Unexplainable run-time segfault error

This is really bizarre... :wtf:
Ok, so I have this function called "strIsFollowedByASpace()" which returns true if a given search string is followed by a space in a source string, else it returns false.
At runtime in my program (since I managed to compile it fine), there was this segfault that kept cropping up, so I traced it to its source (by adding cout lines until I managed to ouput one of them), and the source was inside the strIsFollowedByASpace(). Note that this function had already been previously tested, and worked before. So I wasn't expecting for there to be an error in it, but I examined it more closely anyway once I discovered where the segfault was generating from. I had (since it was working) briefly modified it, but not anything that I thought would generate such an error. All I did was I changed some ints to unsigned ints. But anyway, I examined it anyway, and discovered where in the function the segfault was coming from, and tried to figure out a way to fix it, but could not. And you'll see why in a moment.

I added some cout lines to the function to see if it entered an if statement that looks like this:

Code:

---

if (strOfSource.at(endIndex + 1) != ' ') { //the character after the last character of searchString in sourceString is not a space
    cout<< "sourceString[" << endIndex << " + 1] != ' '." <<endl;
    return false;
}

---

And it did not output that statement before seg-faulting, which gives evidence that it didn't enter it for some reason. Now here's why this is so weird:

The same exact function/usecase works with no seg-fault when I tested it at codepad.org:

C++ code - 56 lines - codepad

I pass the same exact same c-string to the function as that passed in my real code, to try to replicate the real circumstances as much as possible, and there the function works as expected (no seg-fault)!
In the real code, I pass to the function the buffer of a line read from file which contains the string "enum DAY {" (no quotes), which has 10 characters. It outputs the size of the sourceString (in my real code, as well as at the link I posted) as 10, which is correct, and it outputs the index of the last character of the search string (which is "enum") as 3. Hence, when I say access sourceString[endIndex + 1] that means in this case, it is sourceString[3 + 1] which is element 4. Since the array is 0-based, that is really character 5 (e = 0, n = 1, u = 2, m = 3, space-character = 4), which is indeed a space, and the codepad.org function proves this by doing the correct thing. That is why this is so bizarre in my function. And when I change the sourceString[endIndex + 1] line in my real code to sourceString[endIndex] instead, it no longer seg-faults, but of course it enters the wrong statement (i.e. if the "enum" string is not followed by a space) hence wrecking havoc on the result of my code.

So I guess my question is: Why the heck is this function seg-faulting when there is clearly no problem with the code? :D

Hmm...nevermind. The function is working ok in the real code. It enters the else statement that indicates that the "enum" string IS followed by a space, thus returning true. The seg-fault appears to be somewhere else. Ignore my ramblings...

What the hell! Ok, now we got something else unexplainable (for real, this time). :) This statement should NOT be entered on the first line read from file which contains the string "enum DAY {", but its entered anyway as evidenced by a debug cout statement I put inside it:

Code:

---

if (containsString(filestreamInBuffer, ";")) { //line which contains 'enum' contains the terminating semicolon also
                    cout<< "This statement has been entered anyway. WTF???" <<endl;
                    //do stuff...
}

---

Now my program outputs:

Quote:

---

This statement has been entered anyway. WTF???

---

right before the seg-fault. Note that I just copied the containsString() function to codepad.org to double-check if its still working, and it is. When I pass to it the same exact string, the function returns false as it should:

C++ code - 224 lines - codepad

So now I have an if (false) statement that's being bizarrely entered despite being false. Can any of you geniuses explain this one to me?

Ok, I added another cout statement inside the if statement, which outputs the content of the "filestreamInBuffer" c-string, and so it now outputs:

Quote:

---

This statement has been entered anyway. WTF???
filestreamInBuffer contains: enum DAY {

---

Did I break my compiler or what...? :D

When I compiled your code, it seems to run ok. no errors and does not enter the if() in main.
What compiler and/or IDE are you using?

Jim.

Quote:

---

Originally Posted by jimblumberg

When I compiled your code, it seems to run ok. no errors and does not enter the if() in main.
What compiler and/or IDE are you using?

Jim.

---

GCC / Code::Blocks

Note that my whole program compiles fine. Its a run-time error/seg-fault we're talking about here.
And its becoming a real pain in the ass... :(

Have you tried to DEBUG the program using the debugger and stepping through your code?

