Finding Variables of a C File using Grep

Question

For example let's say i have this code

#include <stdio.h>

int main(void)
{
    int x = 99;
    int *p1, *p2;
    char y[10] = "a"
    // some code
    // some code

    return 0;

}

So in This Code, there are two Variables primary of type int and char . Now, Let's say i make a code of atleast 400 - 500 lines in which i have initialised 20-30 variable. Now, i want to make a bash script to find what are the variables in this code. I begin with this way cat code.c | grep int this would print int main void() Function in the above given scenario and also print int x =99 and //some code statement, But i want to strictly print those lines int x=99; and int *p1, *p2 Not Every line.

So now, here in above question i have a Variable data-type char too, Now How do i construct my Bash script in a Way that after cat i grep only particular data-type like int char double float, Is there any OR Method for it? Or are there any easier way than cat and Grep to perform the same??

So my Final output shall be int x=99; int *p1, *p2 , char y[10] = "a" :)

Would be looking forward to the responses.

Answer 1

You can take a approach with grep to find lines that begin with whitespace followed by int or char using a basic regular expression with something as simple as:

$ grep '^[ ][ ]*\(int\|char\)' yourfile.c
int x = 99;
int *p1, *p2;
char y[10] = "a"

If you have initial tab characters, (or mixed spaces and tabs), you can use a Perl Compatible Regular Expression:

$ grep -P '^[ \t]+(int|char)' yourfile.c
int x = 99;
int *p1, *p2;
char y[10] = "a"

I don't know if this is exactly what you are looking for, and if not, let me know. There are other c-code scanners that will pull variables, functions, etc out and summarize, e.g. cproto is one I have used and liked a lot. There are a number of others on sourceforge you may want to check as well.

Explanation (from comment)

Given the example, it was clear that only int, char, etc.. were wanted AFTER whitespace (e.g. to miss the int main () declaration). With that in mind we set up grep with the basic regular expression to require at least 1 (space using BRE) or if mixed tab and space are possible (using PCRE) before the search term. To handle spaces only, with a BRE:

grep '^[ ][ ]*

or if dealing with mixed spaces and tabs, a PCRE of

grep -P `^[ \t]+

Which anchors the search ^ at the beginning of the line and, with the BRE looks for a space with the character class [ ]. To handle at least one space, and for zero or more additional spaces, we add an additional [ ]*.

Where multiple (or mixed) spaces and tabs are involved (which cannot be handled easily with BRE), the PCRE essentially does the same thing for both space and tab, while sacrificing some portability for the expanded expression handling provided by PCRE. ^[ \t]+ will accommodate one or more instances of space or tab characters before the search terms. The + requiring the presence of at least 1 or more matches of the characters within the character class.

When looking for either int or char, the basic regular expression format is

\(int\|char\)

where ( and | do not retain special meaning in a BRE and must be escaped while the PCRE form is simply:

(int|char)

So putting it altogether it simply anchors a search for one-or-more space (or using PCRE mixed space and tab) characters before either int or char in making the determination of which lines to display.

Hope that helped.