Fix many calls to ctype functions

All ctype functions like `isspace()`, `tolower()` and alike expect a
character as an `unsigned char` or EOF, anything else has undefined
behavior.

This commit should fix all calls to the various ctype functions either:

* adding `unsigned char` cast where appropriate;
* removing useless `int` casts that are more confusing than useful.

Some instances required slightly more involved changes though, but the
fix itself is the same.

To find the calls to check, I used an equivalent to the following:

```
git grep -P 'is(alnum|alpha|cntrl|x?digit|graph|lower|print|punct|'\
'space|upper|ascii|blank)|to(low|up)er'
```

...and then proceeded to manually check each one of them to see if the
call, or call chain, was using appropriate values and types.

Units/simple-diff.d/input.diff

@@ -106,7 +106,7 @@ index 0000000..a1bf5f3
 +		if (strncmp ((const char*) cp, DiffDelims[delim], 4u) == 0)
 +		{
 +			cp += 4;
-+			if (isspace ((int) *cp)) continue;
++			if (isspace (*cp)) continue;
 +			/* when original filename is /dev/null use the new one instead */
 +			if (delim == DIFF_DELIM_MINUS &&
 +				strncmp ((const char*) cp, "/dev/null", 9u) == 0 &&

dsl/dsl.c

@@ -803,7 +803,7 @@ static EsObject* caseop (EsObject *o, int (*op)(int))
 		char *r = strdup (s);
 
 		for (char *tmp = r; *tmp != '\0'; tmp++)
-			*tmp = op (*tmp);
+			*tmp = op ((unsigned char) *tmp);
 
 		EsObject *q = es_object_autounref (es_string_new (r));
 		free (r);

dsl/optscript.c

@@ -1160,7 +1160,7 @@ name_or_number_new (const char* s, void *data)
 	const char *t = s;
 	while (*t)
 	{
-		if (!isdigit ((int)*t))
+		if (!isdigit ((unsigned char) *t))
 		{
 			number = false;
 			break;

libreadtags/readtags.c

@@ -135,7 +135,7 @@ static int readtags_fseek(FILE *fp, rt_off_t pos, int whence)
 }
 
 /* Converts a hexadecimal digit to its value */
-static int xdigitValue (char digit)
+static int xdigitValue (unsigned char digit)
 {
 	if (digit >= '0' && digit <= '9')
 		return digit - '0';
@@ -151,39 +151,42 @@ static int xdigitValue (char digit)
  * Reads the first character from the string, possibly un-escaping it, and
  * advances *s to the start of the next character.
  */
-static int readTagCharacter (const char **s)
+static int readTagCharacter (const char **const s)
 {
-	int c = **(const unsigned char **)s;
+	const unsigned char *p = (const unsigned char *) *s;
+	int c = *p;
 
-	(*s)++;
+	p++;
 
 	if (c == '\\')
 	{
-		switch (**s)
+		switch (*p)
 		{
-			case 't': c = '\t'; (*s)++; break;
-			case 'r': c = '\r'; (*s)++; break;
-			case 'n': c = '\n'; (*s)++; break;
-			case '\\': c = '\\'; (*s)++; break;
+			case 't': c = '\t'; p++; break;
+			case 'r': c = '\r'; p++; break;
+			case 'n': c = '\n'; p++; break;
+			case '\\': c = '\\'; p++; break;
 			/* Universal-CTags extensions */
-			case 'a': c = '\a'; (*s)++; break;
-			case 'b': c = '\b'; (*s)++; break;
-			case 'v': c = '\v'; (*s)++; break;
-			case 'f': c = '\f'; (*s)++; break;
+			case 'a': c = '\a'; p++; break;
+			case 'b': c = '\b'; p++; break;
+			case 'v': c = '\v'; p++; break;
+			case 'f': c = '\f'; p++; break;
 			case 'x':
-				if (isxdigit ((*s)[1]) && isxdigit ((*s)[2]))
+				if (isxdigit (p[1]) && isxdigit (p[2]))
 				{
-					int val = (xdigitValue ((*s)[1]) << 4) | xdigitValue ((*s)[2]);
+					int val = (xdigitValue (p[1]) << 4) | xdigitValue (p[2]);
 					if (val < 0x80)
 					{
-						(*s) += 3;
+						p += 3;
 						c = val;
 					}
 				}
 				break;
 		}
 	}
 
