I was using fscanf to read in a file and examine each token of this complete file one at a time for the the occurence of a specific keyword. However I want to read one line from the file. Examine each token in the line and if then when that full line has been examined. I read the next line and examine that until the end of the file.
char word[MAX_LENGTH];
FILE *inputfile;
while( fscanf( inputfile, " %s", word) !=EOF )
{
//Examine word here
}
The above section of the code basically tokenises the complete file until finished. I want it to do one line and then the next.
Is there a way when using fscanf of knowing when I have reach a newline ?
or should i use scanf ?
Basically I need to know when I have ancountered a newline
Cheers Dave
This Question has been solved and asker verified All Experts Exchange premium technology solutions are available to subscription members.
Experts Exchange has been collecting answers to technology questions since 1996…3 million and counting! If you have a question, chances are we already have your answer.
If you can't find the exact answer you're looking for, ask our exclusive community of 50,000 experts. You’ll get a personalized answer from a trusted professional.
Thousands of free tech tips, tricks, how-to’s and tutorials are available in our peer reviewed articles section. See for yourself how smart our experts are, no login required.
Save solutions to your questions, answers you’ve discovered through searching plus helpful articles in your personal knowledgebase for easy future access.
Access the answers to your technology questions today.
30-day free trial. Register in 60 seconds.
Members of the expert community talk about why the experience at Experts Exchange is different than what you will find anywhere else.
Try it out and discover for yourself.
30-day free trial. Register in 60 seconds.
Join the community of experts here and help other tech pros by answering question in your area of expertise. You can earn FREE access to all Experts Exchange's premium features and resources.
If i was to use sscanf, going on example si have looked at. You seem to need to know how many tokens there will be in the input buffer (String) in order to store it somewhere.
Okay -I could use the following:
ifstream inStream(inputDir, ios::in);
while (!inStream.eof())
{
inStream.getline(line, 255, '\n');
//My Scanf here (or strtok)
sscanf( line, " %s", word);
}
I am not sure how to break the string into clean tokens here though.
All I want is to be able to examine each token now from the line. So %s would be okay but when i do the above it stops after the first token and doesnt read the next. How should that line read.
Cheers
As I alluded, a more robust (see also, complicated) format specifier will let you read the entire line with fscanf(). (Personally, I much prefer fgets().)
Using your original code:
char word[MAX_LENGTH];
FILE *inputfile;
while( fscanf( inputfile, " %[^\n]", word) !=EOF )
{
//Examine word here
}
This tells fscanf() to read until newline (\n) is reached.
Kent
Dave,
The way I'd do it is to use:
char terminator;
while( fscanf( inputfile, "%[^ \n]%c", word, &terminator) !=EOF )
This is a lot like Kents but I am reading up to the first space or newline and collecting whatever caused it to stop. You can then check 'terminator' to see what stopped it. If it's a '\n' then, obviously, you hit a newline.
The main difference between this and what you had will be that multiple spaces will be seen as multiple tokens.
Paul
Ahh.... I misunderstood. Thought that you wanted to read an entire line, processes each word, and then read the next line.
Combining your original code and Paul's suggestion:
char word[MAX_LENGTH];
char EndChar;
FILE *inputfile;
while( fscanf( inputfile, " %s%c", word, &EndChar) !=EOF )
{
// Examine Word
if (EndChar == '\n')
// End of line after token
}
Kent
Dave,
One more thought! Use "^ \r\n" just to be sure. In this way it will still work if you open the file in binary mode.
With mine, you wont get spaces. Note that there is a space between the '^' and the '\r'. This will ensure you dont get spaces in your token. The fscanf will stop on the first character that IS in the list, i.e. space, cr, lf, and then put the character it DID get into 'terminator'. You can then check 'terminator' to see what it was.
Paul
Hi Lads
You are both on the right track. I think I have confused you both a bit though .
Sorry for not being clearer with my explantion.
Based on what I am trying above this is what I want.
[1] Open up a file for reading.
[2a] Read (using fscanf) tokens from the file input stream. (Tokens which are sperated by space)
These tokens can with either a normal text or a '\n'
If the token is a '\n' continue (I only want to be aware that I am on a new line)
I ran the code above but in some cases when I perform the
if (EndChar == '\n')
I get true when it is actually a space
Also for some lines it reads the first or maybe the seconds tokens and ignores the rest of the line
and also it stopped half way through the line and just finished.
Any ideas why
This is the code:
DIR *dirp;
FILE *input_file_pointer;
char inputDir[MAX_LENGTH];
char word[MAX_LENGTH];
char terminator;
ifstream inStream(inputDir, ios::in);
input_file_pointer = fopen(inputDir, "r");
while( fscanf( input_file_pointer, " %s%c", word , &terminator) !=EOF )
{
if(terminator == '\n')
printf("New Line \n");
else
printf("[%s]\n",word);
}
>>Also for some lines it reads the first or maybe the seconds tokens and ignores the rest of the line
This is probably the definition of EOF. Use '> 0' instead of '!= EOF' because fscanf actually returns the number of items read. Note, it maytherefore be worthwhile saving the number of items read because if it was equal to '1' then the terminator character hasnt been read; the end of the file was reached first.
