/* * Licensed to the Apache Software Foundation (ASF) under one or more * contributor license agreements. See the NOTICE file distributed with * this work for additional information regarding copyright ownership. * The ASF licenses this file to You under the Apache License, Version 2.0 * (the "License"); you may not use this file except in compliance with * the License. You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ using System; using Document = Lucene.Net.Documents.Document; using FieldSelector = Lucene.Net.Documents.FieldSelector; using Similarity = Lucene.Net.Search.Similarity; using Directory = Lucene.Net.Store.Directory; using FSDirectory = Lucene.Net.Store.FSDirectory; using IndexInput = Lucene.Net.Store.IndexInput; using Lock = Lucene.Net.Store.Lock; namespace Lucene.Net.Index { /// IndexReader is an abstract class, providing an interface for accessing an /// index. Search of an index is done entirely through this abstract interface, /// so that any subclass which implements it is searchable. ///

Concrete subclasses of IndexReader are usually constructed with a call to /// one of the static open() methods, e.g. {@link #Open(String)}. ///

For efficiency, in this API documents are often referred to via /// document numbers, non-negative integers which each name a unique /// document in the index. These document numbers are ephemeral--they may change /// as documents are added to and deleted from an index. Clients should thus not /// rely on a given document having the same number between sessions. ///

An IndexReader can be opened on a directory for which an IndexWriter is /// opened already, but it cannot be used to delete documents from the index then. ///