Jim

Is the code being run within the
if (searchStringIsSpecialString == 0)
conditional the same in all 3 cases?

If it is, then you really need to stop copy/pasting code and create a few more functions along the way.

200+ lines and 10+ levels of indentation - no wonder you're confused.

Quote:

---

Originally Posted by Salem

Is the code being run within the
if (searchStringIsSpecialString == 0)
conditional the same in all 3 cases?

---

Yes.

Quote:

---

If it is, then you really need to stop copy/pasting code and create a few more functions along the way.

---

Lol, I've considered it.

Quote:

---

200+ lines and 10+ levels of indentation - no wonder you're confused.

---

Actually, that's not the confusing part. In fact, none of my code is confusing (to me) except for the unexplainable part where an if(false) statement is entered.

Quote:

---

Originally Posted by jimblumberg

Have you tried to DEBUG the program using the debugger and stepping through your code?

Jim

---

That's no use because code execution doesn't even reach the functions we're talking about when the program is run with no arguments passed in.

If it's command line arguments you are missing the add them to the project (under Project - Set programs' arguments.

Jim

Quote:

---

Originally Posted by jimblumberg

If it's command line arguments you are missing the add them to the project (under Project - Set programs' arguments.

Jim

---

Thanks. I missed that feature.
Yeah, that works.
But its not like I don't already know where the problem begins. It begins at the if statement I already mentioned which is being entered even though the return value of the containsString() function is false.

EDIT: Well, it did help me find the exact line that the seg-fault is on (on the code line right after the output lines I mentioned above).

> In fact, none of my code is confusing (to me) except for the unexplainable part where an if(false) statement is entered.
Yeah - now try leaving it alone for a week.
When you come back to it, you'll have an instant WTF moment.

I might explain why, if I thought it would do any good....

So is this a problem of entering an if statement when it's not supposed to, or a segfault problem?
For either problem please post the smallest piece of code that demonstrates the problem. The code you have posted does not seem to have any problems when I compile and run the code snippet. Also post a sample of any any files the program is reading and or writing and any command line arguments passed to the program.

Jim

Quote:

---

Originally Posted by Salem

> In fact, none of my code is confusing (to me) except for the unexplainable part where an if(false) statement is entered.
Yeah - now try leaving it alone for a week.
When you come back to it, you'll have an instant WTF moment.

---

I doubt it. And I'm still not confused... :p
By the sound of it though, you are confused by my code.

Quote:

---

I might explain why, if I thought it would do any good....

---

Well, I might think it would do some good, if it weren't for the fact that I already know the answer to the why question.
Oh, and for the record, I went ahead and followed your advice and put the "if (searchStringIsSpecialString" stuff in two other functions, which I then call from inside containsString().

Quote:

---

Originally Posted by jimblumberg

So is this a problem of entering an if statement when it's not supposed to, or a segfault problem?

---

Both. The segfault problem seems to be buried in my strippedString() function, which I'm debugging right now.

Quote:

---

For either problem please post the smallest piece of code that demonstrates the problem. The code you have posted does not seem to have any problems when I compile and run the code snippet. Also post a sample of any any files the program is reading and or writing and any command line arguments passed to the program.

Jim

---

I will. Just a sec...

Ok, here is the problem code:

Code:

---

if (containsString(filestreamInBuffer, ";")) { //line which contains 'enum' contains the terminating semicolon also
    cout<< "This statement has been entered anyway. WTF???" <<endl;
    cout<< "filestreamInBuffer contains: " << filestreamInBuffer <<endl;
    currentEnumName = stripString(filestreamInBuffer, 'm', '{'); //we now have the name of the current enum
    //continue doing stuff inside this if statement...
}

---

Ok, now my output statements demonstrate that that if statement is being entered (even though it is clear that the return value of the containsString() function here is false, but the if statement is still entered).
My debugger shows that the line which produces the segfault is the bolded line of code, so the problem appears to be inside stripString().
But the output shows this:

Quote:

---

This statement has been entered anyway. WTF???
filestreamInBuffer contains: enum DAY {
stripString() has been entered.
sizeOfStr is: 10
str[3] == start
str[4] != end
str[5] != end
str[6] != end
str[7] != end
str[8] != end
str[9] == end
strippedString is: DAY
Segmentation fault

