Convert a given string into how Java would interpret it if used in code?

S

steve.chambers

Hi,

Is there any way to do this in Java? I'll try to explain what I mean a
bit better.

Given the following string in a text file:

This is a return:\r\nThis is a tab:\tand this is a backslash:\\

I want Java to interpret this as it would if this string were in code
by changing the escape characters into their single character
representations. I know I could do this by using a number of
String.replaceAll()'s but was wondering if there's a command to parse
the string & produce the result in a nicer way???

Thanks for any help with this...

Cheers,
Steve
 
S

steve.chambers

In fact I've just realised it's a bit more complicated than I thought
if using replaceAll(). If there was a double backslash followed by an
escape character e.g. \\t, which ever way round I do the replacing it's
going to run into problems. If I replace the \\ with \ first then \t
will still be replaced with the literal tab character afterwards. And
replacing the \t first wouldn't work either - what we want to end up
with is a literal backslash followed by a "t". I'm stuck! But clinging
onto the hope that there might be an API call somewhere that will take
the work away...
 
R

Robert Klemme

Please don't top post.

In fact I've just realised it's a bit more complicated than I thought
if using replaceAll(). If there was a double backslash followed by an
escape character e.g. \\t, which ever way round I do the replacing it's
going to run into problems. If I replace the \\ with \ first then \t
will still be replaced with the literal tab character afterwards. And
replacing the \t first wouldn't work either - what we want to end up
with is a literal backslash followed by a "t". I'm stuck! But clinging
onto the hope that there might be an API call somewhere that will take
the work away...
Click to expand...

I don't see the point. If you have the sequence "\\t" in your string
then what you want afterwards is a single literal backslash and a t.
Otherwise you would have to have either one or three backslashes.

The way I typically code this is a single loop that replaces the
sequence "backslash anything" with the appropriate value.

HTH

Regards

robert
 
S

Simon

Robert said:
Please don't top post.



I don't see the point. If you have the sequence "\\t" in your string
then what you want afterwards is a single literal backslash and a t.
Otherwise you would have to have either one or three backslashes.
Click to expand...

Take the string "\\t" and consider Steve's algorithm using replaceAll(). It can
be implemented in to ways:

Option 1:
1) First replace all "\\" by "\"
Result: "\t"
2) Then replace all "\t" by "[TAB]"
Result: "[TAB]"

Option 2:
1) First replace all "\t" by "[TAB]"
Result: "\[TAB]"
2) Then replace all "\t" by "[TAB]"
Result: "\[TAB]"

Both is not what you want ("\t"). It fails because a replacement character may
be replaced again. Iterating once over all characters of the original string and
appending the characters or their replacements to a new string should do the
job. This implementation should also be quick and easy to check for correctness.

Still, the question remains open whether this is implemented somewhere already.
Actually I think I did that a dozen times already...

Cheers,
Simon
 
S

steve.chambers

Take the string "\\t" and consider Steve's algorithm using replaceAll(). It can
be implemented in to ways:

Option 1:
1) First replace all "\\" by "\"
Result: "\t"
2) Then replace all "\t" by "[TAB]"
Result: "[TAB]"

Option 2:
1) First replace all "\t" by "[TAB]"
Result: "\[TAB]"
2) Then replace all "\t" by "[TAB]"
Result: "\[TAB]"

Both is not what you want ("\t"). It fails because a replacement character may
be replaced again. Iterating once over all characters of the original string and
appending the characters or their replacements to a new string should do the
job. This implementation should also be quick and easy to check for correctness.

Still, the question remains open whether this is implemented somewhere already.
Actually I think I did that a dozen times already...

Cheers,
Simon
Click to expand...

Well said simon, that was what I meant. What I could do to botch this
is a replaceAll on the double backslashes first but replace them with a
string that would never be expected and which doesn't include a
backslash character (e.g. "`¬backslash~@") and then replace all these
back to backslashes at the end, having replaced the other escape
characters.

However as this is a bit of a botch I think I'll go with the suggested
looping method instead. Here's my first attempt, which converts the
particular escape characters that I need to be able to use in my text
file but doesn't bother with backspaces, unicode literals etc:


