add utf8_iterator for adapting byte iterators to iterate on unicode characters

1 files changed, 140 insertions, 0 deletions

diff --git a/src/utf8_iterator.hh b/src/utf8_iterator.hh
new file mode 100644
index 00000000..e782e297
--- /dev/null
+++ b/src/utf8_iterator.hh
@@ -0,0 +1,140 @@
+#ifndef utf8_iterator_hh_INCLUDED
+#define utf8_iterator_hh_INCLUDED
+
+#include "utf8.hh"
+
+namespace Kakoune
+{
+
+namespace utf8
+{
+
+// adapter for an iterator on bytes which permits to iterate
+// on unicode codepoints instead.
+template<typename Iterator>
+class utf8_iterator
+{
+public:
+    utf8_iterator() = default;
+    utf8_iterator(Iterator it) : m_it(std::move(it)) {}
+
+    utf8_iterator& operator++()
+    {
+        m_it = utf8::next(m_it);
+        invalidate_value();
+        return *this;
+    }
+
+    utf8_iterator operator++(int)
+    {
+        utf8_iterator save = *this;
+        ++*this;
+        return save;
+    }
+
+    utf8_iterator& operator--()
+    {
+        m_it = utf8::previous(m_it);
+        invalidate_value();
+        return *this;
+    }
+
+    utf8_iterator operator--(int)
+    {
+        utf8_iterator save = *this;
+        --*this;
+        return save;
+    }
+
+    utf8_iterator operator+(int count) const
+    {
+        if (count < 0)
+           return operator-(-count);
+
+        utf8_iterator res = *this;
+        while (count--)
+            ++res;
+        return res;
+    }
+
+    utf8_iterator operator-(int count) const
+    {
+        if (count < 0)
+           return operator+(-count);
+
+        utf8_iterator res = *this;
+        while (count--)
+            --res;
+        return res;
+    }
+
+    bool operator==(const utf8_iterator& other) { return m_it == other.m_it; }
+    bool operator!=(const utf8_iterator& other) { return m_it != other.m_it; }
+
+    bool operator< (const utf8_iterator& other) const
+    {
+        return m_it < other.m_it;
+    }
+
+    bool operator<= (const utf8_iterator& other) const
+    {
+        return m_it <= other.m_it;
+    }
+
+    bool operator> (const utf8_iterator& other) const
+    {
+        return m_it > other.m_it;
+    }
+
+    bool operator>= (const utf8_iterator& other) const
+    {
+        return m_it >= other.m_it;
+    }
+
+    size_t operator-(utf8_iterator other) const
+    {
+        //assert(other < *this);
+        check_invariant();
+        other.check_invariant();
+        size_t dist = 0;
+        while (other.m_it != m_it)
+        {
+            ++dist;
+            ++other;
+        }
+        return dist;
+    }
+
+    Codepoint operator*() const
+    {
+        return get_value();
+    }
+
+    const Iterator& underlying_iterator() const { return m_it; }
+    Iterator& underlying_iterator() { return m_it; }
+
+protected:
+    void check_invariant() const
+    {
+        // always point to a character first byte;
+        // assert(is_character_start(it));
+    }
+
+private:
+    void invalidate_value() { m_value = -1; }
+    Codepoint get_value() const
+    {
+        if (m_value == -1)
+            m_value = utf8::codepoint(m_it);
+        return m_value;
+    }
+
+    Iterator m_it;
+    mutable Codepoint m_value = -1;
+};
+
+}
+
+}
+
+#endif // utf8_iterator_hh_INCLUDED