---

And here is the definition of the stripString() function (note that I narrowed the last output line before the segfault I posted to the else statement at the end of the function):

Code:

---

const string& CenumOperations::stripString(const char str[], const char start, const char end, ET_strippingMode strippingMode) {

    cout<< "stripString() has been entered." <<endl;
    unsigned int sizeOfStr = 0;
    while (str[sizeOfStr] != '\0') {
        sizeOfStr++; //increment sizeOfStr
    }
    cout<< "sizeOfStr is: " << sizeOfStr <<endl;
    bool startHasNotBeenReachedYet = true;
    bool endHasNotBeenReachedYet = true;
    unsigned int i2 = 0;
    string strippedString = ""; //create an empty string
    for (unsigned int i = 0; i < sizeOfStr; i++) {
        //loop without doing anything until 'start' character is found or until the end of the string has been reached...
        if (str[i] == start) {
            cout<< "str[" << i << "] == start" <<endl;
            startHasNotBeenReachedYet = false; //since 'start' has now been found
            if (strippingMode == exclusive) {
                i2 = i+1; //begin the inner do-while loop counter at the index of the next character after the 'begin' character
            }

            else if (strippingMode == inclusive) {
                i2 = i; //begin the inner do-while loop counter at the index of the 'begin' character
            }

            do {
                if (i2 < sizeOfStr - 1) { //we're not at the last character of the string yet
                    if (strippingMode == exclusive) { //meaning we want to exclude the 'begin' and 'end' characters from stripped string
                        if (str[i2] == end) { //the current character matches the 'end' character
                            cout<< "str[" << i2 << "] == end" <<endl;
                            endHasNotBeenReachedYet = false; //since we do not want to add the 'end' character to the stripped string
                            cout<< "strippedString is: " << strippedString <<endl;
                            return strippedString;
                        }

                        else if (str[i2] != end) { //the current character does not match the 'end' character
                            cout<< "str[" << i2 << "] != end" <<endl;
                            if (str[i2] != ' ') { //since we don't want spaces in the stripped string...
                                strippedString += str[i2]; //add the current character of the string to the stripped string
                            }
                        }
                    }

                    else if (strippingMode == inclusive) { //meaning we want to include the 'begin' and 'end' characters from stripped string
                        if (str[i2] == end) { //the current character matches the 'end' character
                            cout<< "str[" << i2 << "] == end" <<endl;
                            if (str[i2] != ' ') { //since we don't want spaces in the stripped string...
                                strippedString += str[i2]; //add the 'end' character to the stripped string
                            }
                            endHasNotBeenReachedYet = false; //since we added the 'end' character and we're done
                        }

                        else if (str[i2] != end) { //the current character does NOT match the 'end' character
                            cout<< "str[" << i2 << "] != end" <<endl;
                            if (str[i2] != ' ') { //since we don't want spaces in the stripped string...
                                strippedString += str[i2]; //add the remaining character we want to add
                            }
                        }
                    }
                }

                else if (i2 == sizeOfStr - 1) { //we're at the last character of 'str'...
                    if (strippingMode == exclusive) { //meaning we want to exclude the 'begin' and 'end' characters from stripped string
                        if (str[i2] == end) { //the last character of the string matches the 'end' character
                            cout<< "str[" << i2 << "] == end" <<endl;
                            endHasNotBeenReachedYet = false; //since we do not want to add the 'end' character to the stripped string
                            cout<< "strippedString is: " << strippedString <<endl;
                            return strippedString;
                        }

                        else if (str[i2] != end) { //the current character does not match the 'end' character
                            cout<< "str[" << i2 << "] != end" <<endl;
                            if (str[i2] != ' ') { //since we don't want spaces in the stripped string...
                                strippedString += str[i2]; //add the last character of the string to the stripped string
                            }

                            break; //out of the loop, since we just added the last character (if it wasn't a space)
                        }
                    }

                    else if (strippingMode == inclusive) { //meaning we want to include the 'begin' and 'end' characters from stripped string
                        if (str[i2] == end) { //the last character of the string matches the 'end' character
                            cout<< "str[" << i2 << "] == end" <<endl;
                            strippedString += str[i2]; //add the 'end' character to the stripped string
                            endHasNotBeenReachedYet = false; //since we added the 'end' character and we're done
                        }