Paul
Hi Lads
Okay Brilliant I am nearly working fully. What you ahve suggested worked
I put in the following code and it works very well. The only small thing that I need is
If there is a blank line in the file. How would I detect it and print it out ?
while( fscanf( input_file_pointer, " %s%c", word , &terminator) > 0 )
{
if(terminatorFlag == TRUE)
{
//reset Flag
terminatorFlag = FALSE;
printf("\n");
}
if(terminator == '\n')
{
terminatorFlag = TRUE;
printf(" ^%s^",word);
}
else
{
printf(" [%s]",word);
}
}
Note that in Kents code above, if the last word does not have a space or a \n at the end, the fscanf will only read 1 field and will leave the terminator as it was. This would only miscalculate your line count if this word was alone on the line but I'd probably make this simple change:
while( (count = fscanf( input_file_pointer, " %s%c", word , &terminator)) > 0)
...
if(count > 1 && terminator == '\n') /* Also, you should get a value placed in *word* */
...
Paul
Dave,
>>The count always returns 2 as it trys to read a %s and a %c for each scan.
What happens if your file just contains one word, no spaces, no returns? I'd bet it returns 1.
>>It doesnt appear to read a blank line - Just lines with text on them
That's because you are using an initial space in " %s%c". This space eats all whitespace characters until a non-whitespace, i.e. blank lines. Using "%[^ \r\n]%c" would act slightly differently, as would "%s%c" (no initial space).
If you are happy with the solution you have then stick with it. Dont worry about the details. Deep knowledge of scanf formats will come with years of use.
Paul
Hey Paul
I noticed a small problem but an imprtant one for me.
I can submit this as a new question if youd like as I know you have spent enough time alread on it.
The question is that it misses the very last token in the file
So for example if there was closing brace on the last file as follows
//
} <- it does not read this one
while( (fscanf( inStream, " %s%c", word , &terminator)) !=EOF )
{
//Do Stuff here
}
Dave
Hi Dave,
If you check, you'll probably find that a lot of other characters are also getting treated as a terminating character OR combined into a concatenated string. Exact syntax and spacing is an issue here.
"abcd efgh" and "abcdefgh" are clearly different strings.
"abcd 1111" and "abcd1111" are also different.
"while (" and "while(" are too.
The solution is a really creative format string, or to parse the data yourself.
Kent
Hi Dave,
It looks like you're trying to parse C source code. Or perhaps other text that is quite C like.
scanf() poses challenges here that are tricky to overcome.
-- 'A' is really a single token.
-- "This is really a single token, too"
-- "As is this, even when you \"quote\" embedded text"
-- X0 is a valid C identifier,
-- 0X is not.
-- 0f isn't either, but it IS a valid constant
etc....
Yeah I am reading in Java files. Examing each line and then ouputting the line to a new file in a different directory.
Hence I need to know when I have encountered the end of a line.
So I read each word of the java file and when I encounter certain key words I may do something to it and then I output to file.
I nearly have it working perfect until I discovered one or two files didnt compile afterwards.
This was because for example I had to lines as follows:
// 1 corresponds to second field on summary screen, face value \n <- new line char here
treePane.setTransferInfo(v
When I parsed the lines and outputted (unchanged) to file. it came out like all in one line which causes a compile error. For obvious reasons the fscanf loop failed to detect the \n at the end of the first line as it was two spaces on from the end of the line and
my parser was looking for '%s%c' instead of '%s %c'
How would I sort this out once and for all
Cheers Dave
I put the \n in abovefor demonstration purposes
See above comment as well :
Paul this code:
while( (fscanf( inStream, "%[^ \r\n]%c", word , &terminator)) !=EOF )
{
}
Reads the tokens until the first \n but then just loops on the last word. How do I get it to go to the next line and continue on ?
I think it must be staring me in the face but I its not clicking
Sorry
Some tokens are really multi-character tokens.
a *= b;
For parsing, you don't care if the operator '*=' is treated a one or two tokens.
a *= b;
a*=b;
a * = b ;
are all valid and should produce the same result.
a *= /* compute residual */ b;
But in the line above, '/*' and '*/' clear DO get treated as a single entity.
Similarly, you don't really want to parse anything past '//', instead you want to copy the rest of the line. (This appears to have bitten you in your example above.)
I tend to parse this kind of text myself. The code gets a bit long, but to me, it's easier than grabbing generic text from the source file and trying to keep track of all of the possible embedded statuses.
Having processed:
a =
a *=
You have the same status, but the second line has an embedded status because of the pending multiplication. A parser will simply put this on the stack, but your process doesn't sound recursive and doesn't seem to utilize a stack.
In general, I have a generic function called get_token(). With each call it checks to see if there is any data in the "line buffer" and if not it reads a line from the source file If there is data in the line buffer, it continues with the current text. I then grab the first character and build my token based on the character.
while (1)
{
if (LineEmpty)
{
fgets (LineBuffer, LINE_LENGTH, stdin);
if (feof (stdin))
break;
ptr = LineBuffer;
}
while (*ptr == ' ' || *ptr == '\t') // skip space and tab chars
*ptr++;
if (*ptr == '\n')
continue;
end = ptr;
switch (*ptr)
{
case '(':
case '{':
case '[':
// etc. Handle single character tokens
break;
case '=': // check to see if '=' or '==' token
break;
case '-':
case '*:
case '+' // check for +, +=, -, -=, *, *=
break;
case '/': // check for /, /=, /*, //
break;
case '#' // check for #include, #define, #if, #pragma, etc
break;
case '"': // extract string
break;
case '/'': // extract quoted single character
break;
default:
if (isapha (*ptr))
// handle alpha data
else if (isdigit (*ptr))
// handle numeric data
else
// error...
Of course, this may be overkill for what you're trying to do....
Kent
Business Accounts
Answer for Membership
by: 9703403Posted on 2005-02-21 at 06:45:47ID: 13363062
TYPO: I meant sscanf above not scanf