/// Doug Cutting /// /// $Id: IndexReader.java 497612 2007-01-18 22:47:03Z mikemccand $ /// public abstract class IndexReader { private class AnonymousClassFindSegmentsFile : SegmentInfos.FindSegmentsFile { private void InitBlock(bool closeDirectory) { this.closeDirectory = closeDirectory; } private bool closeDirectory; internal AnonymousClassFindSegmentsFile(bool closeDirectory, Lucene.Net.Store.Directory Param1) : base(Param1) { InitBlock(closeDirectory); } public override System.Object DoBody(System.String segmentFileName) { SegmentInfos infos = new SegmentInfos(); infos.Read(directory, segmentFileName); if (infos.Count == 1) { // index is optimized return SegmentReader.Get(infos, infos.Info(0), closeDirectory); } else { // To reduce the chance of hitting FileNotFound // (and having to retry), we open segments in // reverse because IndexWriter merges & deletes // the newest segments first. IndexReader[] readers = new IndexReader[infos.Count]; for (int i = infos.Count - 1; i >= 0; i--) { try { readers[i] = SegmentReader.Get(infos.Info(i)); } catch (System.IO.IOException e) { // Close all readers we had opened: for (i++; i < infos.Count; i++) { readers[i].Close(); } throw e; } } return new MultiReader(directory, infos, closeDirectory, readers); } } } private class AnonymousClassFindSegmentsFile1 : SegmentInfos.FindSegmentsFile { internal AnonymousClassFindSegmentsFile1(System.IO.FileInfo Param1):base(Param1) { } public override System.Object DoBody(System.String segmentFileName) { return (long) FSDirectory.FileModified(fileDirectory, segmentFileName); } } private class AnonymousClassFindSegmentsFile2 : SegmentInfos.FindSegmentsFile { private void InitBlock(Lucene.Net.Store.Directory directory2) { this.directory2 = directory2; } private Lucene.Net.Store.Directory directory2; internal AnonymousClassFindSegmentsFile2(Lucene.Net.Store.Directory directory2, Lucene.Net.Store.Directory Param1):base(Param1) { InitBlock(directory2); } public override System.Object DoBody(System.String segmentFileName) { return (long) directory2.FileModified(segmentFileName); } } public sealed class FieldOption { private System.String option; internal FieldOption() { } internal FieldOption(System.String option) { this.option = option; } public override System.String ToString() { return this.option; } // all fields public static readonly FieldOption ALL = new FieldOption("ALL"); // all indexed fields public static readonly FieldOption INDEXED = new FieldOption("INDEXED"); // all fields which are not indexed public static readonly FieldOption UNINDEXED = new FieldOption("UNINDEXED"); // all fields which are indexed with termvectors enables public static readonly FieldOption INDEXED_WITH_TERMVECTOR = new FieldOption("INDEXED_WITH_TERMVECTOR"); // all fields which are indexed but don't have termvectors enabled public static readonly FieldOption INDEXED_NO_TERMVECTOR = new FieldOption("INDEXED_NO_TERMVECTOR"); // all fields where termvectors are enabled. Please note that only standard termvector fields are returned public static readonly FieldOption TERMVECTOR = new FieldOption("TERMVECTOR"); // all field with termvectors wiht positions enabled public static readonly FieldOption TERMVECTOR_WITH_POSITION = new FieldOption("TERMVECTOR_WITH_POSITION"); // all fields where termvectors with offset position are set public static readonly FieldOption TERMVECTOR_WITH_OFFSET = new FieldOption("TERMVECTOR_WITH_OFFSET"); // all fields where termvectors with offset and position values set public static readonly FieldOption TERMVECTOR_WITH_POSITION_OFFSET = new FieldOption("TERMVECTOR_WITH_POSITION_OFFSET"); } /// Constructor used if IndexReader is not owner of its directory. /// This is used for IndexReaders that are used within other IndexReaders that take care or locking directories. /// /// /// Directory where IndexReader files reside. /// protected internal IndexReader(Directory directory) { this.directory = directory; } /// Constructor used if IndexReader is owner of its directory. /// If IndexReader is owner of its directory, it locks its directory in case of write operations. /// /// /// Directory where IndexReader files reside. /// /// Used for write-l /// /// closeDirectory /// internal IndexReader(Directory directory, SegmentInfos segmentInfos, bool closeDirectory) { Init(directory, segmentInfos, closeDirectory, true); } internal virtual void Init(Directory directory, SegmentInfos segmentInfos, bool closeDirectory, bool directoryOwner) { this.directory = directory; this.segmentInfos = segmentInfos; this.directoryOwner = directoryOwner; this.closeDirectory = closeDirectory; } private Directory directory; private bool directoryOwner; private bool closeDirectory; protected internal IndexFileDeleter deleter; private SegmentInfos segmentInfos; private Lock writeLock; private bool stale; private bool hasChanges; /// Used by commit() to record pre-commit state in case /// rollback is necessary /// private bool rollbackHasChanges; private SegmentInfos rollbackSegmentInfos; /// Returns an IndexReader reading the index in an FSDirectory in the named /// path. /// public static IndexReader Open(System.String path) { return Open(FSDirectory.GetDirectory(path), true); } /// Returns an IndexReader reading the