Extend Split to work with multiple separators

24.x
Martin Leitner-Ankerl 3 years ago
parent 12455acca2
commit b7ab9db545
No known key found for this signature in database
GPG Key ID: FBEAAD7FC6FFFE81

@ -2396,6 +2396,19 @@ BOOST_AUTO_TEST_CASE(test_SplitString)
BOOST_CHECK_EQUAL(result.size(), 1);
BOOST_CHECK_EQUAL(result[0], "AAA");
}
// multiple split characters
{
using V = std::vector<std::string>;
BOOST_TEST(SplitString("a,b.c:d;e", ",;") == V({"a", "b.c:d", "e"}));
BOOST_TEST(SplitString("a,b.c:d;e", ",;:.") == V({"a", "b", "c", "d", "e"}));
BOOST_TEST(SplitString("a,b.c:d;e", "") == V({"a,b.c:d;e"}));
BOOST_TEST(SplitString("aaa", "bcdefg") == V({"aaa"}));
BOOST_TEST(SplitString("x\0a,b"s, "\0"s) == V({"x", "a,b"}));
BOOST_TEST(SplitString("x\0a,b"s, '\0') == V({"x", "a,b"}));
BOOST_TEST(SplitString("x\0a,b"s, "\0,"s) == V({"x", "a", "b"}));
BOOST_TEST(SplitString("abcdefg", "bcd") == V({"a", "", "", "efg"}));
}
}
BOOST_AUTO_TEST_CASE(test_LogEscapeMessage)

@ -8,6 +8,7 @@
#include <span.h>
#include <string>
#include <string_view>
#include <vector>
namespace spanparsing {
@ -36,7 +37,7 @@ bool Func(const std::string& str, Span<const char>& sp);
*/
Span<const char> Expr(Span<const char>& sp);
/** Split a string on every instance of sep, returning a vector.
/** Split a string on any char found in separators, returning a vector.
*
* If sep does not occur in sp, a singleton with the entirety of sp is returned.
*
@ -44,13 +45,13 @@ Span<const char> Expr(Span<const char>& sp);
* "foo(bar(1),2),3) on ',' will return {"foo(bar(1)", "2)", "3)"}.
*/
template <typename T = Span<const char>>
std::vector<T> Split(const Span<const char>& sp, char sep)
std::vector<T> Split(const Span<const char>& sp, std::string_view separators)
{
std::vector<T> ret;
auto it = sp.begin();
auto start = it;
while (it != sp.end()) {
if (*it == sep) {
if (separators.find(*it) != std::string::npos) {
ret.emplace_back(start, it);
start = it + 1;
}
@ -60,6 +61,19 @@ std::vector<T> Split(const Span<const char>& sp, char sep)
return ret;
}
/** Split a string on every instance of sep, returning a vector.
*
* If sep does not occur in sp, a singleton with the entirety of sp is returned.
*
* Note that this function does not care about braces, so splitting
* "foo(bar(1),2),3) on ',' will return {"foo(bar(1)", "2)", "3)"}.
*/
template <typename T = Span<const char>>
std::vector<T> Split(const Span<const char>& sp, char sep)
{
return Split<T>(sp, std::string_view{&sep, 1});
}
} // namespace spanparsing
#endif // BITCOIN_UTIL_SPANPARSING_H

@ -14,6 +14,7 @@
#include <locale>
#include <sstream>
#include <string>
#include <string_view>
#include <vector>
[[nodiscard]] inline std::vector<std::string> SplitString(std::string_view str, char sep)
@ -21,6 +22,11 @@
return spanparsing::Split<std::string>(str, sep);
}
[[nodiscard]] inline std::vector<std::string> SplitString(std::string_view str, std::string_view separators)
{
return spanparsing::Split<std::string>(str, separators);
}
[[nodiscard]] inline std::string_view TrimStringView(std::string_view str, std::string_view pattern = " \f\n\r\t\v")
{
std::string::size_type front = str.find_first_not_of(pattern);

Loading…
Cancel
Save