                        else if (str[i2] != end) { //the last character of the string does NOT match the 'end' character
                            cout<< "str[" << i2 << "] != end" <<endl;
                            if (str[i2] != ' ') { //since we don't want spaces in the stripped string...
                                strippedString += str[i2]; //add the last character of the string to the stripped string
                            }

                            break; //out of the loop, since we just added the last character
                        }
                    }
                }

                i2++; //increment the inner-do while loop iterator
            } while(endHasNotBeenReachedYet);

            break; //out of the for loop since we already scanned through the rest of the string after 'start'
        }

    }

    if (startHasNotBeenReachedYet) { //meaning the 'start' character was not found after iterating through the whole string
        return "";
    }

    else {
        cout<< "strippedString is: " << strippedString <<endl;
        return strippedString;
    }

}

---

And here is the content of the file I'm passing to the program:

Quote:

---

enum DAY {
saturday,
sunday = 0,
monday,
tuesday,
wednesday,
thursday,
friday
} workday;

enum Work {
hard = 0,
lazy = -1
};

enum weirdenum //this is a weird enum... :)
{

weirdenum1,
weirdenum2 = 2,
weirdenum3 = 0

};

---

EDIT: Oh, and my program has not got as far as outputting to file yet (even though it created the output file, like intended). But here is the command I used to run ConvertEnumToStrings:

Code:

---

./ConvertEnumToStrings "/home/gorilla/Documents/test.cpp" "results"

---

The result file was "results.h" (created in the same directory as the inputted file), but as already mentioned, it has not got as far as outputting anything to it since the segfault stops the program before it gets a chance to.

Quote:

---

Ok, now my output statements demonstrate that that if statement is being entered (even though it is clear that the return value of the containsString() function here is false, and so the if statement is not entered.

---

O_o

Quote:

---

that [the] if statement is being entered

---

Quote:

---

even though

---

Quote:

---

the if statement is not entered

---

o_O

Would you like to revise the following statements?

Quote:

---

I'm still not confused...

---

Quote:

---

I am both a musician and a programmer.

---

I'm curious. Do you think that the compiler is just being unkind?

Soma

Quote:

---

Originally Posted by phantomotap

O_o





o_O

Would you like to revise the following statements?

---

Damnit! :) I knew someone was going to reply to that. I already edited and fixed that remark to what I really intended before you posted.

Quote:

---

I'm curious. Do you think that the compiler is just being unkind?

Soma

---

Apparently... :p
Because the computer IS entering a if(false) statement that it shouldn't be.
I don't know...I seem to have the most unique problems around here.

O_o

Arrogant, inexperienced, and confused?

Are you the unholy newbie of legend?

Soma

Quote:

---

Originally Posted by phantomotap

O_o

Arrogant, inexperienced, and confused?

Are you the unholy newbie of legend?

Soma

---

hahaha! That's a good one.
For the record though, I'm not arrogant or confused. Inexperienced, yes, but not arrogant or confused (at least not about the topic at hand).
And you annoying little prick...my whole life I've had a very low self-esteem. Perhaps you should look in the mirror for a glance at those same attributes you mentioned? :)

Your main difficulty with finding your bug is that you're quite literally writing more than ten times more code than you should be!
The following replaces 48 lines of your code (Lines 91 to 137):

Code:

---

char c = sourceString[i];
if ('A' <= ch && ch <= 'Z' || 'a' <= ch && ch <= 'z')
    return true;
if (i == sourceStringSize - 1)
    return false;

---

Now consider that those same lines of code appear three times and that the five lines I just showed should appear just once, in a function of its own, and there's a 25 - 30x reduction in number of lines of code.

Lines 9 to 17 could be replaced by two strlen's.

Lines 74 to 83 could be replaced with the one-liner:

Code:

---

return string(sourceString).find(searchString) != string::npos;

---

The whole thing could probably be about 30 lines long rather than 224 lines long.

Stuff like this is precisely why you are having problems.

Quote:

---

That's a good one.

---

Thank you; it is very kind of you to notice.

Quote:

---

annoying little prick

---

That's "annoying medium prick" to you! ^_^v

Quote:

---

my whole life I've had a very low self-esteem

---

Right...

Quote:

---

Perhaps you should look in the mirror for a glance at those same attributes you mentioned?

---

I'd never forget my numerous faults. I value them in a way. Just in case I'm tempted to forget, I keep a framed list above my desk.