index in an FSDirectory in the named /// path. /// public static IndexReader Open(System.IO.FileInfo path) { return Open(FSDirectory.GetDirectory(path), true); } /// Returns an IndexReader reading the index in the given Directory. public static IndexReader Open(Directory directory) { return Open(directory, false); } private static IndexReader Open(Directory directory, bool closeDirectory) { return (IndexReader) new AnonymousClassFindSegmentsFile(closeDirectory, directory).run(); } /// Returns the directory this index resides in. public virtual Directory Directory() { return directory; } /// Returns the time the index in the named directory was last modified. /// Do not use this to check whether the reader is still up-to-date, use /// {@link #IsCurrent()} instead. /// public static long LastModified(System.String directory) { return LastModified(new System.IO.FileInfo(directory)); } /// Returns the time the index in the named directory was last modified. /// Do not use this to check whether the reader is still up-to-date, use /// {@link #IsCurrent()} instead. /// public static long LastModified(System.IO.FileInfo fileDirectory) { return (long) ((System.Int64) new AnonymousClassFindSegmentsFile1(fileDirectory).run()); } /// Returns the time the index in the named directory was last modified. /// Do not use this to check whether the reader is still up-to-date, use /// {@link #IsCurrent()} instead. /// public static long LastModified(Directory directory2) { return (long) ((System.Int64) new AnonymousClassFindSegmentsFile2(directory2, directory2).run()); } /// Reads version number from segments files. The version number is /// initialized with a timestamp and then increased by one for each change of /// the index. /// /// /// where the index resides. /// /// version number. /// /// IOException if segments file cannot be read public static long GetCurrentVersion(System.String directory) { return GetCurrentVersion(new System.IO.FileInfo(directory)); } /// Reads version number from segments files. The version number is /// initialized with a timestamp and then increased by one for each change of /// the index. /// /// /// where the index resides. /// /// version number. /// /// IOException if segments file cannot be read public static long GetCurrentVersion(System.IO.FileInfo directory) { Directory dir = FSDirectory.GetDirectory(directory); long version = GetCurrentVersion(dir); dir.Close(); return version; } /// Reads version number from segments files. The version number is /// initialized with a timestamp and then increased by one for each change of /// the index. /// /// /// where the index resides. /// /// version number. /// /// IOException if segments file cannot be read. public static long GetCurrentVersion(Directory directory) { return SegmentInfos.ReadCurrentVersion(directory); } /// Version number when this IndexReader was opened. public virtual long GetVersion() { return segmentInfos.GetVersion(); } /// Check whether this IndexReader still works on a current version of the index. /// If this is not the case you will need to re-open the IndexReader to /// make sure you see the latest changes made to the index. /// /// /// IOException public virtual bool IsCurrent() { return SegmentInfos.ReadCurrentVersion(directory) == segmentInfos.GetVersion(); } /// Checks is the index is optimized (if it has a single segment and no deletions) /// true if the index is optimized; false otherwise /// public virtual bool IsOptimized() { return segmentInfos.Count == 1 && HasDeletions() == false; } /// Return an array of term frequency vectors for the specified document. /// The array contains a vector for each vectorized field in the document. /// Each vector contains terms and frequencies for all terms in a given vectorized field. /// If no such fields existed, the method returns null. The term vectors that are /// returned my either be of type TermFreqVector or of type TermPositionsVector if /// positions or offsets have been stored. /// /// /// document for which term frequency vectors are returned /// /// array of term frequency vectors. May be null if no term vectors have been /// stored for the specified document. /// /// IOException if index cannot be accessed /// /// abstract public TermFreqVector[] GetTermFreqVectors(int docNumber); /// Return a term frequency vector for the specified document and field. The /// returned vector contains terms and frequencies for the terms in /// the specified field of this document, if the field had the storeTermVector /// flag set. If termvectors had been stored with positions or offsets, a /// TermPositionsVector is returned. /// /// /// document for which the term frequency vector is returned /// /// field for which the term frequency vector is returned. /// /// term frequency vector May be null if field does not exist in the specified /// document or term vector was not stored. /// /// IOException if index cannot be accessed /// /// abstract public TermFreqVector GetTermFreqVector(int docNumber, System.String field); /// Returns true if an index exists at the specified directory. /// If the directory does not exist or if there is no index in it. /// false is returned. /// /// the directory to check for an index /// /// true if an index