accre/spark_ttree/TTreeDataSourceIntegrationTest.java

@@ -72,13 +72,15 @@ public void testVectorLoad() {
 
     @Test
     public void testTwoFiles() {
+        // If the files are duplicated in the input list, they shouldn't be
+        // after reading
         Dataset<Row> df = spark
                 .read()
                 .format("root")
                 .option("tree",  "Events")
                 .option("threadCount", "0")
                 .load("testdata/all-types.root", "testdata/all-types.root");
-        assertEquals(18, df.count());
+        assertEquals(9, df.count());
     }
 
     @AfterClass