/**
* Replaces the following escape characters in a string with their
literal
* equivalents:
* \\f -> \f (form feed)
* \\n -> \n (new line)
* \\r -> \r (carriage return)
* \\t -> \t (tab)
* \\' -> \' (single quote)
* \\" -> \" (double quote)
* \\\\ -> \\ (backslash)
*
* @param inputString The string in which the replacements will be
made
* @return The string with all escape characters replaced by their
* equivalent literals
*/
public static String replaceEscapesWithLiterals(String inputString)
{
String returnString = "";
int inputStringLength = inputString.length();

int charNum = 0;
while (charNum < inputStringLength) {
char currentChar = inputString.charAt(charNum);
char literal = '\0';
if ((currentChar == '\\') && (charNum + 1 <
inputStringLength)) {
char nextChar = inputString.charAt(charNum + 1);
switch (nextChar) {
case 'f':
literal = '\f';
break;
case 'n':
literal = '\n';
break;
case 'r':
literal = '\r';
break;
case 't':
literal = '\t';
break;
case '\'':
literal = '\'';
break;
case '\"':
literal = '\"';
break;
case '\\':
literal = '\\';
break;
}
}
if (literal == '\0') {
returnString += currentChar;
charNum++;
} else {
returnString += literal;
charNum += 2;
}
}

return returnString;
}
 
R

Roland de Ruiter

Take the string "\\t" and consider Steve's algorithm using replaceAll(). It can
be implemented in to ways:

Option 1:
1) First replace all "\\" by "\"
Result: "\t"
2) Then replace all "\t" by "[TAB]"
Result: "[TAB]"

Option 2:
1) First replace all "\t" by "[TAB]"
Result: "\[TAB]"
2) Then replace all "\t" by "[TAB]"
Result: "\[TAB]"

Both is not what you want ("\t"). It fails because a replacement character may
be replaced again. Iterating once over all characters of the original string and
appending the characters or their replacements to a new string should do the
job. This implementation should also be quick and easy to check for correctness.

Still, the question remains open whether this is implemented somewhere already.
Actually I think I did that a dozen times already...

Cheers,
Simon
Click to expand...

Well said simon, that was what I meant. What I could do to botch this
is a replaceAll on the double backslashes first but replace them with a
string that would never be expected and which doesn't include a
backslash character (e.g. "`¬backslash~@") and then replace all these
back to backslashes at the end, having replaced the other escape
characters.

However as this is a bit of a botch I think I'll go with the suggested
looping method instead. Here's my first attempt, which converts the
particular escape characters that I need to be able to use in my text
file but doesn't bother with backspaces, unicode literals etc:

[...]
Click to expand...

Here's one I did earlier ;-)

// begin of ConvertEscapedChars.java
public class ConvertEscapedChars {
public static void main(String[] args) {
final String orig = "This is a return:\\r\\nThis is "
+ "a tab:\\tand this is a backslash:\\\\.\r\n"
+ "Others are the backspace \\b\r\nthe formfeed \\f\r\n"
+ "single quote \\'\r\ndouble quote \\\"\r\n"
+ "one character octal \\7\r\ntwo character octal \\76\r\n"
+ "three character octal \\176\r\n"
+ "two character octal escape followed by octal digit \\767\r\n"
+ "\t(note that this is not a 3-char octal escape because "
+ "first digit is bigger than 3)\r\n"
+ "two character octal escape followed by non-octal digit "
+ "\\768\r\n"
+ "one character octal escape followed by non-octal digit "
+ "\\78\r\n"
+ "an invalid escape sequence \\w\r\n"
+ "an unterminated escape sequence \\";

System.out.println("------- Original -------");
System.out.println(orig);
System.out.println("------------------------");
System.out.println();

final String conv = convertEscapedChars(orig);
System.out.println("------- Converted ------");
System.out.println(conv);
System.out.println("------------------------");
}

/**
* Replaces escape sequences in the given String <code>s</code> by
* the actual characters they represent. The escape sequences that
* are recognized are those that are defined in section 3.10.6 of
* the JLS, 3rd ed. Invalid escape sequences are left untouched.
*
* @param s
* the String to convert
* @return a String where escape sequences have been replaced by
* their actual character
* @see
http://java.sun.com/docs/books/jls/third_edition/html/lexical.html#101089
*/
public static String convertEscapedChars(String s) {
int n = s == null ? 0 : s.length();
if (n == 0) {
return s; // null or empty string
}
StringBuffer result = new StringBuffer(n);
for (int i = 0; i < n; i++) {
char c = s.charAt(i);
if (c != '\\') {
result.append(c);
} else {
if (++i < n) {
c = s.charAt(i);
switch (c) {
case 'b':
result.append('\b');
break;
case 't':
result.append('\t');
break;
case 'n':
result.append('\n');
break;
case 'f':
result.append('\f');
break;
case 'r':
result.append('\r');
break;
case '\"':
result.append('\"');
break;
case '\'':
result.append('\'');
break;
case '\\':
result.append('\\');
break;
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
// OctalEscape
StringBuffer octal =
new StringBuffer(3).append(c);
if (i + 1 < n && (c = s.charAt(i + 1)) >= '0'
&& c <= '7') {
i++;
octal.append(c);
if (i + 1 < n && (c = s.charAt(i + 1))>= '0'
&& c <= '7') {
i++;
octal.append(c);
}
}
if (octal.length()==3 && octal.charAt(0)>'3') {
i--;
octal.setLength(2);
}
result.append(
(char) Integer.parseInt(octal.toString(),
8));
break;
default:
System.err.println(
"Invalid escape sequence: \\" + c);
result.append('\\').append(c);
break;
}
} else {
System.err.println("Unterminated escape sequence");
result.append('\\');
}
}
}
return result.toString();
}
}
// end of ConvertEscapedChars.java
 