exists; false otherwise /// public static bool IndexExists(System.String directory) { return IndexExists(new System.IO.FileInfo(directory)); } /// Returns true if an index exists at the specified directory. /// If the directory does not exist or if there is no index in it. /// /// the directory to check for an index /// /// true if an index exists; false otherwise /// public static bool IndexExists(System.IO.FileInfo directory) { if (System.IO.Directory.Exists(directory.FullName)) { return SegmentInfos.GetCurrentSegmentGeneration(System.IO.Directory.GetFileSystemEntries(directory.FullName)) != - 1; } else { return false; } } /// Returns true if an index exists at the specified directory. /// If the directory does not exist or if there is no index in it. /// /// the directory to check for an index /// /// true if an index exists; false otherwise /// /// IOException if there is a problem with accessing the index public static bool IndexExists(Directory directory) { return SegmentInfos.GetCurrentSegmentGeneration(directory) != - 1; } /// Returns the number of documents in this index. public abstract int NumDocs(); /// Returns one greater than the largest possible document number. /// This may be used to, e.g., determine how big to allocate an array which /// will have an element for every document number in an index. /// public abstract int MaxDoc(); /// Returns the stored fields of the nth /// Document in this index. /// public virtual Document Document(int n) { return Document(n, null); } /// Get the {@link Lucene.Net.Documents.Document} at the nth position. The {@link Lucene.Net.Documents.FieldSelector} /// may be used to determine what {@link Lucene.Net.Documents.Field}s to load and how they should be loaded. /// /// NOTE: If this Reader (more specifically, the underlying {@link FieldsReader} is closed before the lazy {@link Lucene.Net.Documents.Field} is /// loaded an exception may be thrown. If you want the value of a lazy {@link Lucene.Net.Documents.Field} to be available after closing you must /// explicitly load it or fetch the Document again with a new loader. /// /// /// /// Get the document at the nth position /// /// The {@link Lucene.Net.Documents.FieldSelector} to use to determine what Fields should be loaded on the Document. May be null, in which case all Fields will be loaded. /// /// The stored fields of the {@link Lucene.Net.Documents.Document} at the nth position /// /// IOException If there is a problem reading this document /// /// /// /// /// /// /// /// /// /// //When we convert to JDK 1.5 make this Set public abstract Document Document(int n, FieldSelector fieldSelector); /// Returns true if document n has been deleted public abstract bool IsDeleted(int n); /// Returns true if any documents have been deleted public abstract bool HasDeletions(); /// Returns true if there are norms stored for this field. public virtual bool HasNorms(System.String field) { // backward compatible implementation. // SegmentReader has an efficient implementation. return Norms(field) != null; } /// Returns the byte-encoded normalization factor for the named field of /// every document. This is used by the search code to score documents. /// /// /// /// public abstract byte[] Norms(System.String field); /// Reads the byte-encoded normalization factor for the named field of every /// document. This is used by the search code to score documents. /// /// /// /// public abstract void Norms(System.String field, byte[] bytes, int offset); /// Expert: Resets the normalization factor for the named field of the named /// document. The norm represents the product of the field's {@link /// Fieldable#SetBoost(float) boost} and its {@link Similarity#LengthNorm(String, /// int) length normalization}. Thus, to preserve the length normalization /// values when resetting this, one should base the new value upon the old. /// /// /// /// /// /// public void SetNorm(int doc, System.String field, byte value_Renamed) { lock (this) { if (directoryOwner) AquireWriteLock(); hasChanges = true; DoSetNorm(doc, field, value_Renamed); } } /// Implements setNorm in subclass. protected internal abstract void DoSetNorm(int doc, System.String field, byte value_Renamed); /// Expert: Resets the normalization factor for the named field of the named /// document. /// /// /// /// /// /// public virtual void SetNorm(int doc, System.String field, float value_Renamed) { SetNorm(doc, field, Similarity.EncodeNorm(value_Renamed)); } /// Returns an enumeration of all the terms in the index. /// The enumeration is ordered by Term.compareTo(). Each term /// is greater than all that precede it in the enumeration. /// public abstract TermEnum Terms(); /// Returns an enumeration of all terms after a given term. /// The enumeration is ordered by Term.compareTo(). Each term /// is greater than all that precede it in the enumeration. /// public abstract TermEnum Terms(Term t); /// Returns the number of documents containing the term t. public abstract int DocFreq(Term t); /// Returns an enumeration of all the documents which contain /// term. For each document, the document number, the frequency of /// the term in that document is also provided, for use in search scoring. /// Thus, this method implements the mapping: ///

