bytestring: Split bytestrings by a set of character delimiters

This introduces the functions bytestring_splitchrs_first() and
bytestring_splitchrs_next() which can be used to iterate through substrings
of a bytestring separated by any of a given set of delimiter characters.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>

ccan/bytestring/bytestring.c

@@ -66,3 +66,41 @@ struct bytestring bytestring_splitchr_next(struct bytestring whole,
 
 	return _splitchr(whole, delim, (prev.ptr - whole.ptr) + prev.len + 1);
 }
+
+static struct bytestring _splitchrs(struct bytestring whole,
+				    struct bytestring delim, size_t start)
+{
+	struct bytestring remainder;
+	size_t n;
+
+	assert(start <= whole.len);
+
+	remainder = bytestring_slice(whole, start, whole.len);
+	n = bytestring_cspn(remainder, delim);
+	return bytestring_slice(whole, start, start + n);
+}
+
+struct bytestring bytestring_splitchrs_first(struct bytestring whole,
+					     struct bytestring delim)
+{
+	if (whole.len == 0)
+		return bytestring_NULL;
+
+	return _splitchrs(whole, delim, 0);
+}
+
+struct bytestring bytestring_splitchrs_next(struct bytestring whole,
+					    struct bytestring delim,
+					    struct bytestring prev)
+{
+	if (!prev.ptr)
+		return bytestring_NULL;
+
+	/* prev has to be a substring of whole */
+	assert(prev.ptr >= whole.ptr);
+
+	if ((prev.ptr + prev.len) == (whole.ptr + whole.len))
+		return bytestring_NULL;
+
+	return _splitchrs(whole, delim, (prev.ptr - whole.ptr) + prev.len + 1);
+}

ccan/bytestring/bytestring.h

@@ -246,4 +246,34 @@ struct bytestring bytestring_splitchr_next(struct bytestring whole,
 	     (_s).ptr;					       \
 	     (_s) = bytestring_splitchr_next((_w), (_delim), (_s)))
 
+/**
+ * bytestring_splitchrs_first - split a bytestring on a set of delimiter
+ *                              characters
+ * @whole: a bytestring
+ * @delim: delimiter characters
+ *
+ * Returns the first substring of @whole delimited by any character in
+ * @delim.
+ */
+struct bytestring bytestring_splitchrs_first(struct bytestring whole,
+					     struct bytestring delim);
+
+/**
+ * bytestring_splitchr_next - split a bytestring on a set of delimiter
+ *                            characters
+ * @whole: a bytestring
+ * @delim: delimiter character
+ * @prev: last substring
+ *
+ * Returns the next @delim delimited substring of @whole after @prev.
+ */
+struct bytestring bytestring_splitchrs_next(struct bytestring whole,
+					    struct bytestring delim,
+					    struct bytestring prev);
+
+#define bytestring_foreach_splitchrs(_s, _w, _delim) \
+	for ((_s) = bytestring_splitchrs_first((_w), (_delim)); \
+	     (_s).ptr;					       \
+	     (_s) = bytestring_splitchrs_next((_w), (_delim), (_s)))
+
 #endif /* CCAN_BYTESTRING_H_ */

ccan/bytestring/test/run.c

@@ -17,7 +17,7 @@ int main(void)
 	int n;
 
 	/* This is how many tests you plan to run */
-	plan_tests(89);
+	plan_tests(109);
 
 	bs = bytestring(str1, sizeof(str1) - 1);
 	ok1(bs.ptr == str1);
@@ -165,6 +165,39 @@ int main(void)
 	}
 	ok1(n == 4);
 
+	bs7 = bytestring_splitchrs_first(bs, BYTESTRING(" \0"));
+	ok1(bs7.ptr == bs.ptr);
+	ok1(bytestring_eq(bs7, BYTESTRING("test")));
+	bs7 = bytestring_splitchrs_next(bs, BYTESTRING(" \0"), bs7);
+	ok1(bs7.ptr == bs.ptr + 5);
+	ok1(bytestring_eq(bs7, BYTESTRING("string")));
+	bs7 = bytestring_splitchrs_next(bs, BYTESTRING(" \0"), bs7);
+	ok1(!bs7.ptr);
+	bs7 = bytestring_splitchrs_next(bs, BYTESTRING(" \0"), bs7);
+	ok1(!bs7.ptr);
+
+	bs7 = bytestring_splitchrs_first(bs2, BYTESTRING(" \0"));
+	ok1(bs7.ptr == bs2.ptr);
+	ok1(bytestring_eq(bs7, BYTESTRING("abc")));
+	bs7 = bytestring_splitchrs_next(bs2, BYTESTRING(" \0"), bs7);
+	ok1(bs7.ptr == bs2.ptr + 4);
+	ok1(bytestring_eq(bs7, BYTESTRING("def")));
+	bs7 = bytestring_splitchrs_next(bs2, BYTESTRING(" \0"), bs7);
+	ok1(!bs7.ptr);
+	bs7 = bytestring_splitchrs_next(bs2, BYTESTRING(" \0"), bs7);
+	ok1(!bs7.ptr);
+
+	bs7 = bytestring_splitchrs_first(BYTESTRING(""), BYTESTRING(" \0"));
+	ok1(!bs7.ptr);
+
+	n = 0;
+	bytestring_foreach_splitchrs(bs7, BYTESTRING(" \0 \0 "),
+				     BYTESTRING("\0 ")) {
+		n++;
+		ok1(bs7.ptr && !bs7.len);
+	}
+	ok1(n == 6);
+
 	/* This exits depending on whether all tests passed */
 	return exit_status();
 }