+	*s = (const char *) p;
+
 	return c;
 }
 
@@ -610,12 +613,12 @@ static tagResult parseTagLine (tagFile *file, tagEntry *const entry, int *err)
 				else
 					++p;
 			}
-			else if (isdigit ((int) *(unsigned char*) p))
+			else if (isdigit (*(unsigned char*) p))
 			{
 				/* parse line number */
 				entry->address.pattern = p;
 				entry->address.lineNumber = atol (p);
-				while (isdigit ((int) *(unsigned char*) p))
+				while (isdigit (*(unsigned char*) p))
 					++p;
 				if (p)
 				{

main/args.c

@@ -31,7 +31,7 @@ static char *nextStringArg (const char** const next)
 	const char* start;
 
 	Assert (*next != NULL);
-	for (start = *next  ;  isspace ((int) *start)  ;  ++start)
+	for (start = *next  ;  isspace ((unsigned char) *start)  ;  ++start)
 		;
 	if (*start == '\0')
 		*next = start;
@@ -40,7 +40,7 @@ static char *nextStringArg (const char** const next)
 		size_t length;
 		const char* end;
 
-		for (end = start ;  *end != '\0'  &&  ! isspace ((int) *end)  ;  ++end)
+		for (end = start; *end != '\0' && ! isspace ((unsigned char) *end); ++end)
 			;
 		length = end - start;
 		Assert (length > 0);
@@ -156,7 +156,7 @@ static char* nextFileLine (FILE* const fp)
 
 static bool isCommentLine (char* line)
 {
-	while (isspace(*line))
+	while (isspace((unsigned char) *line))
 		++line;
 	return (*line == '#');
 }

main/field.c

@@ -878,7 +878,7 @@ static const char* renderCompactInputLine (vString *b,  const char *const line)
 
 				/*  Consume repeating white space.
 				 */
-				while (next = *(p+1) , isspace (next)  &&  next != NEWLINE)
+				while (next = (unsigned char) *(p+1), isspace (next) && next != NEWLINE)
 					++p;
 				c = ' ';  /* force space character for any white space */
 			}
@@ -1328,7 +1328,7 @@ extern int defineField (fieldDefinition *def, langType language)
 	Assert (def->name);
 	for (i = 0; i < strlen (def->name); i++)
 	{
-		Assert ( isalpha (def->name [i]) );
+		Assert ( isalpha ((unsigned char) def->name [i]) );
 	}
 	def->letter = NUL_FIELD_LETTER;
 

main/keyword.c

@@ -74,17 +74,17 @@ static hashEntry *getHashTableEntry (unsigned long hashedValue)
 static unsigned int hashValue (const char *const string, langType language,
 	unsigned int maxLen, bool *maxLenReached)
 {
-	const signed char *p;
+	const char *p;
 	unsigned int h = 5381;
 
 	Assert (string != NULL);
 
 	/* "djb" hash as used in g_str_hash() in glib */
-	for (p = (const signed char *)string; *p != '\0'; p++)
+	for (p = string; *p != '\0'; p++)
 	{
-		h = (h << 5) + h + tolower (*p);
+		h = (h << 5) + h + (signed char) tolower ((unsigned char) *p);
 
