在 ftp 服务器中不递归列出文件、目录、子文件和子目录的想法
Ideas for listing files, directories, sub files & sub directories without recursion in ftp server
我正在尝试生成给定目录的文件列表及其在 ftp 服务器中的子目录。
服务器运行良好,我已经成功生成了当前目录的文件列表。当我尝试列出子目录和它们的文件时,事情变得复杂了。
我被要求不要使用递归算法,所以我自己做了一些研究。我曾尝试使用线程(对于找到的每个目录,启动一个新线程),但我无法保持连接稳定和打开。关于如何使用线程或其他替代方法正确执行此操作的任何想法?
编辑:下面是我的代码,当使用递归语句(最后一行代码)时,它有效
class TEST {
public static synchronized void main(String[] args) {
String server = args[0]; //server,path will be given as an arguments
String pass = "SOMEPASS";
String user = "SOMEUSER";
int port = 21;
FTPClient ftpClient = new FTPClient();
try {
ftpClient.connect(server, port);
showServerReply(ftpClient);
int replyCode = ftpClient.getReplyCode();
if (!FTPReply.isPositiveCompletion(replyCode)) {
System.out.println("Connect failed");
return;
}
boolean success = ftpClient.login(user, pass);
showServerReply(ftpClient);
if (!success) {
System.out.println("Could not login to the server");
return;
}
/*START THE FILE LISTING HERE*/
} catch (IOException ex) {
System.out.println("Oops! Something wrong happened");
ex.printStackTrace();
} finally {
// logs out and disconnects from server
try {
if (ftpClient.isConnected()) {
ftpClient.logout();
ftpClient.disconnect();
}
} catch (IOException ex) {
ex.printStackTrace();
}
}
}
private static void showServerReply(FTPClient ftpClient) {
String[] replies = ftpClient.getReplyStrings();
if (replies != null && replies.length > 0) {
for (String aReply : replies) {
System.out.println("SERVER: " + aReply);
}
}
}
private static void scanDir(FTPClient client, String path) throws IOException {
FTPFile[] files = client.listFiles(path); // Search all the files in the current directory
for (int j = 0; j < files.length; j++) {
System.out.println(files[j].getName()); // Print the name of each files
}
FTPFile[] directories = client.listDirectories(path); // Search all the directories in the current directory
for (int i = 0; i < directories.length; i++) {
String dirPath = directories[i].getName();
System.out.println(dirPath); // Print the path of a sub-directory
scanDir(client,dirPath); // Call recursively the method to display the files in the sub-directory DONT WANT TO DO THAT...
}
}
}
好的,这是一个示例,说明如何使用列表非递归地处理它。
请注意,此示例基于/访问本地文件系统,但可以很容易地 rewritten/extended 用于任何类型的 hierarchial/recursive 结构。
package Whosebug.nonrecursivefilesearch;
import java.io.File;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.ArrayList;
import java.util.stream.Stream;
public class NonRecursiveFileSearch {
public static void main(final String[] args) throws IOException {
final File searchDir = new File("D:\test\maven-test"); // set one
System.out.println("\nOld Java");
printDirs(listFiles_old(searchDir, true, true), "OLD: Depth first, include dirs");
printDirs(listFiles_old(searchDir, true, false), "OLD: Breadth first, include dirs");
printDirs(listFiles_old(searchDir, false, true), "OLD: Depth first, exclude dirs");
printDirs(listFiles_old(searchDir, false, false), "OLD: Breadth first, exclude dirs");
System.out.println("\nNew java.io with streams");
printDirs(listFiles_newIO(searchDir, true), "Java NIO, include dirs");
printDirs(listFiles_newIO(searchDir, false), "Java NIO, exclude dirs");
}
/**
* this is the way to 'manually' find files in hierarchial/recursive structures
*
* reminder: "Depth First" is not a real depth-first implementation
* real depth-first would iterate subdirs immediately.
* this implementation iterates breadth first, but descends into supdirs before it handles same-level directories
* advantage of this implementation is its speed, no need for additional lists etc.
*
* in case you want to exclude recursion traps made possible by symbolic or hard links, you could introduce a hashset/treeset with
* visited files (use filename strings retrieved with canonicalpath).
