Updated version 3.0 from 'upstream/3.0'

with Debian dir 0f18938e98f5a466f36719f60cef0490163ab845

1 files changed, 43 insertions, 0 deletions

diff --git a/louloulibs/utils/encoding.hpp b/louloulibs/utils/encoding.hpp
new file mode 100644
index 0000000..586edd8
--- /dev/null
+++ b/louloulibs/utils/encoding.hpp
@@ -0,0 +1,43 @@
+#pragma once
+
+
+#include <string>
+
+namespace utils
+{
+  /**
+   * Return the size, in bytes, of the next UTF-8 codepoint, based on
+   * the given char.
+   */
+  std::size_t get_next_codepoint_size(const unsigned char c);
+  /**
+   * Returns true if the given null-terminated string is valid utf-8.
+   *
+   * Based on http://en.wikipedia.org/wiki/UTF-8#Description
+   */
+  bool is_valid_utf8(const char* s);
+  /**
+   * Remove all invalid codepoints from the given utf-8-encoded string.
+   * The value returned is a copy of the string, without the removed chars.
+   *
+   * See http://www.w3.org/TR/xml/#charsets for the list of valid characters
+   * in XML.
+   */
+  std::string remove_invalid_xml_chars(const std::string& original);
+  /**
+   * Convert the given string (encoded is "encoding") into valid utf-8.
+   * If some decoding fails, insert an utf-8 placeholder character instead.
+   */
+  std::string convert_to_utf8(const std::string& str, const char* encoding);
+}
+
+namespace xep0106
+{
+  /**
+   * Decode and encode inplace.
+   */
+  void decode(std::string&);
+  void encode(std::string&);
+}
+
+