S

steve.chambers

Here's one I did earlier ;-)
// begin of ConvertEscapedChars.java
public class ConvertEscapedChars {
public static void main(String[] args) {
final String orig = "This is a return:\\r\\nThis is "
+ "a tab:\\tand this is a backslash:\\\\.\r\n"
+ "Others are the backspace \\b\r\nthe formfeed \\f\r\n"
+ "single quote \\'\r\ndouble quote \\\"\r\n"
+ "one character octal \\7\r\ntwo character octal \\76\r\n"
+ "three character octal \\176\r\n"
+ "two character octal escape followed by octal digit \\767\r\n"
+ "\t(note that this is not a 3-char octal escape because "
+ "first digit is bigger than 3)\r\n"
+ "two character octal escape followed by non-octal digit "
+ "\\768\r\n"
+ "one character octal escape followed by non-octal digit "
+ "\\78\r\n"
+ "an invalid escape sequence \\w\r\n"
+ "an unterminated escape sequence \\";

System.out.println("------- Original -------");
System.out.println(orig);
System.out.println("------------------------");
System.out.println();

final String conv = convertEscapedChars(orig);
System.out.println("------- Converted ------");
System.out.println(conv);
System.out.println("------------------------");
}

/**
* Replaces escape sequences in the given String <code>s</code> by
* the actual characters they represent. The escape sequences that
* are recognized are those that are defined in section 3.10.6 of
* the JLS, 3rd ed. Invalid escape sequences are left untouched.
*
* @param s
* the String to convert
* @return a String where escape sequences have been replaced by
* their actual character
* @see
http://java.sun.com/docs/books/jls/third_edition/html/lexical.html#101089
*/
public static String convertEscapedChars(String s) {
int n = s == null ? 0 : s.length();
if (n == 0) {
return s; // null or empty string
}
StringBuffer result = new StringBuffer(n);
for (int i = 0; i < n; i++) {
char c = s.charAt(i);
if (c != '\\') {
result.append(c);
} else {
if (++i < n) {
c = s.charAt(i);
switch (c) {
case 'b':
result.append('\b');
break;
case 't':
result.append('\t');
break;
case 'n':
result.append('\n');
break;
case 'f':
result.append('\f');
break;
case 'r':
result.append('\r');
break;
case '\"':
result.append('\"');
break;
case '\'':
result.append('\'');
break;
case '\\':
result.append('\\');
break;
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
// OctalEscape
StringBuffer octal =
new StringBuffer(3).append(c);
if (i + 1 < n && (c = s.charAt(i + 1)) >= '0'
&& c <= '7') {
i++;
octal.append(c);
if (i + 1 < n && (c = s.charAt(i + 1))>= '0'
&& c <= '7') {
i++;
octal.append(c);
}
}
if (octal.length()==3 && octal.charAt(0)>'3') {
i--;
octal.setLength(2);
}
result.append(
(char) Integer.parseInt(octal.toString(),
8));
break;
default:
System.err.println(
"Invalid escape sequence: \\" + c);
result.append('\\').append(c);
break;
}
} else {
System.err.println("Unterminated escape sequence");
result.append('\\');
}
}
}
return result.toString();
}
}
// end of ConvertEscapedChars.java
Click to expand...

Thanks Roland, looks like I've reinvented the wheel with a slightly
inferior version!
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Similar Threads

Hello guys ! How do I convert a string from an array into numbers ? Javascript 3
Trouble with prediction code, for the life of me I can't figure out why it isnt running properly. Help would be appreciated. 0
Converting an Array to a String in JavaScript 7
Teaching myself how to code using java but stuck and need help 0
How to sort a CSV file with merge sort JAVA 7
How to convert utf-8 bytes into a java string? 5
insert html into ElementTree without parsing it 1
JSF feeds a Set<String> into a Set<Foo> 2

Members online

No members online now.
Total: 33 (members: 0, guests: 33)
Robots: 462

Forum statistics

Threads
473,768
Messages
2,569,574
Members
45,048
Latest member
verona

Latest Threads

Top