Soma

Quote:

---

Originally Posted by phantomotap

That's "annoying medium prick" to you! ^_^v

---

Whatever works for you.... :)

Quote:

---

Right...

---

Hey, its true! When I was growing up, I was very shy and didn't talk much.
I didn't think that much of myself then (and still don't). And it always ........es me off every time I screw up too (which is often). I have never had that high and lofty thoughts about myself.

Quote:

---

I'd never forget my numerous faults. I value them in a way. Just in case I'm tempted to forget, I keep a framed list above my desk.

Soma

---

Lol...
Good idea. Maybe I'll do that too.

@iMalc: Thanks. I incorporated a lot of your advice in my new code. But it still seg-faults...

What is your current code? Instead of using debugging output statements, use a debugger.

Quote:

---

Originally Posted by laserlight

What is your current code? Instead of using debugging output statements, use a debugger.

---

C++ code - 279 lines - codepad

And I fixed the seg-fault. It turned out to be stripString() indeed. Apparently, it was seg-faulting because it was returning a reference to a local variable. Once I changed the function to return strippedString by value instead, the seg-fault went away. But still, the if statement is being entered when the return value of containsString is returning false, so its messing up the entire results which depend on that.

Eh, do you know that you are unnecessarily verbose in several places? For example:

Code:

---

if (searchStringIsSpecialString(searchString)) {
    //search for a character in sourceString in the range of A-Z, or a-z:
    bool trueOrFalse = sourceStringContainsAlphabetChar(sourceString);
    if (trueOrFalse == true) {
        return true;
    }

    else if (trueOrFalse == false) {
        return false;
    }
}

---

The above code snippet can be trivially simplified to:

Code:

---

if (searchStringIsSpecialString(searchString)) {
    //search for a character in sourceString in the range of A-Z, or a-z:
    return sourceStringContainsAlphabetChar(sourceString);
}

---

Quote:

---

Originally Posted by laserlight

Eh, do you know that you are unnecessarily verbose in several places? For example:

Code:

---

if (searchStringIsSpecialString(searchString)) {
    //search for a character in sourceString in the range of A-Z, or a-z:
    bool trueOrFalse = sourceStringContainsAlphabetChar(sourceString);
    if (trueOrFalse == true) {
        return true;
    }

    else if (trueOrFalse == false) {
        return false;
    }
}

---

The above code snippet can be trivially simplified to:

Code:

---

if (searchStringIsSpecialString(searchString)) {
    //search for a character in sourceString in the range of A-Z, or a-z:
    return sourceStringContainsAlphabetChar(sourceString);
}

---

---

Yes, I know, but I figure my method is more clearer when you're looking at the code.
Anyway, none of this has anything to do with a if(false) statement being entered. So far, no one has commented on this...

Quote:

---

Originally Posted by Programmer_P

Yes, I know, but I figure my method is more clearer when you're looking at the code.

---

No, it isn't. It makes one think what your code is actually trying to do, whereas an immediate return is straightforward.

Quote:

---

Originally Posted by Programmer_P

Anyway, none of this has anything to do with a if(false) statement being entered. So far, no one has commented on this...

---

Simplify your code first. I do not have a compiler with me anyway, so I am literally eyeballing your code only, so please be kind and make it easier to follow.

Quote:

---

Originally Posted by laserlight

No, it isn't. It makes one think what your code is actually trying to do, whereas an immediate return is straightforward.

---

That is a matter of opinion, but I changed the code:
http://codepad.org/VoH75k1Z

And before you mention it, the "const" keyword was removed from the return value of stripString, the post was edited, and the new link given. :)
I had it like that because it was originally returning a reference, then when I changed it back to return just a string, I forgot to remove the "const" keyword. But that's fixed now.

