HTML to TXT

**eobergfell** · 06-27-2007

Hello Everybody,
I have a question regarding the conversion of files from HTML to TXT. Im fairly new to the C language but understand the basics. Im a little confused with what to do.I have to write a program that asks the user to input a html file and then the program converts it to txt.
Thanks-

**MacGyver** · 06-27-2007

You may have a question, but you failed to ask it.

**eobergfell** · 06-27-2007

huh?

**mike_g** · 06-27-2007

You would want to open the HTML file for reading. Read it and output the contents to another file with a .txt extension.

You can find out how to do that here:
http://faq.cprogramming.com/cgi-bin/...&id=1043284392

**MacGyver** · 06-27-2007

Originally Posted by eobergfell

huh?

That's the first question you actually asked.

Look at your first post. There is no question. You just said you wanted to make a new txt file from an HTML file. Someone concluded that meant you just want a copy of the same file with a different file extension. I think it could mean you want to strip all the tags and create a new txt file with just straight text from the original file.

You were first too ambiguous in your initial description of what you wanted to do. Your second fault was to not even ask a question, and somehow hope we would just know what to say.

Have a look at this: http://www.catb.org/~esr/faqs/smart-questions.html

**Adak** · 06-27-2007

Originally Posted by eobergfell

Hello Everybody,
I have a question regarding the conversion of files from HTML to TXT. Im fairly new to the C language but understand the basics. Im a little confused with what to do.I have to write a program that asks the user to input a html file and then the program converts it to txt.
Thanks-

Your program will be very similar to one's that count up words. You'll need to keep track of the current position in the file, and whether that position is INSIDE an html tag, (in which case the text char's found there do NOT need to be copied to the text file you're program will make), or that position is OUTSIDE any html tag, and therefore is just text char's that DO need to be copied into the text file.

So your program will scan the html file, and keep an integer flag variable "Inside", equal to 1 (yes, it's inside currently), or equal to 0, (no, it is not inside, currently).

Now all you have to do is learn what are the border char's, that always distinguish html tags?

The more specific your questions are, the better we can answer your questions. Be sure to post up some code of yours as you ask, as well. That will also help guide our communications, and show that you are working on it, also.

**eobergfell** · 06-27-2007

Thanks mike and adak and to MacGyver keep being a sweetheart okay?

**eobergfell** · 06-27-2007

Yea I have to strip off all the tags <> and make it into a readable plan TXT file

**dwks** · 06-27-2007

If you want to strip XML (or HTML, same thing basically) properly, you need a finite state machine (never mind, it's just a term) with three states.

Outside any tag.
Inside a tag.
Inside an attribute.

If you don't have the last state, your code would be fooled by

Code:

<tag data=">" />

Anyway, the logic is quite simple. If you're outside a tag, look for a '<', which would involve changing to mode inside-a-tag. If you're inside a tag, look for a ' or ", which would make the mode inside-an-attribute. Ditto for the closing characters.

**eobergfell** · 06-27-2007

Code:

#include<stdio.h>
int main()
{
        int a, b;
        int input;
        FILE* outfile;
        char filename[80];
        printf("Enter Filename: ");
        gets(filename);
        outfile = fopen(filename, "w");
        scanf("&#37;d", &input);

        fprintf(outfile, "%d\n",input );
        fclose(outfile);
}

thats what ive done so far...I understand the logic behind this "If you're outside a tag, look for a '<', which would involve changing to mode inside-a-tag. If you're inside a tag, look for a ' or ", which would make the mode inside-an-attribute" But I don't know how to write it out.
Help, Anybody?

**dwks** · 06-27-2007

Don't use gets(). It's bad. See the FAQ.

How are you going to read from the file if you open it in "w" mode? Or are you reading from stdin?

Regardless, the logic goes something like this.

Code:

int mode = outside;

while(get a character) {
    if(mode == outside) {
        if(char == '<') mode = inside;
        else write character
    }
    else if(mode == inside) {
        if(char == '>') mode = outside;
        else if(char == '"' || char == '\'') mode = attribute;
    }
    else {  /* mode = attribute */
        /* ... */
    }
}

An enum would be useful here. http://www.cprogramming.com/tutorial/enum.html

Code:

enum {
    MODE_OUTSIDE,
    MODE_INTAG,
    MODE_INATTRIBUTE
} mode = MODE_OUTSIDE;

But you can use ordinary variables too of course.

**Adak** · 06-27-2007

Using gets() on a C board is sort of like putting a match out on a burn victim - very touchy about that. Check the faq on how to use fgets() which is safe, for input from stdin (when needed in your program). You won't get much help beyond "don't use gets()!" until that's done.

You'll definitely want a variable named "InsideTag" (or OutsideTag), and InsideAttrib, (or OutsideAttrib). Please get those added and initialized to out.

While you're trying to code this, I'd temporarily /* remark out */ the file names, etc., until you get a handle on the logic. It will greatly reduce your frustration, and speed everything up.

**dwks** · 06-28-2007

You'll definitely want a variable named "InsideTag" (or OutsideTag), and InsideAttrib, (or OutsideAttrib). Please get those added and initialized to out.

It makes more sense to have a single variable to hold three states, since the program can only be in one state at once. But two variables would work as well.

**eobergfell** · 06-28-2007

Hey everybody thanks, im having trouble with using the "enum function"

Code:

enum Status{OUTSIDE = ?, INSIDE = ?);

I dont know what to set OUTSIDE and INSIDE equals to.
If anyone could help I would really appreciate it. Thanks-

**dwks** · 06-28-2007

The first value in an enum is automatically set to zero. Succeeding values are automatically set to one more than the previous value. But you can override this if you like.

Code:

enum numbers {
    minusone = -1,
    zero,
    three = 3,
    four,
    five
};

But the point of enums is that the values don't matter. You just have something like

Code:

enum {
    MODE_OUTSIDE,
    MODE_INTAG,
    MODE_INATTRIBUTE
} mode = MODE_OUTSIDE;

and you always compare the enum to those values.

Code:

if(mode == MODE_OUTSIDE) mode = MODE_INTAG;
if(mode != MODE_INTAG) puts("not in tag");

Using enums the way they are intended to be used, it should not matter what the values of the enumerations are.

Thread: HTML to TXT

Thread Tools

Search Thread

Display

HTML to TXT

?

Similar Threads

Please Help - C code creates dynamic HTML

Writing an HTML Preprocessor

Parsing HTML files

Design + HTML