-		if (p > (const signed char *)string + maxLen)
+		if (p > string + maxLen)
 		{
 			*maxLenReached = true;
 			return 0;

main/kind.c

@@ -91,7 +91,7 @@ static void initRoleObject (roleObject *robj, roleDefinition *rdef, freeRoleDefF
 #ifdef DEBUG
 	size_t len = strlen (rdef->name);
 	for (int i = 0; i < len; i++)
-		Assert (isalnum (rdef->name [i]));
+		Assert (isalnum ((unsigned char) rdef->name [i]));
 #endif
 	robj->def = rdef;
 	robj->free = freefunc;

main/lregex.c

@@ -824,7 +824,7 @@ static void pre_ptrn_flag_guest_long (const char* const s, const char* const v,
 	{
 		const char *n_tmp = v + 1;
 		const char *n = n_tmp;
-		for (; isdigit (*n_tmp); n_tmp++);
+		for (; isdigit ((unsigned char) *n_tmp); n_tmp++);
 		char c = *n_tmp;
 		*(char *)n_tmp = '\0';
 		if (!strToInt (n, 10, &(guest->lang.spec.patternGroup)))
@@ -886,7 +886,7 @@ static void pre_ptrn_flag_guest_long (const char* const s, const char* const v,
 		{
 			char *n = tmp;
 
-			for (; isdigit (*tmp); tmp++);
+			for (; isdigit ((unsigned char) *tmp); tmp++);
 			char c = *tmp;
 			*tmp = '\0';
 			if (!strToInt (n, 10, &(current->patternGroup)))
@@ -1599,7 +1599,7 @@ static vString* substitute (
 	const char* p;
 	for (p = out  ;  *p != '\0'  ;  p++)
 	{
-		if (*p == '\\'  &&  isdigit ((int) *++p))
+		if (*p == '\\'  &&  isdigit ((unsigned char) *++p))
 		{
 			const int dig = *p - '0';
 			if (0 < dig  &&  dig < nmatch  &&  pmatch [dig].rm_so != -1)
@@ -2294,7 +2294,7 @@ static regexPattern *addTagRegexInternal (struct lregexControlBlock *lcb,
 	{
 		if (p == kindName)
 		{
-			if (!isalpha(*p))
+			if (!isalpha((unsigned char) *p))
 				error (FATAL,
 					   "A kind name doesn't start with an alphabetical character: "
 					   "'%s' in \"--%s-%s\" option",
@@ -2310,7 +2310,7 @@ static regexPattern *addTagRegexInternal (struct lregexControlBlock *lcb,
 			 * in which Exuberant-ctags users define kind names with whitespaces.
 			 * "FATAL" error breaks the compatibility.
 			 */
-			if (!isalnum(*p))
+			if (!isalnum((unsigned char) *p))
 				error (/* regptype == REG_PARSER_SINGLE_LINE? WARNING: */ FATAL,
 					   "Non-alphanumeric char is used in kind name: "
 					   "'%s' in \"--%s-%s\" option",
@@ -2433,7 +2433,7 @@ static void addTagRegexOption (struct lregexControlBlock *lcb,
 		const char *c;
 		for (c = pattern; *c; c++)
 		{
-			if (! (isalnum(*c) || *c == '_'))
+			if (! (isalnum((unsigned char) *c) || *c == '_'))
 			{
 				if (*c &&  (*(c + 1) != '^'))
 				{
@@ -2652,7 +2652,7 @@ extern void addRegexTable (struct lregexControlBlock *lcb, const char *name)
 {
 	const char *c;
 	for (c = name; *c; c++)
-		if (! (isalnum(*c) || *c == '_'))
+		if (! (isalnum((unsigned char) *c) || *c == '_'))
 			error (FATAL, "`%c' in \"%s\" is not acceptable as part of table name", *c, name);
 
 	if (getTableIndexForName(lcb, name) >= 0)

main/options.c

@@ -2604,7 +2604,7 @@ static void processIgnoreOption (const char *const list, int IgnoreOrDefine)
 		addIgnoreListFromFile (lang, fileName);
 	}
 #if defined (WIN32)
-	else if (isalpha (list [0])  &&  list [1] == ':')
+	else if (isalpha ((unsigned char) list [0])  &&  list [1] == ':')
 		addIgnoreListFromFile (lang, list);
 #endif
 	else if (strcmp (list, "-") == 0)