Here is what I might simplify the first part of your code to:

Code:

---

#include <string>
#include <cstring>
#include <cctype>
#include <iostream>
#include <cassert>

using namespace std;

enum ET_strippingMode {
    exclusive = 0,
    inclusive = 1
};

bool searchStringIsSpecialString(const char searchString[]) {
    return strcmp(searchString, "A-Za-z") == 0;
}

bool sourceStringContainsAlphabetChar(const char sourceString[]) {
    //we are assuming 'sourceString' is pointed at a null-terminated char array...
    unsigned int sourceStringSize = strlen(sourceString);

    for (unsigned int i = 0; i < sourceStringSize; i++) {
        if (isalpha(sourceString[i])) {
            return true; //at the first char in the range of A-Z or a-z
        }
    }

    return false;
}

bool containsString(const char sourceString[], const char searchString[]) {
    //assume that 'sourceString' and 'searchString' are pointing at two null-terminated char arrays...
    unsigned int sourceStringSize = strlen(sourceString);
    unsigned int searchStringSize = strlen(searchString);

    if (searchStringSize < sourceStringSize) {
        //if the special string "A-Za-z" was passed into the 'searchString' parameter
        //then search for a character in sourceString in the range of A-Z, or a-z,
        //otherwise check to see if the searchString is contained in the sourceString.
        return searchStringIsSpecialString(searchString) && sourceStringContainsAlphabetChar(sourceString)
            || sourceString.find(searchString) != string::npos;
    }
    else if (searchStringSize == sourceStringSize) { //meaning the two strings have the same length
        //if the special string "A-Za-z" was passed into the 'searchString' parameter
        //then search for a character in sourceString in the range of A-Z, or a-z,
        //otherwise check to see if the source string and the search string are the same.
        return searchStringIsSpecialString(searchString) && sourceStringContainsAlphabetChar(sourceString)
            || strcmp(sourceString, searchString) == 0;
    }

    //since the string searched for is longer string than the one being checked,
    //there is no possibility of the shorter string fully containing the longer string.
    return false;
}

---

Now, for the stripString function: I suggest that you break it up into two helper functions: one to handle exclusive stripping mode, the other to handle inclusive stripping mode.

EDIT:
Oh, and another thing for stripString: instead of nesting a do while loop in a for loop, first use one loop to find the start character, and then use a separate loop for the rest of the parsing.

Quote:

---

Originally Posted by laserlight

Here is what I might simplify the first part of your code to:

Code:

---

#include <string>
#include <cstring>
#include <cctype>
#include <iostream>
#include <cassert>

using namespace std;

enum ET_strippingMode {
    exclusive = 0,
    inclusive = 1
};

bool searchStringIsSpecialString(const char searchString[]) {
    return strcmp(searchString, "A-Za-z") == 0;
}

bool sourceStringContainsAlphabetChar(const char sourceString[]) {
    //we are assuming 'sourceString' is pointed at a null-terminated char array...
    unsigned int sourceStringSize = strlen(sourceString);

    for (unsigned int i = 0; i < sourceStringSize; i++) {
        if (isalpha(sourceString[i])) {
            return true; //at the first char in the range of A-Z or a-z
        }
    }

    return false;
}

---

---

Good idea for those two functions. Changed my code to that.

Quote:

---

Code:

---

bool containsString(const char sourceString[], const char searchString[]) {
    //assume that 'sourceString' and 'searchString' are pointing at two null-terminated char arrays...
    unsigned int sourceStringSize = strlen(sourceString);
    unsigned int searchStringSize = strlen(searchString);

    if (searchStringSize < sourceStringSize) {
        //if the special string "A-Za-z" was passed into the 'searchString' parameter
        //then search for a character in sourceString in the range of A-Z, or a-z,
        //otherwise check to see if the searchString is contained in the sourceString.
        return searchStringIsSpecialString(searchString) && sourceStringContainsAlphabetChar(sourceString)
            || sourceString.find(searchString) != string::npos;
    }
    else if (searchStringSize == sourceStringSize) { //meaning the two strings have the same length
        //if the special string "A-Za-z" was passed into the 'searchString' parameter
        //then search for a character in sourceString in the range of A-Z, or a-z,
        //otherwise check to see if the source string and the search string are the same.
        return searchStringIsSpecialString(searchString) && sourceStringContainsAlphabetChar(sourceString)
            || strcmp(sourceString, searchString) == 0;
    }
    //since the string searched for is longer string than the one being checked,
    //there is no possibility of the shorter string fully containing the longer string.
    return false;
}

---

---

Thanks. That is a lot cleaner and shorter code. But you should note though that you forgot to handle the case where the searchString is longer than the sourceString, but searchString is the special string. You just automatically assumed it should return false in that case.

Quote:

---

Now, for the stripString function: I suggest that you break it up into two helper functions: one to handle exclusive stripping mode, the other to handle inclusive stripping mode.

