I have a text file, for example : file.txt, I want to read a line, for example line 7, have any way to read directly line 7 without reading other lines? I want to save memory from this work.
Because of the way JME been cut down you cannot do this. You would have to read whole file. Only other way, but may not be well suitable is to read the file , store it in RecordStore new entry per line, but is it really worth...
I think it is possible however you need to use a hashtable which might result to more heap usage.
Anyway, first, the contents of the text file should be stored in a char array. Then, second, the contents of the char array must be moved to the hashtable.
Each line in the text file is separated by a new line. In the char array, the new line (maybe) is translated to '\n'. Concatenate the characters in the array until the new line character is reached. The concatenated characters (minus '\n') will form the string in the first line. There should also be a counter here which should have been initialized to 0 (or 1, whatever you prefer). Save the text to the hashtable; The value will be the string that has been created and the key will be the counter. Increment the counter afterwards. Repeat this process for the remainder of the array until the end of file is reached.
With the hashtable, you can now get the string at line 7 without going through the other lines. Well, basically, each line has been read once. But, at least, you don't have to traverse each line once they have been stored in the hashtable.
Like what I have said earlier, doing this might increase heap usage especially if the text file is very large.
[And, by the way, sorry for the very late response. This is my first time here (I mean I just registered and answered this question) :D ]
Common Code
private String readLine(InputStream _inStream, int lineNum)
throws IOException {
if (null == _inStream) {
throw new IOException("Inputstream null.");
}
if (lineNum < 0) {
return ("Cannot read line a number " + lineNum);
}
final StringBuffer buf = new StringBuffer();
byte c;
int curLine = 1;
while (((c = (byte) _inStream.read()) != -1)) {
//System.out.println((char)c);
if (c == '\n') {
++curLine;
if (curLine > lineNum) {
break;
} else if (curLine < lineNum) {
continue;
}
} else if (curLine != lineNum) {
continue;
}
buf.append((char) c);
}
if (0 == buf.length()) {
return null;
} else {
return buf.toString().trim();
}
}
.
private String readLineWithSkip(InputStream _inStream, long skipCharacters)
throws IOException {
if (null == _inStream) {
throw new IOException("Inputstream null.");
}
if (skipCharacters < 1) {
return ("Cannot skip stream of " + skipCharacters + " characters");
}
final StringBuffer buf = new StringBuffer();
byte c;
_inStream.skip(skipCharacters);
while ((c = (byte) _inStream.read()) != '\n') {
//System.out.println((char)c);
buf.append((char) c);
}
if (0 == buf.length()) {
return null;
} else {
return buf.toString().trim();
}
}
.
InputStream inStream = null;
int fileLength = 39;
int skipCharacters = 10;
int lineNumber = 3;
String myLine = "No line read.";
try {
inStream = Class.class.getResourceAsStream("/test.txt");
if (null != inStream) {
inStream.mark(fileLength);
//For Approach II
myLine = readLine(inStream, lineNumber);
inStream.reset();
//For Approach I
myLine = readLineWithSkip(inStream, skipCharacters);
}
} catch (SecurityException se) {
se.printStackTrace();
} catch (IOException ioe) {
ioe.printStackTrace();
} finally {
try {
inStream.close();
} catch (IOException ex) {
ex.printStackTrace();
} catch (NullPointerException e) {
e.printStackTrace();
}
inStream = null;
System.out.println(myLine);
}
.
Approach I: Map the line number with no of cumulative characters
Run the file through a code that maps the line number with the last of charater in that line from the 0th position of the file (to be used as skip() value) all +2 ('\r\n\') for each line. You can store this Mapping table either at the start or at the end of the same file.
The run the above common code with method readLineWithSkip(inStream,
skipCharacters); ONLY and judiciously commenting other method calls.
Points to Consider:
Skips to the desired position in the inputstream
Has an overhead of parsing the file and storing the mapping table.
.
Approach II: Read each line till the Nth line is read
The run the above common code with method readLine(inStream,
lineNumber); ONLY and judiciously commenting other method calls.
Points to Consider:
Slow since it has to read each character till it reaches the desired line
No overhead of parsing the file and no storing of the mapping table.
I would like to further simplify the problem of reading the chars without requiring any mapping of characters with line numbers.
...
Form form=new Form("filename");
InputStream fin=fconn.openDataInputStream();
StringBuffer buf =new StringBuffer();
int c;
int counter=-1;
while((c=fin.read())!=-1)
{
counter++;
if(counter==23)
{
form.append(buf.toString());
buf=null;counter=0;
continue;
}
buf.append((char)c);
}
if(counter<23 && counter>=0) // write the skipped chars between the last number read and the end of file
{
form.append(buf.toString());
buf=null;counter=0;
}
fin.close();
...
