Hi guys, newbie here, some questions!

Hi there :)

Ok, basically I'm pretty new to C and have created a mashed-together program using different elements from different existing programs:

Code:

#include <stdio.h> #define ALPHABET_TOTAL 26 /* Defines the total number of letters in the alphabet */ main() { int text[ALPHABET_TOTAL] = {0}, /* Creates an array of size 26, acting as counters initialized as 0*/ text_input, /* Reads in the text input as characters */ ASCII_count, /* ASCII number of the letters in the alphabet */ alpha; /* boolean flag which turns to 0 (false) in case of a non-alpha character */ /* Setting values to 0 */ text_input = 0; ASCII_count = 0; alpha = 0; /* Get input - this needs to change!! */ FILE *userfile; /* declares file pointer */ printf("Enter file name to load: "); scanf("%s", filename); userfile = fopen("%s", "r"); /* Perform text analysis*/ while( text_input != EOF ) { alpha = 1; text_input = getchar(); /* If input is alphabetical, convert to a number 0-25 so that it can be used as a locator along the array */ if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; /* Non-alphabetical character, set the alpha flag to false */ else alpha = 0; /* Increment the appropriate counter in the text array as long as the character is alphabetical */ if(alpha) text[text_input]++; } /* Display results */ for (ASCII_count; ASCII_count <= 25; ASCII_count++) printf("\nTotal %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); }

As you can probably tell, I'm trying to prompt the user for a file name then conduct a letter frequency analysis on the text contained in said file.
What's wrong with what I've got? I'm guessing it's the code prompting for the user file.

Also, If I wanted to add in a counter to keep track of the total number of characters in the text, whereabouts in the various loops would it go?

Any help would be greatly appreciated. I really want to get to grips with C!

OK. First, main should return int, so it should be int main.
For reading with scanf, see http://cboard.cprogramming.com/showp...37&postcount=9
The open line is totally wrong. You're trying to open file "%s", which is totally illegal on Windows filesystems, and some others too. But not the one you want to open anyway. You stored that filename into the "filename" variable, so pass that instead.
Then two suggestions: Indent with more than one space (the ifs) and keep it consitent (that is, indent using the same number of spaces everywher; makes it easier to read). And secondly, indent the comments too.
And comments can be placed after code, too, not always just on a new line.

As for your counter, it's very simple. Why not actually read and try to understand the code? If you do, you'll see it becomes very easy.

I think before 'mashing' together anything you need to go over the concepts of basic C.
It would be good practice to use {} when starting and ending if's and loops.

Start the main function like so:

Code:

int main (void) { return 0; }

Loops

Code:

while (i != 10) { DO SOMETHING i++; }

Opening files

Code:

FILE *fo; fo = fopen("filename.$", "#"); //# can be a, r or w, or some others, research !!! //$ can be only basic text files like rtf or txt

Copy from one file, writing to another

Code:

FILE *fo; FILE *fw; char c; fo = fopen("C:\\Directory\\Directory\\Directory\\Filename.txt","r"); fw = fopen("C:\\Directory\\Directory\\Directory\\Output.txt", "w"); while(1){ c = fgetc(fo); if(c == EOF){ break; } else{ PERFORM ACTIONS ON CHARACTERS HERE fputc(c,fw); } }

Hope some of that helps

Cheers guys - I think I do need to try and comprehend the program as a whole more

Quick question - is it possible to includemore than one 'else if' branch?

so in this context:

Code:

while( text_input != EOF ) { alpha = 1; text_input = getchar(); if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; else if(text_input != ' ') char_counter++1 else alpha = 0; if(alpha) text[text_input]++; }

??

Quote:

Originally Posted by fuso

Quick question - is it possible to includemore than one 'else if' branch?

As many as you like. Just try to indent a little better.

Code:

while( text_input != EOF ) { alpha = 1; text_input = getchar();

in this code you firstly process the EOF as regular char, and only then - exit the loop.
In general - if should be opposit - first thing after input - check the EOF and only if the input is ok - do the rest...

