Patch to add the "-iname" option to find

[Bernhard Fischer]

On Thu, Dec 06, 2007 at 10:07:13AM +0100, Alexander Griesser wrote:
>Hi!
>
>I always wondered why the "-iname" option isn't supported by busybox's
>find and as I'm using it heavily I've written a small patch to get it in.
>
>I'm sure that this patch isn't optimal as it simply duplicates the
>action function for the "-name" parameter, but I don't know of any
>other way to do that. The significant change is to add the flag
>FNM_CASEFOLD to the fnmatch function call.
>
>Any objections?
>
>ciao,
>-- 
>Alexander Griesser (Netzwerkadministration)
>E-Mail: alexander.griesser at lkh-vil.or.at | Web: http://www.lkh-vil.or.at
>KABEG LKH Villach | Nikolaigasse 43 | 9500 Villach
>Tel.:   +43 4242 208 3061 | Fax.:   +43 4242 971 3061

>diff -Nur old/findutils/find.c new/findutils/find.c
>--- old/findutils/find.c	2007-12-06 10:04:10.000000000 +0100
>+++ new/findutils/find.c	2007-12-06 10:04:25.000000000 +0100
>@@ -77,6 +77,7 @@
> #define ACTF(name)         static int func_##name(const char *fileName, struct stat *statbuf, action_##name* ap)
>                          ACTS(print)
>                          ACTS(name,  const char *pattern;)
>+                         ACTS(iname,  const char *pattern;)
> USE_FEATURE_FIND_PATH(   ACTS(path,  const char *pattern;))
> USE_FEATURE_FIND_REGEX(  ACTS(regex, regex_t compiled_pattern;))
> USE_FEATURE_FIND_PRINT0( ACTS(print0))
>@@ -190,6 +191,20 @@
> 	}
> 	return fnmatch(ap->pattern, tmp, FNM_PERIOD) == 0;
> }
>+
>+ACTF(iname)
>+{
>+	const char *tmp = bb_basename(fileName);
>+	if (tmp != fileName && !*tmp) {
>+		tmp--;
>+		while (tmp != fileName && *--tmp != '/')
>+			continue;
>+		if (*tmp == '/')
>+			tmp++;
>+	}
>+	return fnmatch(ap->pattern, tmp, FNM_PERIOD | FNM_CASEFOLD) == 0;
>+}
>+
> #if ENABLE_FEATURE_FIND_PATH
> ACTF(path)
> {
>@@ -458,6 +473,7 @@
> 	USE_FEATURE_FIND_PAREN(  PARM_char_brace,)
> 	/* All options starting from here require argument */
> 	                         PARM_name      ,
>+	                         PARM_iname     ,
> 	USE_FEATURE_FIND_PATH(   PARM_path      ,)
> 	USE_FEATURE_FIND_REGEX(  PARM_regex     ,)
> 	USE_FEATURE_FIND_TYPE(   PARM_type      ,)
>@@ -490,6 +506,7 @@
> 	USE_FEATURE_FIND_PAREN(  "(\0"       )
> 	/* All options starting from here require argument */
> 	                         "-name\0"
>+	                         "-iname\0"
> 	USE_FEATURE_FIND_PATH(   "-path\0"   )
> 	USE_FEATURE_FIND_REGEX(  "-regex\0"  )
> 	USE_FEATURE_FIND_TYPE(   "-type\0"   )
>@@ -659,6 +676,11 @@
> 			ap = ALLOC_ACTION(name);
> 			ap->pattern = arg1;
> 		}
>+		else if (parm == PARM_iname) {
>+			action_iname *ap;
>+			ap = ALLOC_ACTION(iname);
>+			ap->pattern = arg1;
>+		}
> #if ENABLE_FEATURE_FIND_PATH
> 		else if (parm == PARM_path) {
> 			action_path *ap;
>diff -Nur old/include/usage.h new/include/usage.h
>--- old/include/usage.h	2007-12-06 10:04:02.000000000 +0100
>+++ new/include/usage.h	2007-12-06 10:04:20.000000000 +0100
>@@ -998,6 +998,7 @@
>      "\n	-maxdepth N	Descend at most N levels. -maxdepth 0 applies" \
>      "\n			tests/actions to command line arguments only") \
>      "\n	-name PATTERN	File name (w/o directory name) matches PATTERN" \
->+     "\n	-iname PATTERN	File name (w/o directory name) matches PATTERN (case insensitive)" \
+>+     "\n	-iname PATTERN	Like -name but case insensitive" \
Is shorter and as easy to understand.
What's the bloatcheck for this change?

Not sure if reusing ACTF(name) and trying to deduce FNM_CASEFOLD somehow
would be smaller or cleaner..