Hope this would help others.
Related
I am trying to understand how this convertingStringToInt method works. I am reading a file, storing the values in an array and am to pass those values to the method to be converted. In the parameters of convertingStringToInt, I have (String number) I don't get where the String "number" is getting its values. So I am passing in a string called numbers, but how is that newly created String associated with any of the values in my file...?!?
I am trying to understand the cause all the return numbers are the error code -460 except the last digit in the file. So the String numbers is associated with the file somehow I just don't get how...
public static void read_file()
{
try {
File file = new File("randomNumbers.txt");
Scanner scan = new Scanner(file);
int amountOfNumbersInFile = convertingStringToInt(scan.nextLine()); // read the first line which is 100 to set array size
global_numbers = new int[amountOfNumbersInFile]; // set the array size equal to the first line read which is 100
for (int index = 0; index < amountOfNumbersInFile; index++)
{
String line = scan.nextLine();
global_numbers [index] = convertingStringToInt(line);
}
} catch (IOException ex) {
ex.printStackTrace();
}
}
public static int convertingStringToInt(String numbers) //what does string "number" equal? why/where is it declared?
{
String numbers = scan.nextInt();
try {
return Integer.parseInt(numbers);
} catch (NumberFormatException n) {
return -460;
}
}
I have global_numbers declared as a global variable.
so the first thing u need understand is what u have in your txt file
if in this file you have only number is ok use stringToInt
but if you have words this never work properly
In case of an error I want to generate in my compiler an error message like e.g. clang, which contains the complete offending source code line.
Example:
1.c:3:7: error: use of undeclared identifier 'x'
if ( x== y) {
^
I have extended the ANTLR BaseErrorListener, but I have no idea
how to get the offending source code line (besides of the line number) as a string.
That's a simple thing. You get line number and char position in your error info. Use that to locate the position in your input. Then scan back and forward for line breaks. The text between those linebreaks is your source code line.
I created 2 methods in my extension of BaseErrorListener. One to get the source code line and another to get the '^' sign:
private String getSourceLine(String src, int line) {
String currentLine = "";
try (Scanner fileScanner = new Scanner(new File(src))) {
int currentLineNumber = 1;
while (fileScanner.hasNextLine()) {
currentLine = fileScanner.nextLine();
if (currentLineNumber == line) {
return currentLine + "\n";
}
currentLineNumber++;
}
} catch (Exception e) {
currentLine = "\n";
}
return currentLine;
}
private String getPointer(int charPosition) {
return StringUtils.repeat(' ', charPosition) + '^';
}
Calling them with:
String source = recognizer.getInputStream().getSourceName();
errorMessage += this.getSourceLine(source, line);
errorMessage += this.getPointer(charPositionInLine);
I am having an issue with comparing strings in my program. I receive in serial data and save it to a string:
void serialEvent() {
if(!stringComplete){
while (Serial.available()) {
// get the new byte:
char inChar = (char)Serial.read();
// add it to the inputString:
inputString += inChar;
// if the incoming character is a newline, set a flag
// so the main loop can do something about it:
if (inChar == '\n') {
stringComplete = true;
Serial.println("COMPLETE");
}
I then do a compare on the string that was stored from the serialEvent function:
void setCMD(String a){
if(a == "01*00"){
busACTIVE=0;
// clear the string:
inputString = "";
stringComplete = false;
}
else if(a.equals("01*01")){
busACTIVE=1;
// clear the string:
inputString = "";
stringComplete = false;
}
I have several else if statements and then a else statement at the end:
else{
Serial.println("Command not Found");
Serial.println(a);
// clear the string:
inputString = "";
stringComplete = false;
}
I tried both == operator and equals() and neither will compare properly. below is a serial output:
Serial Output
As you can see one of my comparison statements looks for 01*01 and that is also what you see in the serial output window but the if statement did not equate to true. Can anyone help figure out why this is not working. Thanks
Try to edit this:
inputString += inChar;
// if the incoming character is a newline, set a flag
// so the main loop can do something about it:
if (inChar == '\n') {
stringComplete = true;
Serial.println("COMPLETE");
}
into this:
// if the incoming character is a newline, set a flag
// so the main loop can do something about it:
if (inChar == '\n') {
stringComplete = true;
Serial.println("COMPLETE");
}
else
inputString += inChar;
The reason is that if you are comparing "01*00" with "01*00\n" and, of course, the comparison fails.