Quote:

Originally Posted by Elysia

As many as you like. Just try to indent a little better.

Just to note, nested if/else code blocks are limited to 15 for c89 and increased to 127 for c99 c coding standard I believe.

If you happen to have that many try converting those to a switch statement.

Quote:

Originally Posted by slingerland3g

Just to note, nested if/else code blocks are limited to 15 for c89 and increased to 127 for c99 c coding standard I believe.

If you happen to have that many try converting those to a switch statement.

Where is this stated? The only limit of 15 is the level of include nesting. There are many other limits that may come into the number nested if-statement, but certainly section 5.2.4.1[which I beleive is the relevant section] doesn't list such a limit.

--
Mats

Quote:

Originally Posted by matsp

Where is this stated? The only limit of 15 is the level of include nesting. There are many other limits that may come into the number nested if-statement, but certainly section 5.2.4.1[which I beleive is the relevant section] doesn't list such a limit.

--
Mats

The c99 ISO standard I have reflects

"[#1] The implementation shall be able to translate and
execute at least one program that contains at least one
instance of every one of the following limits:12)

-- 127 nesting levels of blocks "
-- 63 nesting levels of conditional inclusion
:
:

You are right about the #include nested statements.

63 nesting levels is quite a lot however, and it's a MINIMUM, not a MAXIMUM. So as long as you give the compiler sufficient memory, and it's a modern compiler, it probably copes with a lot more.

I just wrote a "C code generator" to generate a LONG chain of if/else if statemements, along this concept:

Code:

E:\Temp>codegen 3 #include <stdio.h> int main() { int x; printf("enter a number:"); scanf("%d", &x); if (x == 0) printf("x = 0\n"); else if (x == 1) printf("x = 1\n"); else if (x == 2) printf("x = 2\n"); return 0; }

It compiles and generates correct code in gcc for 3000 if-statements.

Edit: MS "cl" version 13 fails to compile the same source. It does compile with about 500 if-statements.

--
Mats

Quick question regarding file opening;

My compiler tells me I have an undefined symbol 'filename' in function main; how (and where) do I define it?

Code:

printf("Type a file name, followed by the <enter> key : "); gets(filename); fp=fopen(&filename[0],"a+");

If anyone can help then thanks.

Quote:

I have an undefined symbol 'filename'

Have you defined filename variable?
do not use gets - read FAQ

Code:

char filename[MAX_PATH]; if(fgets(filename, sizeof filename,stdin) == NULL) { /* failed to read a name */ }

Don't use gets(something), use fgets(soemthing, sizeof(something), stdin). That prevents someone leaning on the 'a'-key for 2 minutes and then hitting enter from crashing your program [aka "the Mattias method", which involves putting your entire lower arm straight across the keyboard].

A good declaration for filename would be

Code:

#include <limits.h> // Should be around your other #include statements. ... char filename[MAX_PATH]; ...

As to where you declare filename[]: it should be declared in the scope where it's needed - so in your case, I suppose at the top of the function main().

Code:

fp=fopen(&filename[0],"a+");

Can be written as:

Code:

fp=fopen(filename,"a+");

And that is what you really should be doing.

--
Mats

I think for my purposes, gets() should be fine - the filename will not be longer than 99 characters.

Do I define 'filename' as a string?

Quote:

Originally Posted by fuso

I think for my purposes, gets() should be fine - the filename will not be longer than 99 characters.

No! Never.
The reason everyone is suggesting fgets (which you should use) is because it is not prone to buffer overruns. Buffer overruns is a security issue, meaning that others can take control of your (!) and others (!!) computer, or even worse stuff.
So never ever use gets. Always use fgets. Fgets is safe.
In the Christian mythology, gets is the devil. It should never have been invented. You should stay away from it. Forget it exists. Toss it into the bowels of oblivion and teach all your friends to do the same.

Quote:

Do I define 'filename' as a string?

