Add character classes for tr

commands/simple/tr.c

@@ -117,11 +117,47 @@ register unsigned char *string1, *string2;
   }
 }
 
+static int starts_with(const char *s1, const char *s2)
+{
+	while (*s1 && *s1 == *s2)
+	{
+		s1++;
+		s2++;
+	}
+	return *s1 == 0;
+}
+
+/* 
+ * character classes from 
+ * http://www.opengroup.org/onlinepubs/009695399/utilities/tr.html
+ * missing: blank, punct, cntrl, graph, print, space 
+ */
+static struct
+{
+	const char *keyword;
+	char first;
+	char last;
+} expand_keywords[] = {
+	{ "[:alnum:]", 'A', 'Z' },
+	{ "[:alnum:]", 'a', 'z' },
+	{ "[:alnum:]", '0', '9' },
+	{ "[:alpha:]", 'A', 'Z' },
+	{ "[:alpha:]", 'a', 'z' },
+	{ "[:digit:]", '0', '9' },
+	{ "[:lower:]", 'a', 'z' },
+	{ "[:upper:]", 'A', 'Z' },
+	{ "[:xdigit:]", '0', '9' },		
+	{ "[:xdigit:]", 'A', 'F' },
+	{ "[:xdigit:]", 'a', 'f' }
+};
+
+#define LENGTH(a) ((sizeof((a))) / (sizeof((a)[0])))
+
 void expand(arg, buffer)
 register char *arg;
 register unsigned char *buffer;
 {
-  int i, ac;
+  int i, ac, keyword_index;
 
   while (*arg) {
 	if (*arg == '\\') {
@@ -136,6 +172,24 @@ register unsigned char *buffer;
 		} else if (*arg != '\0')
 			*buffer++ = *arg++;
 	} else if (*arg == '[') {
+		/* does one of the keywords match? */
+		keyword_index = -1;
+		for (i = 0; i < LENGTH(expand_keywords); i++)
+			if (starts_with(expand_keywords[i].keyword, arg))
+			{
+				/* we have a match, remember and expand */
+				keyword_index = i;
+				ac = expand_keywords[i].first;
+				while (ac <= expand_keywords[i].last)
+					*buffer++ = ac++;
+			}
+
+		/* skip keyword if found, otherwise expand range */
+		if (keyword_index >= 0)
+			arg += strlen(expand_keywords[keyword_index].keyword);
+		else
+		{
+			/* expand range */
 			arg++;
 			i = *arg++;
 			if (*arg++ != '-') {
@@ -146,6 +200,7 @@ register unsigned char *buffer;
 			ac = *arg++;
 			while (i <= ac) *buffer++ = i++;
 			arg++;		/* Skip ']' */
+		}
 	} else
 		*buffer++ = *arg++;
   }

man/man1/tr.1

@@ -45,3 +45,12 @@ Use [ and ] if you want to be portable, because a
 .I tr
 that doesn't need them will still accept the syntax and mindlessly
 translate [ into [ and ] into ].
+.PP
+MINIX tr supports the following character classes: alnum, alpha, digit, lower,
+upper and xdigit. If any of these keywords is encountered between backets and 
+colons, it is replaced by respectively alphanumeric characters, alphabetic
+characters, decimal digits, lowercase letters, uppercase letters and 
+hexadecimal digits. The following are equivalent with the given examples:
+.EX "tr \(fm[:upper:]\(fm \(fm[:lower:]\(fm <x >y     " "Convert upper case to lower case"
+.EX "tr \-d \(fm[:digit:]\(fm <f1 >f2  " "Delete all digits from \fIf1\fR"
+