///

The enumeration is ordered by document number. Each document number /// is greater than all that precede it in the enumeration. ///

public virtual TermDocs TermDocs(Term term) { TermDocs termDocs = TermDocs(); termDocs.Seek(term); return termDocs; } /// Returns an unpositioned {@link TermDocs} enumerator. public abstract TermDocs TermDocs(); /// Returns an enumeration of all the documents which contain /// term. For each document, in addition to the document number /// and frequency of the term in that document, a list of all of the ordinal /// positions of the term in the document is available. Thus, this method /// implements the mapping: /// ///

///

This positional information faciliates phrase and proximity searching. ///

The enumeration is ordered by document number. Each document number is /// greater than all that precede it in the enumeration. ///

public virtual TermPositions TermPositions(Term term) { TermPositions termPositions = TermPositions(); termPositions.Seek(term); return termPositions; } /// Returns an unpositioned {@link TermPositions} enumerator. public abstract TermPositions TermPositions(); /// Tries to acquire the WriteLock on this directory. /// this method is only valid if this IndexReader is directory owner. /// /// /// IOException If WriteLock cannot be acquired. private void AquireWriteLock() { if (stale) throw new System.IO.IOException("IndexReader out of date and no longer valid for delete, undelete, or setNorm operations"); if (this.writeLock == null) { Lock writeLock = directory.MakeLock(IndexWriter.WRITE_LOCK_NAME); if (!writeLock.Obtain(IndexWriter.WRITE_LOCK_TIMEOUT)) // obtain write lock { throw new System.IO.IOException("Index locked for write: " + writeLock); } this.writeLock = writeLock; // we have to check whether index has changed since this reader was opened. // if so, this reader is no longer valid for deletion if (SegmentInfos.ReadCurrentVersion(directory) > segmentInfos.GetVersion()) { stale = true; this.writeLock.Release(); this.writeLock = null; throw new System.IO.IOException("IndexReader out of date and no longer valid for delete, undelete, or setNorm operations"); } } } /// Deletes the document numbered docNum. Once a document is /// deleted it will not appear in TermDocs or TermPostitions enumerations. /// Attempts to read its field with the {@link #document} /// method will result in an error. The presence of this document may still be /// reflected in the {@link #docFreq} statistic, though /// this will be corrected eventually as the index is further modified. /// public void DeleteDocument(int docNum) { lock (this) { if (directoryOwner) AquireWriteLock(); hasChanges = true; DoDelete(docNum); } } /// Implements deletion of the document numbered docNum. /// Applications should call {@link #DeleteDocument(int)} or {@link #DeleteDocuments(Term)}. /// protected internal abstract void DoDelete(int docNum); /// Deletes all documents that have a given term indexed. /// This is useful if one uses a document field to hold a unique ID string for /// the document. Then to delete such a document, one merely constructs a /// term with the appropriate field and the unique ID string as its text and /// passes it to this method. /// See {@link #DeleteDocument(int)} for information about when this deletion will /// become effective. /// /// the number of documents deleted /// public int DeleteDocuments(Term term) { TermDocs docs = TermDocs(term); if (docs == null) return 0; int n = 0; try { while (docs.Next()) { DeleteDocument(docs.Doc()); n++; } } finally { docs.Close(); } return n; } /// Undeletes all documents currently marked as deleted in this index. public void UndeleteAll() { lock (this) { if (directoryOwner) AquireWriteLock(); hasChanges = true; DoUndeleteAll(); } } /// Implements actual undeleteAll() in subclass. protected internal abstract void DoUndeleteAll(); /// Should internally checkpoint state that will change /// during commit so that we can rollback if necessary. /// internal virtual void StartCommit() { if (directoryOwner) { rollbackSegmentInfos = (SegmentInfos) segmentInfos.Clone(); } rollbackHasChanges = hasChanges; } /// Rolls back state to just before the commit (this is /// called by commit() if there is some exception while /// committing). /// internal virtual void RollbackCommit() { if (directoryOwner) { for (int i = 0; i < segmentInfos.Count; i++) { // Rollback each segmentInfo. Because the // SegmentReader holds a reference to the // SegmentInfo we can't [easily] just replace // segmentInfos, so we reset it in place instead: segmentInfos.Info(i).Reset(rollbackSegmentInfos.Info(i)); } rollbackSegmentInfos = null; } hasChanges = rollbackHasChanges; } /// Commit changes resulting from delete, undeleteAll, or /// setNorm operations /// /// If an exception is hit, then either no changes or all /// changes will have been committed to the index /// (transactional semantics). /// /// /// IOException public void Commit() { lock (this) { if (hasChanges) { if (deleter == null) { // In the MultiReader case, we share this deleter // across all SegmentReaders: SetDeleter(new IndexFileDeleter(segmentInfos, directory)); } if (directoryOwner) { // Should not be necessary: no prior commit should // have left pending files, so just defensive: deleter.ClearPendingFiles(); System.String oldInfoFileName = segmentInfos.GetCurrentSegmentFileName(); System.String nextSegmentsFileName = segmentInfos.GetNextSegmentFileName(); // Checkpoint the state we are about to change, in // case we have to roll back: StartCommit(); bool success = false; try { DoCommit(); segmentInfos.Write(directory); success = true; } finally { if (!success) { // Rollback changes that were made to // SegmentInfos but failed to get [fully] // committed. This way this reader instance // remains consistent (matched to what's // actually in the index): RollbackCommit(); // Erase any pending files that we were going to delete: deleter.ClearPendingFiles(); // Remove possibly partially written next // segments file: deleter.DeleteFile(nextSegmentsFileName); // Recompute deletable files & remove them (so // partially written .del files, etc, are // removed): deleter.FindDeletableFiles(); deleter.DeleteFiles(); } } // Attempt to delete all files we just obsoleted: deleter.DeleteFile(oldInfoFileName); deleter.CommitPendingFiles(); if (writeLock != null) { writeLock.Release(); // release write lock writeLock = null; } } else DoCommit(); } hasChanges = false; } } protected internal virtual void SetDeleter(IndexFileDeleter deleter) { this.deleter = deleter; } protected internal virtual IndexFileDeleter GetDeleter() { return deleter; } /// Implements commit. protected internal abstract void DoCommit(); /// Closes files associated with this index. /// Also saves any new deletions to disk. /// No other methods should be called after this has been called. /// public void Close() { lock (this) { Commit(); DoClose(); if (closeDirectory) directory.Close(); } } /// Implements close. protected internal abstract void DoClose(); /// Release the write lock, if needed. ~IndexReader() { try { if (writeLock != null) { writeLock.Release(); // release write lock writeLock = null; } } finally { } } /// Get a list of unique field names that exist in this index and have the specified /// field option information. /// /// specifies which field option should be available for the returned fields /// /// Collection of Strings indicating the names of the fields. /// /// /// public abstract System.Collections.ICollection GetFieldNames(FieldOption fldOption); /// Returns true iff the index in the named directory is /// currently locked. /// /// the directory to check for a lock /// /// IOException if there is a problem with accessing the index public static bool IsLocked(Directory directory) { return directory.MakeLock(IndexWriter.WRITE_LOCK_NAME).IsLocked(); } /// Returns true iff the index in the named directory is /// currently locked. /// /// the directory to check for a lock /// /// IOException if there is a problem with accessing the index public static bool IsLocked(System.String directory) { Directory dir = FSDirectory.GetDirectory(directory); bool result = IsLocked(dir); dir.Close(); return result; } /// Forcibly unlocks the index in the named directory. ///