If it is used to hold text, then yes. That's what strings are for, to hold text, characters and strings.

Quote:

Originally Posted by fuso

I think for my purposes, gets() should be fine - the filename will not be longer than 99 characters.

Do I define 'filename' as a string?

And what happens if the user leans on the keyboard producing more than 99 chars?

Yes, I agree it's the users fault for leaning on the keyboard, leaving something like this

Code:

gfffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff

in the filename, but it's still bad practice to use input functions that allow the application to crash with an error message when the user happens to input too much data.

And why do you believe that a filename should always be less than 99 chars? Admittedly it's unpractical, but MAX_PATH in windows is 260 chars, and in Linux somehting like 1024 chars - so if the OS can support the filename, why should your application deny it from being used in your app?

"string" is not a concept that really exists in C, as it is part of C++. None of your other code is C++.

--
Mats

Ok, I think I'm using fgets properly...

This program is doing my head in. I'm struggling. It doesn't work. Any freindly help would be greatly appreciated!

Code:

FILE *fp; int text[ALPHABET_TOTAL] = {0}, text_input, ASCII_count, alpha; char filename[20]; /*initializing*/ text_input = 0; ASCII_count = 0; alpha = 0; /*user prompt*/ printf("Type a file name, followed by the <enter> key, Ctrl+Z, and then <enter key> again : "); fgets(filename,100, stdin); fp=fopen(&filename[0],"r"); /*text analysis*/ while( text_input != EOF ) { alpha = 1; text_input = getchar(); if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; else alpha = 0; if(alpha) text[text_input]++; } /*display results*/ for (ASCII_count; ASCII_count <= 25; ASCII_count++) printf("\nTotal %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); }

You're not using it right:

fgets(filename,100, stdin);
Should be
fgets(filename,sizeof(filename), stdin);

Remember that if you lie about the buffer size to fgets, it's not better than gets.

fp=fopen(&filename[0],"r");
Is the same as
fp=fopen(filename,"r");

Also from what I can see, you're not actually doing anything with the translated characters. What are you supposed to do with them? Output to screen?

Did you even read my reply?

Well, if you declare a filename[20], then you should not use fgets(filename, 100, stdin), should you? That will allow 80 chars more than what you have space for to be read in.

I suggested that you use sizeof(filename), and that you use limits.h as an include, then use MAX_PATH to size your filename - that way you are GUARANTEED that all filenames that are allowed on that machine can fit within your fileaname, not just ones that fit in some arbitrary size that you think is big enough, be that 20, 99 or some other number.

Code:

fp=fopen(&filename[0],"r");

is still not fixed [although it technically won't cause any problem].

I think the point I missed with fgets() is that it leaves a newline at the end of the string, so you need:

Code:

len = strlen(filename); if (filename[len] == '\n') filename[len-1] = 0; else // ... filename didn't fit in the variable...

--
Mats

Quote:

if (filename[len] == '\n') filename[len-1] = 0;

It's wrong

len-1 should be in both places...

But I prefer

Code:

char* p = strchr(filename, '\n'); if(p) *p = '\0';

Are these ammendments a necessity in order for the program to work? Or are they sophisticated tricks?

I don't think I'm going about the right way of reading characters from the file either; do I need

Code:

text_input = getchar(filename);

or something along those lines? Actually that's definately wrong.

Quote:

Originally Posted by vart

It's wrong

len-1 should be in both places...

But I prefer

Code:

char* p = strchr(filename, '\n'); if(p) *p = '\0';

Good spot - and strchr() and strlen() are about the same performance in the trivial implementation. strlen() can be made more optimal because we KNOW we're searching for a zero, and there are special tricks for doing that on multiword basis. But for this purpose, it's irrelevant.

--
Mats

These are not tricks. Not the newline remove. It is required to open your file.
Oh and you need to use fgetc to read a character from a file.

Here is the appended version (I hope!)

It now freezes after returning a file name...