* in the loop, check if the current canonical filename string is contained in the hash/treeset
*/
static public ArrayList<File> listFiles_old(final File pDir, final boolean pIncludeDirectories, final boolean pDepthFirst) {
final ArrayList<File> found = new ArrayList<>();
final ArrayList<File> todo = new ArrayList<>();
todo.add(pDir);
while (todo.size() > 0) {
final int removeIndex = pDepthFirst ? todo.size() - 1 : 0;
final File currentDir = todo.remove(removeIndex);
if (currentDir == null || !currentDir.isDirectory()) continue;
final File[] files = currentDir.listFiles();
for (final File file : files) {
if (file.isDirectory()) {
if (pIncludeDirectories) found.add(file);
// additional directory filters go here
todo.add(file);
} else {
// additional file filters go here
found.add(file);
}
}
}
return found;
}
static private void printDirs(final ArrayList<File> pFiles, final String pTitle) {
System.out.println("====================== " + pTitle + " ======================");
for (int i = 0; i < pFiles.size(); i++) {
final File file = pFiles.get(i);
System.out.println(i + "\t" + file.getAbsolutePath());
}
System.out.println("============================================================");
}
/**
* this is the java.nio approach. this is NOT be a good solution for cases where you have to retrieve/handle files in your own code.
* this is only useful, if the any NIO class provides support. in this case, NIO class java.nio.file.Files helps handling local files.
* if NIO or your target system does not offer such helper methods, this way is harder to implement, as you have to set up the helper method yourself.
*/
static public Stream<Path> listFiles_newIO(final File pDir, final boolean pIncludeDirectories) throws IOException {
final Stream<Path> stream = Files.find(pDir.toPath(), 100,
(path, basicFileAttributes) -> {
final File file = path.toFile(); // conversion to File for easier access (f.e. isDirectory()), could also use NIO methods
return (pIncludeDirectories || !file.isDirectory() /* additional filters go here */ );
});
return stream;
}
static private void printDirs(final Stream<Path> pStream, final String pTitle) {
System.out.println("====================== " + pTitle + " ======================");
pStream.forEach(System.out::println);
System.out.println("============================================================");
}
}
AND,必须添加,java.nio.file.Files.find()
可以递归实现。但是因为它只是一个调用,所以这也可以算作 'non-recursive'。
此外,正如 OP 在评论中所述,可以使用 Stack 或其他 FIFO/LIFO 集合。混合深度优先的 LIFO,广度优先方法的 FIFO。
我正在尝试生成给定目录的文件列表及其在 ftp 服务器中的子目录。
服务器运行良好,我已经成功生成了当前目录的文件列表。当我尝试列出子目录和它们的文件时,事情变得复杂了。
我被要求不要使用递归算法,所以我自己做了一些研究。我曾尝试使用线程(对于找到的每个目录,启动一个新线程),但我无法保持连接稳定和打开。关于如何使用线程或其他替代方法正确执行此操作的任何想法?
编辑:下面是我的代码,当使用递归语句(最后一行代码)时,它有效
class TEST {
public static synchronized void main(String[] args) {
String server = args[0]; //server,path will be given as an arguments
String pass = "SOMEPASS";
String user = "SOMEUSER";
int port = 21;
FTPClient ftpClient = new FTPClient();
try {
ftpClient.connect(server, port);
showServerReply(ftpClient);
int replyCode = ftpClient.getReplyCode();
if (!FTPReply.isPositiveCompletion(replyCode)) {
System.out.println("Connect failed");
return;
}
boolean success = ftpClient.login(user, pass);
showServerReply(ftpClient);
if (!success) {
System.out.println("Could not login to the server");
return;
}
/*START THE FILE LISTING HERE*/
} catch (IOException ex) {
System.out.println("Oops! Something wrong happened");
ex.printStackTrace();
} finally {
// logs out and disconnects from server
try {
if (ftpClient.isConnected()) {
ftpClient.logout();
ftpClient.disconnect();
}
} catch (IOException ex) {
ex.printStackTrace();
}
}
}
private static void showServerReply(FTPClient ftpClient) {
String[] replies = ftpClient.getReplyStrings();
if (replies != null && replies.length > 0) {
for (String aReply : replies) {
System.out.println("SERVER: " + aReply);
}
}
}
private static void scanDir(FTPClient client, String path) throws IOException {
FTPFile[] files = client.listFiles(path); // Search all the files in the current directory
for (int j = 0; j < files.length; j++) {
System.out.println(files[j].getName()); // Print the name of each files
}
FTPFile[] directories = client.listDirectories(path); // Search all the directories in the current directory
for (int i = 0; i < directories.length; i++) {
String dirPath = directories[i].getName();
System.out.println(dirPath); // Print the path of a sub-directory
scanDir(client,dirPath); // Call recursively the method to display the files in the sub-directory DONT WANT TO DO THAT...