/// Caution: this should only be used by failure recovery code, /// when it is known that no other process nor thread is in fact /// currently accessing this index. ///

public static void Unlock(Directory directory) { directory.MakeLock(IndexWriter.WRITE_LOCK_NAME).Release(); } /// Prints the filename and size of each file within a given compound file. /// Add the -extract flag to extract files to the current working directory. /// In order to make the extracted version of the index work, you have to copy /// the segments file from the compound index into the directory where the extracted files are stored. /// /// Usage: Lucene.Net.Index.IndexReader [-extract] <cfsfile> /// [STAThread] public static void Main(System.String[] args) { System.String filename = null; bool extract = false; for (int i = 0; i < args.Length; ++i) { if (args[i].Equals("-extract")) { extract = true; } else if (filename == null) { filename = args[i]; } } if (filename == null) { System.Console.Out.WriteLine("Usage: Lucene.Net.Index.IndexReader [-extract] "); return ; } Directory dir = null; CompoundFileReader cfr = null; try { System.IO.FileInfo file = new System.IO.FileInfo(filename); System.String dirname = new System.IO.FileInfo(file.FullName).DirectoryName; filename = file.Name; dir = FSDirectory.GetDirectory(dirname); cfr = new CompoundFileReader(dir, filename); System.String[] files = cfr.List(); System.Array.Sort(files); // sort the array of filename so that the output is more readable for (int i = 0; i < files.Length; ++i) { long len = cfr.FileLength(files[i]); if (extract) { System.Console.Out.WriteLine("extract " + files[i] + " with " + len + " bytes to local directory..."); IndexInput ii = cfr.OpenInput(files[i]); System.IO.FileStream f = new System.IO.FileStream(files[i], System.IO.FileMode.Create); // read and write with a small buffer, which is more effectiv than reading byte by byte byte[] buffer = new byte[1024]; int chunk = buffer.Length; while (len > 0) { int bufLen = (int) System.Math.Min(chunk, len); ii.ReadBytes(buffer, 0, bufLen); byte[] byteArray = new byte[buffer.Length]; for (int index=0; index < buffer.Length; index++) byteArray[index] = (byte) buffer[index]; f.Write(byteArray, 0, bufLen); len -= bufLen; } f.Close(); ii.Close(); } else System.Console.Out.WriteLine(files[i] + ": " + len + " bytes"); } } catch (System.IO.IOException ioe) { System.Console.Error.WriteLine(ioe.StackTrace); } finally { try { if (dir != null) dir.Close(); if (cfr != null) cfr.Close(); } catch (System.IO.IOException ioe) { System.Console.Error.WriteLine(ioe.StackTrace); } } } } }