Anyway I'd avoid using variable-sized buffers. I definitely prefer using fixed-size buffers for performance reasons. And also because microcontrollers are... micro! Don't waste their scarce resources on mallocs and frees...
I have textfiles, which have attributes saved in Strings. Those Strings have a pattern like this:
[attributeName]:[value]
I can't generalize the [value], because it could be of any primitive datatype.
Saving the effectively values is not my concern, because it's depending on the user which attribute has to be loaded. The same file won't be loaded very often.
Now I have 2 problems:
1) For some reason the program which creates those files sometimes adds spaces around the : at some attributes and [value] could also contain spaces, so I have to get rid of those
2) Making the reading of those attributes more performant:
I've come up with this method:
public String getAttribute(File file, String attribute)
{
try
{
BufferedReader reader = new BufferedReader(new FileReader(file), 1024);
String line;
Pattern p = Pattern.compile(Pattern.quote(attribute), Pattern.CASE_INSENSITIVE);
while ((line = reader.readLine()) != null)
{
int i = line.indexOf(":");
if(line.charAt(i-1) == ' ')
line = line.substring(0,i-2) + line.substring(i);
if(line.charAt(i+1) == ' ')
line = line.substring(0,i) + line.substring(i+2);
if (p.matcher(line).find())
{
return line.replace(attribute, "").trim();
}
}
} catch (IOException e)
{
e.printStackTrace();
}
return null;
}
However, this method will probably be one of the most called by my application, so I can't leave it so unperformant as it is right now,
Thanks for any help!
I modified code to find appropriate line. Check example code below.
If you have a lot of files and attributes in these files you could think about saving somewhere pair attribute=value in code. In example code I provided very primitive cache by using Table interface from guava library.
Example code:
# guava library
import com.google.common.collect.Table;
import com.google.common.collect.HashBasedTable;
# apache commons lang
import static org.apache.commons.lang.StringUtils.startsWithIgnoreCase;
# apache commons io
import static org.apache.commons.io.IOUtils.closeQuietly;
[...]
# very primitive cache implementation. To find a value in table you have to
# pass row and column keys. In this implementation row is equal to file
# absolute path (because you can have 2 files with the same name) and column
# is equal to attribute name.
# If you have a lot of files and attributes probably you will have to clear
# from time to time the cache otherwise you will get out of memory
private static final Table<String, String, String> CACHE = HashBasedTable.create();
[...]
public String getAttribute(File file, String attribute) {
# get value for the given attribute from the given file
String value = CACHE.get(file.getAbsolutePath(), attribute);
# if cache does not contain value, method will read value from file
if (null == value) {
BufferedReader reader = null;
String value = null;
try {
reader = new BufferedReader(new FileReader(file), 1024);
String line;
while ((line = reader.readLine()) != null) {
# From your description I understood that each line in file
# starts with attribute name
if (startsWithIgnoreCase(line, attribute) {
# if we found correct line we simple split it by ':'
String[] array = line.split(":");
# this is to ensure that line contains attribute name
# and value
if (array.length >= 2) {
# we found value for attribute and we remove spaces
value = array[1].trim();
# we put value to the cache to speed up finding
# value for the same attribute in the future
CACHE.put(file.getAbsolutePath(), attribute, value);
break;
}
}
}
} catch (IOException e) {
e.printStackTrace();
} finally {
# you should always close
closeQuietly(reader);
}
}
return value;
}
void BinaryTree::InitializeFromFile(string Filename){
ifstream inFile;
treenode* Freq[256];
inFile.open(Filename.c_str(), fstream::binary);
if(inFile.fail()){
cout<<"Error in opening file "<<Filename;
return;
}
for(int i=0;i<=255;i++){
Freq[i]->weight=0;
Freq[i]->data = '0'+i;
Freq[i]->LChild = NULL; Freq[i]->RChild=NULL; Freq[i]->Parent=NULL;
}
char c;
inFile.get(c);
while(!inFile.eof()){
Freq[c]->weight ++;
inFile.get(c);
}
}
I'm getting the Access Violation Exception in the for loop. Even when I comment out certain lines it'll give me an error on the next line in that loop.
Edit: Also is the line Freq[c]->weight ++; valid? Can I go to a specific part of the array based on the char value?
You seem to never initialize your Freq table. It contains random pointers. Dereferncing an uninitialized pointer leads to undefined behaviour.
You ought to add Freq[i] = new treenode before Freq[i]->weight=0;.