}
}
}
好的,这是一个示例,说明如何使用列表非递归地处理它。
请注意,此示例基于/访问本地文件系统,但可以很容易地 rewritten/extended 用于任何类型的 hierarchial/recursive 结构。
package Whosebug.nonrecursivefilesearch;
import java.io.File;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.ArrayList;
import java.util.stream.Stream;
public class NonRecursiveFileSearch {
public static void main(final String[] args) throws IOException {
final File searchDir = new File("D:\test\maven-test"); // set one
System.out.println("\nOld Java");
printDirs(listFiles_old(searchDir, true, true), "OLD: Depth first, include dirs");
printDirs(listFiles_old(searchDir, true, false), "OLD: Breadth first, include dirs");
printDirs(listFiles_old(searchDir, false, true), "OLD: Depth first, exclude dirs");
printDirs(listFiles_old(searchDir, false, false), "OLD: Breadth first, exclude dirs");
System.out.println("\nNew java.io with streams");
printDirs(listFiles_newIO(searchDir, true), "Java NIO, include dirs");
printDirs(listFiles_newIO(searchDir, false), "Java NIO, exclude dirs");
}
/**
* this is the way to 'manually' find files in hierarchial/recursive structures
*
* reminder: "Depth First" is not a real depth-first implementation
* real depth-first would iterate subdirs immediately.
* this implementation iterates breadth first, but descends into supdirs before it handles same-level directories
* advantage of this implementation is its speed, no need for additional lists etc.
*
* in case you want to exclude recursion traps made possible by symbolic or hard links, you could introduce a hashset/treeset with
* visited files (use filename strings retrieved with canonicalpath).
* in the loop, check if the current canonical filename string is contained in the hash/treeset
*/
static public ArrayList<File> listFiles_old(final File pDir, final boolean pIncludeDirectories, final boolean pDepthFirst) {
final ArrayList<File> found = new ArrayList<>();
final ArrayList<File> todo = new ArrayList<>();
todo.add(pDir);
while (todo.size() > 0) {
final int removeIndex = pDepthFirst ? todo.size() - 1 : 0;
final File currentDir = todo.remove(removeIndex);
if (currentDir == null || !currentDir.isDirectory()) continue;
final File[] files = currentDir.listFiles();
for (final File file : files) {
if (file.isDirectory()) {
if (pIncludeDirectories) found.add(file);
// additional directory filters go here
todo.add(file);
} else {
// additional file filters go here
found.add(file);
}
}
}
return found;
}
static private void printDirs(final ArrayList<File> pFiles, final String pTitle) {
System.out.println("====================== " + pTitle + " ======================");
for (int i = 0; i < pFiles.size(); i++) {
final File file = pFiles.get(i);
System.out.println(i + "\t" + file.getAbsolutePath());
}
System.out.println("============================================================");
}
/**
* this is the java.nio approach. this is NOT be a good solution for cases where you have to retrieve/handle files in your own code.
* this is only useful, if the any NIO class provides support. in this case, NIO class java.nio.file.Files helps handling local files.
* if NIO or your target system does not offer such helper methods, this way is harder to implement, as you have to set up the helper method yourself.
*/
static public Stream<Path> listFiles_newIO(final File pDir, final boolean pIncludeDirectories) throws IOException {
final Stream<Path> stream = Files.find(pDir.toPath(), 100,
(path, basicFileAttributes) -> {
final File file = path.toFile(); // conversion to File for easier access (f.e. isDirectory()), could also use NIO methods
return (pIncludeDirectories || !file.isDirectory() /* additional filters go here */ );
});
return stream;
}
static private void printDirs(final Stream<Path> pStream, final String pTitle) {
System.out.println("====================== " + pTitle + " ======================");
pStream.forEach(System.out::println);
System.out.println("============================================================");
}
}
AND,必须添加,java.nio.file.Files.find()
可以递归实现。但是因为它只是一个调用,所以这也可以算作 'non-recursive'。
此外,正如 OP 在评论中所述,可以使用 Stack 或其他 FIFO/LIFO 集合。混合深度优先的 LIFO,广度优先方法的 FIFO。