beenull/ladybird
1
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird.git synced 2024-11-21 23:20:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 6

AK: Add Utf8View::for_each_split_view() method

Browse source 
Returns one Utf8View at a time, using a callback function to identify
code points to split on.
This commit is contained in:
Sam Atkins 2024-11-15 15:43:59 +00:00 committed by Andreas Kling
parent ec5101a1d3
commit 3f10a5701d
Notes: github-actions[bot] 2024-11-15 22:27:09 +00:00 
Author: https://github.com/AtkinsSJ
Commit: https://github.com/LadybirdBrowser/ladybird/commit/3f10a5701d9
Pull-request: https://github.com/LadybirdBrowser/ladybird/pull/2368
2 changed files with 70 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

47
AK/Utf8View.h
View file

@ -10,6 +10,7 @@
#include <AK/ByteString.h> #include <AK/ByteString.h>
#include <AK/Debug.h> #include <AK/Debug.h>
#include <AK/Format.h> #include <AK/Format.h>
#include <AK/Function.h>
#include <AK/StringView.h> #include <AK/StringView.h>
#include <AK/Types.h> #include <AK/Types.h>
@ -139,6 +140,52 @@ public:
bool validate(size_t& valid_bytes, AllowSurrogates allow_surrogates = AllowSurrogates::Yes) const; bool validate(size_t& valid_bytes, AllowSurrogates allow_surrogates = AllowSurrogates::Yes) const;
template<typename Callback>
auto for_each_split_view(Function<bool(u32)> splitter, SplitBehavior split_behavior, Callback callback) const
{
bool keep_empty = has_flag(split_behavior, SplitBehavior::KeepEmpty);
bool keep_trailing_separator = has_flag(split_behavior, SplitBehavior::KeepTrailingSeparator);
auto start_offset = 0u;
auto offset = 0u;
auto run_callback = [&]() {
auto length = offset - start_offset;
if (length == 0 && !keep_empty)
return;
auto substring = unicode_substring_view(start_offset, length);
// Reject splitter-only entries if we're not keeping empty results
if (keep_trailing_separator && !keep_empty && length == 1 && splitter(*substring.begin()))
return;
callback(substring);
};
auto iterator = begin();
while (iterator != end()) {
if (splitter(*iterator)) {
if (keep_trailing_separator)
++offset;
run_callback();
if (!keep_trailing_separator)
++offset;
start_offset = offset;
++iterator;
continue;
}
++offset;
++iterator;
}
run_callback();
}
private: private:
friend class Utf8CodePointIterator; friend class Utf8CodePointIterator;

23
Tests/AK/TestUtf8.cpp
View file

@ -312,3 +312,26 @@ TEST_CASE(trim)
EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), "\u180E"); EXPECT_EQ(view.trim(whitespace, TrimMode::Right).as_string(), "\u180E");
} }
} }
static bool is_period(u32 code_point) { return code_point == '.'; }
TEST_CASE(for_each_split_view)
{
Utf8View view { "...Well..hello.friends!..."sv };
auto gather = [&](auto split_behavior) {
Vector<StringView> results;
view.for_each_split_view(is_period, split_behavior, [&](auto part) {
results.append(part.as_string());
});
return results;
};
EXPECT_EQ(gather(SplitBehavior::Nothing),
Vector({ "Well"sv, "hello"sv, "friends!"sv }));
EXPECT_EQ(gather(SplitBehavior::KeepEmpty),
Vector({ ""sv, ""sv, ""sv, "Well"sv, ""sv, "hello"sv, "friends!"sv, ""sv, ""sv, ""sv }));
EXPECT_EQ(gather(SplitBehavior::KeepTrailingSeparator),
Vector({ "Well."sv, "hello."sv, "friends!."sv }));
EXPECT_EQ(gather(SplitBehavior::KeepEmpty | SplitBehavior::KeepTrailingSeparator),
Vector({ "."sv, "."sv, "."sv, "Well."sv, "."sv, "hello."sv, "friends!."sv, "."sv, "."sv, ""sv }));
}