Using scanf with enum types

Hi all,

I've got my own enum type set up as follows

Code:

---

typedef enum aa
{
        a= 0, ala= 0, alanine= 0,
        r= 1, arg= 1, arginine= 1,
        n= 2, asn= 2, asparagine= 2,
        d= 3, asp= 3, aspartic_acid= 3,
        c= 4, cys= 4, cysteine= 4,
        q= 5, gln= 5, glutamine= 5,
        e= 6, glu= 6, glutamic_acid= 6,
        g= 7, gly= 7, glycine= 7,
        h= 8, his= 8, histidine= 8,
        i= 9, lle= 9, isoleucine= 9,
        l= 10, leu= 10, leucine= 10,
        k= 11, lys= 11, lysine= 11,
        m= 12, met= 12, methionine= 12,
        f= 13, phe= 13, phenylalanine= 13,
        p= 14, pro= 14, proline= 14,
        s= 15, ser= 15, serine= 15,
        t= 16, thr= 16, threonine= 16,
        w= 17, trp= 17, tryptophan= 17,
        y= 18, tyr= 18, tyrosine= 18,
        v= 19, val= 19, valine= 19
       
} aa;

---

And now I'd like to use scanf to read a line and store one part of it as this datatype aa.

The data I'm reading in takes the format of integercharacter, such as 19a, 23c, 243q, etc.

Currently I'm reading it in as follows

Code:

---


int focal_position;
char focal_amino_acid;

if (sscanf(string, "%d%c", &focal_position, &focal_amino_acid)!=2)
        {
                printf("could not read in data from string in function tokenizer()\n");
                exit(255);
        }

---

i.e. I'm reading in the a in 19a (for example) as a character, but I'd prefer to read it in as this enum type. Is there easy way to do that?

Thanks,
Brad

There is no way to tokenize a string in C, i.e. you can't read in a character or string of user input and be able to use what they entered as an identifier (vairable name, enum value, etc). You can either create a mapping that relates the string 19a to alanine or whatever, or you can try an interpreted language like PHP that allows tokenizing strings.

I don't need to be able to identify what they entered as an specific identifier. I know that the number will always represent the variable position, and that the character that immediately follows it will always represent the amino acid at the specified position. What I want to be able to do is read in the character that comes immediately after the specified position (e.g. the a that follows 19, in 19a) as an 'a' for the typedef enum specified above, and not an a.

I can think of crude/brute force approaches using a case statement, but I figured that there must be a more elegant way to read in a character and store it's value in an enum'd datatype as appropriate.

By the way, strtok allows you to tokenize strings in C, doesn't it?

Quote:

---

Originally Posted by bhdavis1978

By the way, strtok allows you to tokenize strings in C, doesn't it?

---

Yes, but the strings are broken into tokens that have nothing to do with the tokens in your C program.

Ahh, yes, but with a different meaning for "tokenize". As strtok means it, it breaks the string up into tokens, or chunks, based on certain characters, e.g. strtok("this is a string", " ") would break the string up into words at the space characters. The way I'm using tokenize refers to tokens from the compilers standpoint. A word the compiler can understand and map to a known identifier. If i have something like (this might not compile, it's for instructional use only):

Code:

---

enum {
    x = 17
};
char c = 'x';

if (c == x)
...

---

There is no way for the compiler to take the value in c and use that as an identifier. c refers to the ASCII character 'x', having a numerical value of 120, while the identifier x has a numerical value of 17. If this were PHP (a language I suck at), I think something like

Code:

---

$x = 17;
$c = "x";

if ($$c == $x) // this would evaluate to true.

---

There are clever tricks with something like X-macros (C preprocessor - Wikipedia, the free encyclopedia), or you could store the mappings explicitly somewhere, in a giant array at the top of yoru file (kinda ugly), or a text file to be read in and processed at startup (a bit nicer).

Not knowing your end game, how I'd approach this:

Code:

---

#include <stdio.h>

typedef struct AminoAcid_
{
    char abb1;
    char abb2[4];
    char fullName[14];
} AminoAcid;

static const AminoAcid AminoAcids[] =
{
    { 'a', "ala", "alanine" },
    { 'r', "arg", "arginine" },
    { 'n', "asn", "asparagine" },
    { 'd', "asp", "aspartic acid" },
    { 'c', "cys", "cysteine" },
    { 'q', "gln", "glutamine" }
};

static int getAminoAcidIndex(const char aa)
{
    int i = 0;
    while (i < sizeof(AminoAcids)/sizeof(AminoAcids[0]))
    {
        if (AminoAcids[i].abb1 == aa)
            return i;
        ++i;
    }
    return -1;
}

int main(void)
{
    int focal_position = 0;
    char focal_aa = ' ';
    while (1)
    {
        printf("Enter a number/AA combo (-1 to stop): ");
        if (scanf("%d%c", &focal_position, &focal_aa) == 2)
        {
            if (focal_position == -1) break;

            int aaIndex = getAminoAcidIndex(focal_aa);
            if (aaIndex != -1)
            {
                printf("Found %s (%s) @ %d\n",
                      AminoAcids[aaIndex].fullName,
                      AminoAcids[aaIndex].abb2, focal_position);
            }
        }
        int c = 0;
        while ((c = getchar()) != '\n') ;
    }

    return 0;
}

---

You can't read it directly as an enum, but you can convert a character into an appropriate number easily enough, like so.

Code:

---

char ch;
sscanf( string, "%c", &ch );
int val = ch - 'a';    // 'A' if your characters are uppercase.

---

You can then assign val to your enum variable.

If the format of the given string is strict ie with no spaces between the number and amino-acid name, instead of making an enum, i would rather treat it as a static C-string array with each element's index as its focal position.

Code:

---

#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <errno.h>

static const char *const amino_acids[ ] = {
                                    "alamine",
                                    "arginine",
                                    "asparagine"
                                  };

int parse_string( char s[ ] )
{
    char *p;
    long pos;
   
    pos = -1;
    p = NULL;
    if ( s == NULL ) {
        return -1;
    }
    errno = 0;
    pos = strtol( s, &p, 10 );
   
    if ( errno == ERANGE  || pos < 0 || p == NULL || *p == '\0' || ( size_t )pos >= sizeof amino_acids / sizeof *amino_acids
        || strstr( amino_acids[ pos ], s + pos ) == NULL ) {
        return -1;
    }

    return pos;

}

int main( void )
{
    char s[ BUFSIZ ];

    while ( fgets( s, sizeof s, stdin ) != NULL ) {
          char *p;
          int t;

          p = NULL;
          if ( ( p = strchr( s, '\n' ) ) != NULL ) {
            *p = '\0';
          }
          if ( ( t = parse_string( s ) ) != -1 ) {
              printf( "%s \n",amino_acids[ t ] );
          } else {
            puts( "Not found. Incorrect format" );
          }
    }

    return 0;
}

---

edit:
Caveats( bugs ):
1. If the name of an amino acid happens to be a part of another amino acid's name and the focal position matches, then it will incorrectly accept it.
2. If your abbreviations format changes: for eg: alanine is abbreviated as aln, then you may want to develop your own function to deal with amino acids abbreviation.