如果我有2个文件说ABCD.txt和DEF.txt。我需要检查DEF.txt中是否存在字符串“ABCD”以及ABCD.txt中是否存在字符串“DEF”,并将该组合写入文件。
我总共有大约15000个文件,每个文件包含近50 - 3000行必须被搜索。我写了一段代码,它的工作原理......但显示整个列表需要一个小时......
有更好的方法来执行此操作吗?请建议我。
public void findCyclicDependency(){
Hashtable<String, String> htFileNameList_1 = new Hashtable<String, String>();
Hashtable<String, String> htCyclicNameList = new Hashtable<String, String>();
FileWriter fwCyclicDepen = null;
PrintWriter outFile = null;
FileInputStream fstream = null;
FileInputStream fstream_1 = null;
DataInputStream in = null;
BufferedReader br = null;
DataInputStream in_1 = null;
BufferedReader br_1 = null;
String strSV_File_CK="";
boolean bFound = false;
File fileToSearch = null;
String strSVFileNameForComparison = "";
String strSVDependencyFileLine = "";
String strSVDFileLineExisting = "";
String strCyclicDependencyOut = "";
try {
File baseInputDirectory = new File(strInputPath);
List<File> baseInputDirListing = FileListing.getFileListing(baseInputDirectory);
// Printing out the filenames for the SodaSystem
for (File swPackage : baseInputDirListing)
{
if (swPackage.isDirectory() && swPackage.getName().endsWith("Plus")) {
List<File> currSwPackageFileListing = FileListing.getFileListing(swPackage);
System.out.println("\n swPackage File --> " + swPackage.getName() );
strCyclicDependencyOut = strOutputPath + "_"+ swPackage.getName() + "_CyclicDependency.xml";
System.out.println("\n strCyclicDependencyOut File --> " + strCyclicDependencyOut );
fwCyclicDepen = new FileWriter(strCyclicDependencyOut);
outFile = new PrintWriter(new BufferedWriter(fwCyclicDepen));
outFile.write("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
outFile.write("<CyclicDependencyFile>");
for (File DependentFile : currSwPackageFileListing) {
strSV_File_CK = DependentFile.getName().substring(0, (DependentFile.getName().length() - 4)).trim();
htFileNameList_1.put(strSV_File_CK.toUpperCase(),strSV_File_CK.toUpperCase());
}
for (File DependentFile : currSwPackageFileListing)
{
fstream = new FileInputStream(DependentFile);
// Get the object of DataInputStream
in = new DataInputStream(fstream);
br = new BufferedReader(new InputStreamReader(in));
strSVFileNameForComparison = DependentFile.getName().substring(0, (DependentFile.getName().length() - 4)).trim();
//Read File Line By Line
while ((strSVDependencyFileLine = br.readLine()) != null)
{
bFound = false;
if (strSVDependencyFileLine.toUpperCase().indexOf("INDICES") == -1)
{
//Check the current line matches any of the file name in software package folder
if (htFileNameList_1.contains(strSVDependencyFileLine.trim().toUpperCase())
&& strSVDependencyFileLine.compareTo(strSVFileNameForComparison) != 0)
{
bFound = true;
// Get the file to search
for (File searchFile : currSwPackageFileListing)
{
if((searchFile.getName().substring(0, (searchFile.getName().length() - 4)).trim()).equals(strSVDependencyFileLine))
{
fileToSearch = searchFile;
break;
}
}
// Read the file where the file name is found
fstream_1 = new FileInputStream(fileToSearch);
in_1 = new DataInputStream(fstream_1);
br_1 = new BufferedReader(new InputStreamReader(in_1));
while ((strSVDFileLineExisting = br_1.readLine()) != null)
{
if (strSVDFileLineExisting.toUpperCase().indexOf("EXTRA") == -1)
{
if (htFileNameList_1.contains(strSVDFileLineExisting.trim().toUpperCase()) && bFound
&& strSVDFileLineExisting.compareTo(strSVDependencyFileLine) != 0
&& strSVDFileLineExisting.compareTo(strSVFileNameForComparison) == 0 )
{
if(!htCyclicNameList.containsKey(strSVDependencyFileLine) &&
!htCyclicNameList.containsValue(strSVDFileLineExisting))
{
htCyclicNameList.put(strSVDFileLineExisting,strSVDependencyFileLine);
outFile.write("<CyclicDepedency FileName = \"" + strSVDFileLineExisting + "\""+ " CyclicFileName = \"" +
strSVDependencyFileLine + "\" />");
break;
}
}
}
}
}
}
else
{
bFound = false;
}
}//if current line <>
}// reach each line in the current file
outFile.write("</CyclicDependencyFile>");
}
outFile.flush();
outFile.close();
}
}
catch(Exception e){
e.printStackTrace();
}
}
由于 RAMM
答案 0 :(得分:1)
为索引所有文件,然后查询文件名并使用结果来检测循环依赖项,可能效率更高。
答案 1 :(得分:1)
您的设计存在一些问题。最重要的一个是重复扫描文件系统。请尝试下面的代码。
static public void findCyclicDependency2() {
PrintWriter outFile = null;
Map<String,File> fileNames = new HashMap<String,File>();
Map<String,Set<String>> fileBackward = new HashMap<String,Set<String>>();
Map<String,Set<String>> fileForward = new HashMap<String,Set<String>>();
try {
File baseInputDirectory = new File(strInputPath);
List<File> baseInputDirListing = getFileListing(baseInputDirectory);
// Printing out the filenames for the SodaSystem
for(File swPackage:baseInputDirListing) {
if (! (swPackage.isDirectory()
|| swPackage.getName().endsWith("Plus"))) continue;
System.out.println("Loading file names");
List<File> currSwPackageFileListing = getFileListing(swPackage);
for(File dependentFile:currSwPackageFileListing) {
String name = trimName(dependentFile);
fileNames.put(name,dependentFile);
BufferedReader br = new BufferedReader(new FileReader(dependentFile));
String line;
Set<String> contFor = new HashSet<String>();
Set<String> contBack = new HashSet<String>();
while( (line=br.readLine()) != null ) {
line = line.toUpperCase().trim();
if( line.equals("EXTRA") ) continue;
if( line.equals("INDICES") ) continue;
if( line.equals(name) ) continue;
if( line.compareTo(name) == 1 ) {
contFor.add(line);
} else {
contBack.add(line);
}
}
fileBackward.put(name,contBack);
fileForward.put(name,contFor);
}
String strCyclicDependencyOut = strOutputPath + "_"
+ swPackage.getName() + "_CyclicDependency.xml";
outFile = new PrintWriter(new BufferedWriter(new FileWriter(strCyclicDependencyOut)));
outFile.write("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
outFile.write("<CyclicDependencyFile>");
for(Entry<String,Set<String>> entry : fileForward.entrySet()) {
String curr = entry.getKey();
for(String other : entry.getValue()) {
Set<String> otherRefs = fileBackward.get(other);
if( otherRefs == null ) continue;
if( otherRefs.contains(curr) ) {
outFile.write("<CyclicDepedency FileName = \""
+ fileNames.get(curr).getPath()
+ "\""
+ " CyclicFileName = \""
+ fileNames.get(other).getPath()
+ "\" />");
}
}
}
outFile.write("</CyclicDependencyFile>");
outFile.flush();
outFile.close();
}
} catch (Exception e) {
e.printStackTrace();
}
}