---

Hmm...I may do that, but I think it would be a ...........

Quote:

---

EDIT:
Oh, and another thing for stripString: instead of nesting a do while loop in a for loop, first use one loop to find the start character, and then use a separate loop for the rest of the parsing.

---

So I guess set a bool var to true for that then? Yeah, that should work...

Taking my own advice, I see that it is simpler than expected:

Code:

---

#include <string>
#include <cstring>
#include <cctype>
#include <iostream>
#include <stdexcept>

using namespace std;

enum ET_strippingMode {
    exclusive = 0,
    inclusive = 1
};

bool searchStringIsSpecialString(const char searchString[]) {
    return strcmp(searchString, "A-Za-z") == 0;
}

bool sourceStringContainsAlphabetChar(const char sourceString[]) {
    //we are assuming 'sourceString' is pointed at a null-terminated char array...
    unsigned int sourceStringSize = strlen(sourceString);

    for (unsigned int i = 0; i < sourceStringSize; i++) {
        if (isalpha(sourceString[i])) {
            return true; //at the first char in the range of A-Z or a-z
        }
    }

    return false;
}

bool containsString(const char sourceString[], const char searchString[]) {
    //assume that 'sourceString' and 'searchString' are pointing at two null-terminated char arrays.

    //if the special string "A-Za-z" was passed into the 'searchString' parameter
    //then search for a character in sourceString in the range of A-Z, or a-z.
    if (searchStringIsSpecialString(searchString) && sourceStringContainsAlphabetChar(sourceString)) {
        return true;
    }
    else {
        size_t sourceStringSize = strlen(sourceString);
        size_t searchStringSize = strlen(searchString);
        if (searchStringSize < sourceStringSize) {
            //check to see if the searchString is contained in the sourceString.
            return strstr(sourceString, searchString) != NULL;
        }
        else if (searchStringSize == sourceStringSize) {
            //check to see if the source string and the search string are the same.
            return strcmp(sourceString, searchString) == 0;
        }
        else {
            //there is no possibility of the shorter string fully containing the longer string.
            return false;
        }
    }
}

namespace
{
    string stripStringExclusive(const char str[], size_t size, const char end, size_t i) {
        string strippedString = "";
        for (; i < size; ++i) {
            if (str[i] == end) {
                return strippedString;
            }
            else if (str[i] != ' ') { //skip spaces
                strippedString.push_back(str[i]);
            }
        }
        return ""; //end char is not present
    }

    string stripStringInclusive(const char str[], size_t size, const char end, size_t i) {
        string strippedString = "";
        for (; i < size; ++i) {
            if (str[i] == end) {
                strippedString.push_back(end);
                return strippedString;
            }
            else if (str[i] != ' ') { //skip spaces
                strippedString.push_back(str[i]);
            }
        }
        return ""; //end char is not present
    }
}

string stripString(const char str[], const char start, const char end, ET_strippingMode strippingMode = exclusive) {
    char* start_ptr = strchr(str, start);
    if (start_ptr == NULL) {
        return "";
    }

    if (strippingMode == exclusive) {
        return stripStringExclusive(str, strlen(str), end, str - start_ptr + 1);
    }
    else if (strippingMode == inclusive) {
        return stripStringInclusive(str, strlen(str), end, str - start_ptr);
    }
    else {
        throw invalid_argument("stripping mode must be either exclusive or inclusive");
    }
}

int main() {
    char filestreamInBuffer[] = "enum DAY {";
    if (containsString(filestreamInBuffer, "enum")) {
        cout<< "Yes, \"" << filestreamInBuffer << "\" does indeed contain \"enum\"." <<endl;
    }
    else {
        cout<< "No, \"" << filestreamInBuffer << "\" does not contain \"enum\"." <<endl;
    }

    if (containsString(filestreamInBuffer, ";")) {
        cout<< "Yes, \"" << filestreamInBuffer << "\" does indeed contain \";\"." <<endl;
    }
    else {
        cout<< "No, \"" << filestreamInBuffer << "\" does not contain \";\"." <<endl;
    }

    string currentEnumName = stripString(filestreamInBuffer, 'm', '{');
    cout<< "The name of the current enum is: " << currentEnumName <<endl;

    return 0;
}

---

Your code didn't output the stripped string...

http://codepad.org/uCWA3Ttd

Quote:

---

Originally Posted by laserlight

That is expected since the end character is not present in your test string.