Code:

#include <stdio.h> #define ALPHABET_TOTAL 26 int main() { /*defining variables*/ FILE *fp; int text[ALPHABET_TOTAL] = {0}, text_input, ASCII_count, alpha; char filename[20]; char* p = strchr(filename, '\n'); if(p) *p = '\0'; /*initializing*/ text_input = 0; ASCII_count = 0; alpha = 0; /*user prompt*/ printf("Type a file name, followed by the <enter> key : "); fgets(filename,sizeof(filename),stdin); fp=fopen(filename,"r"); /*text analysis*/ while( text_input != EOF ) { alpha = 1; text_input = fgetc(filename); if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; else alpha = 0; if(alpha) text[text_input]++; } /*display results*/ for (ASCII_count; ASCII_count <= 25; ASCII_count++) printf("\nTotal %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); }

arrrgrhgrgraararararrrrgg!!!

Code:

char* p = strchr(filename, '\n'); if(p) *p = '\0';

This portion needs to be AFTER the input of the filename.

You may have to declare char *p up at the top of the function, and then use p =strchr(...) after the fgets(). The purpose of the above snippet is to make sure the newline that fgets() took in is removed from the string itself.

--
Mats

Make sure you actually opened the file successfully before trying to read from it; otherwise you are going to get an endless loop.
(If you can't open the file, you should probably print an error, and optionally, let the user try again.)

Ok so now it's physically opening the file I type in (i.e. it loads notepad), which is intereting - a good thing you say?

How come it's ignoring the rest of my code after that?

You may need to add "getchar()" at the end of the program to prevent the console window from closing. Try that first.

Alternatively, for console programs, download cygwin, and compile with gcc, then u don't have a problem of console's closing.

Hmm, ok it doesn't notepad anymore, and in fact just freezes the command prompt

You should pay attention to the warnings of th ecompiler

Code:

#include <stdio.h> #include <string.h> #define ALPHABET_TOTAL 26 int main() { /*defining variables*/ FILE *fp; int text[ALPHABET_TOTAL] = {0}, text_input, ASCII_count, alpha; char filename[20]; /*initializing*/ text_input = 0; ASCII_count = 0; alpha = 0; /*user prompt*/ printf("Type a file name, followed by the <enter> key : "); if(fgets(filename,sizeof(filename),stdin)) { char* p = strchr(filename, '\n'); if(p) *p = '\0'; } else { puts("Failed to read a file name"); return 1; } fp=fopen(filename,"r"); /*text analysis*/ while( (text_input = fgetc(fp)) != EOF ) { if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; else continue; text[text_input]++; } /*display results*/ for (ASCII_count = 0; ASCII_count <= 25; ASCII_count++) printf("\nTotal %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); return 0; }

Well I tried compiling your program and I got implicit declaration of strchr, which means you didn't include some library, probably the strings library.

also it said that :

for (ASCII_count; ASCII_count <= 25; ASCII_count++)

was a statement with no effect

Well you didn't initialize this:

for (ASCII_count = 0; ASCII_count <= 25; ASCII_count++)

Whatever the value is meant to be

What compiler are you using?

Does it ask you for a filename? If not, try addding "fflush(stdout);" after the printf() asking the name.

--
Mats

I'm using the Borland C++

Yeah it does prompt me for a file name, in fact the whole thing works pretty well

Borland C++ is a pretty old, non-recommended compiler.
Try upgrading to, say, Visual Studio. Great IDE.

Ok, another quick question, more to do with the logical side of things.

Basically I've added a few lines in order to make a word count - using the number of spaces + 1, however my program seems to add to the count when it encounters punctuation too. I have no idea why!

Code:

#include <stdio.h> #include <string.h> #define ALPHABET_TOTAL 26 int main() { /*defining variables*/ FILE *fp; int text[ALPHABET_TOTAL] = {0}, text_input, ASCII_count, word_count; char filename[20]; /*initializing*/ text_input = 0; ASCII_count = 0; word_count = 1; /*user prompt*/ printf("Type a file name, followed by the <enter> key : "); if(fgets(filename,sizeof(filename),stdin)) { char* p = strchr(filename, '\n'); if(p) *p = '\0'; } else { puts("Failed to read a file name"); return 1; } fp=fopen(filename,"r"); /*text analysis*/ while( (text_input = fgetc(fp)) != EOF ) { if(text_input >= 'a' && text_input <= 'z') text_input -= 'a'; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A'; else if(text_input = ' ') word_count++; else continue; text[text_input]++; } printf(" Total words: %d " , word_count++); /*display results*/ for (ASCII_count = 0; ASCII_count <= 25; ASCII_count++) printf("\n Total %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); return 0; }

So set it to count a word only when there is a space encountered or a new line. As long as you don't do this:

hi , ok

it will work fine

That's what I thought I'd done...

Code:

else if(text_input = ' ') word_count++;

Quote:

seems to add to the count when it encounters punctuation too. I have no idea why!

Why do you think so?
Show your input and the output

Code:

else if(text_input == ' ')

Turn up them warnings. Some compilers warn about that.
Single = is assign and double == is compare.

Quote:

Originally Posted by Elysia

Code:

else if(text_input == ' ')

Turn up them warnings. Some compilers warn about that.
Single = is assign and double == is compare.

aha!

Thanks!

Ok, another one

Trying to print the 'average word length' which I'm trying to define as total characters/total words

Do all of these variables need to be defined as doubles from the offset?

Code:

#include <stdio.h> #include <string.h> #define ALPHABET_TOTAL 26 int main() { /*defining variables*/ FILE *fp; int text[ALPHABET_TOTAL] = {0}, text_input, ASCII_count; double letter_count, word_count, word_average; char filename[20]; /*initializing*/ text_input = 0; ASCII_count = 0; letter_count = 0; word_count = 1; word_average = 0; /*user prompt*/ printf("Type a file name, followed by the <enter> key : "); if(fgets(filename,sizeof(filename),stdin)) { char* p = strchr(filename, '\n'); if(p) *p = '\0'; } else { puts("Failed to read a file name"); return 1; } fp=fopen(filename,"r"); /*text analysis*/ while( (text_input = fgetc(fp)) != EOF ) { if(text_input >= 'a' && text_input <= 'z') text_input -= 'a', letter_count++; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A', letter_count++; else if(text_input == ' ') word_count++; else continue; text[text_input]++; } word_average = letter_count/word_count; /*display results*/ printf(" Average word length: %d ", word_average); for (ASCII_count = 0; ASCII_count <= 25; ASCII_count++) printf("\n Total %c or %c: %d", ASCII_count + 'a', ASCII_count + 'A',text[ASCII_count]); return 0; }

No, they don't need to be doubles because you're not using decimals.

Code:

if(text_input >= 'a' && text_input <= 'z') text_input -= 'a', letter_count++; else if(text_input >= 'A' && text_input <= 'Z') text_input -= 'A', letter_count++;

Should be

Code:

if(text_input >= 'a' && text_input <= 'z') { text_input -= 'a'; letter_count++; } else if(text_input >= 'A' && text_input <= 'Z') { text_input -= 'A'; letter_count++; }

{ marks the beginning of a block and } marks the end, just like your function.

When I define the variables as ints, the %d is rounded to a whole number in the printf call, why is it doing that?

You need to understand about how division works.
There are two types of numbers in C. One is integers, whole numbers. One is floating point, decimal numbers.
If you need/want the decimals, you need to make sure one variable is a floating point. You can do it via a simple cast.

Then, you need to tell printf that it's actually printing a floating point number. This can be done via %f.

So, all in all, it might look like:

Code:

float word_average = (float)letter_count/word_count; printf("%f", word_average);

If your going to keep adding, might I suggest organizing this into seperate functions.

Also a debugger usually helps to, it allows you to step through your code and look at variables. Visual Studio is good about this and gdb-insight is good as well, although it takes forever to build the files.