---

Umm...actually, it is:

Code:

---

char filestreamInBuffer[] = "enum DAY {";
//...do stuff
string currentEnumName = stripString(filestreamInBuffer, 'm', '{');

---

And anyway, I wanted the stripString function to return the strippedString even if the end char was not found (I have my reasons). Only if the start char was not found, was the function supposed to return nothing.

Right. What is stripString supposed to do?

Quote:

---

Originally Posted by laserlight

Right. What is stripString supposed to do?

---

It is supposed to strip the sourceString from the start char to the end char if the end char is present, otherwise it should strip it until the end of the sourceString.
If the stripping mode is exclusive, it should exclude the start and end chars. Otherwise, if its inclusive, it should include them.

EDIT: Also, I wanted the function to return a completely different string than the one passed. I didn't want it to modify the original string by way of a reference or a pointer passed in its parameter.

What do you mean by "strip the sourceString"?

What my implementation does is that it extracts the characters from start to end, ignoring spaces. If it is inclusive, the start and end chars are extracted, otherwise they are not.

Quote:

---

Originally Posted by laserlight

What do you mean by "strip the sourceString"?

---

Return a stripped string containing chars from start to end (if exclusive). If its in inclusive mode, than the stripped string should contain also the start and end chars.

Quote:

---

What my implementation does is that it extracts the characters from start to end, ignoring spaces. If it is inclusive, the start and end chars are extracted, otherwise they are not.

---

No, it doesn't, because it obviously doesn't work, since it returns nothing, even though both the start and end chars are present in the string passed.

Quote:

---

Looking at your description, I would expect only a space to be extracted, since that is the only thing in between 'm' and '{'.

---

Actually, its not: The other thing is the string "DAY". :D

Grr... logic error: in stripString, subtract str from start_ptr, not the other way around. I'll leave the change to "strip it until the end of the sourceString" up to you.

Quote:

---

Originally Posted by Programmer_P

Yes, I know, but I figure my method is more clearer when you're looking at the code.
Anyway, none of this has anything to do with a if(false) statement being entered. So far, no one has commented on this...

---

It's not clearer...

Do you say to someone:
Is <some number> squared equal to nine?
or
Answer me "yes" if <some number> squared is equal to nine, otherwise answer me "no"?
or
If you would answer "yes" to "Answer 'yes' if <some number> squared is equal to nine, otherwise answer me 'no'" then answer "yes", otherwise answer "no"?

If you start comparing boolean expressions to true, where are you supposed to draw the line?:

Code:

---

if (x*x - 9)                                        //slightly obfuscated (taking things a bit too far)
if (x*x == 9)                                        //perfect
if ((x*x == 9) == true)                                //barely acceptable
if (((x*x == 9) == true) == true)                //very bad
if ((((x*x == 9) == true) == true) == true)        //insanely ridiculous

---

Forgetting that this is a crap example for a moment...
A few people code with a style equivalent to the first or third line above, but most people code in such as style as the second line. The level of redundancy you're coding to puts your code on par with the fourth line!

Code:

---

char* start_ptr = strchr(str, start);

---

Why does the strchr function want a int type passed into its second parameter, and why does passing a char into it work without any compiler error?

@laserlight: Btw, I got your code to work:
C++ code - 124 lines - codepad

Quote:

---

Originally Posted by Programmer_P

Why does the strchr function want a int type passed into its second parameter,

---

Possibly to work with EOF which is not a char? Anyway, many C string functions are like this -- the trend is to favour more inclusive and generic types, I think.

Quote:

---

and why does passing a char into it work without any compiler error?

---

It will generate a warning probably, if you have warnings enabled. Just cast (int).

They pre-date ANSI C (and prototypes). In the K&R world, all parameters tended to get promoted to int or double, unless expicitly cast.

Variadic parameters for the likes of printf etc still follow default promotion rules.

Quote:

---

Originally Posted by Programmer_P

Why does the strchr function want a int type passed into its second parameter

---

MK27's explanation of EOF is feasible, though it might be more likely because character literals in C are of type int, and strchr is taken from the C standard library.

Quote:

---

Originally Posted by Programmer_P

why does passing a char into it work without any compiler error?

---

Conversion from char to int is perfectly fine. It will probably not generate any warnings, and I do not recommend a cast.