diff options
Diffstat (limited to 'src/mnemonics')
| -rw-r--r-- | src/mnemonics/language_base.h | 74 |
1 files changed, 4 insertions, 70 deletions
diff --git a/src/mnemonics/language_base.h b/src/mnemonics/language_base.h index 7d2599e9a..ad09dc5fa 100644 --- a/src/mnemonics/language_base.h +++ b/src/mnemonics/language_base.h @@ -41,6 +41,7 @@ #include <boost/algorithm/string.hpp>
#include "misc_log_ex.h"
#include "fnv1.h"
+#include "common/utf8.h"
/*!
* \namespace Language
@@ -73,78 +74,11 @@ namespace Language return prefix;
}
- template<typename T>
- inline T utf8canonical(const T &s)
- {
- T sc = "";
- size_t avail = s.size();
- const char *ptr = s.data();
- wint_t cp = 0;
- int bytes = 1;
- char wbuf[8], *wptr;
- while (avail--)
- {
- if ((*ptr & 0x80) == 0)
- {
- cp = *ptr++;
- bytes = 1;
- }
- else if ((*ptr & 0xe0) == 0xc0)
- {
- if (avail < 1)
- throw std::runtime_error("Invalid UTF-8");
- cp = (*ptr++ & 0x1f) << 6;
- cp |= *ptr++ & 0x3f;
- --avail;
- bytes = 2;
- }
- else if ((*ptr & 0xf0) == 0xe0)
- {
- if (avail < 2)
- throw std::runtime_error("Invalid UTF-8");
- cp = (*ptr++ & 0xf) << 12;
- cp |= (*ptr++ & 0x3f) << 6;
- cp |= *ptr++ & 0x3f;
- avail -= 2;
- bytes = 3;
- }
- else if ((*ptr & 0xf8) == 0xf0)
- {
- if (avail < 3)
- throw std::runtime_error("Invalid UTF-8");
- cp = (*ptr++ & 0x7) << 18;
- cp |= (*ptr++ & 0x3f) << 12;
- cp |= (*ptr++ & 0x3f) << 6;
- cp |= *ptr++ & 0x3f;
- avail -= 3;
- bytes = 4;
- }
- else
- throw std::runtime_error("Invalid UTF-8");
-
- cp = std::towlower(cp);
- wptr = wbuf;
- switch (bytes)
- {
- case 1: *wptr++ = cp; break;
- case 2: *wptr++ = 0xc0 | (cp >> 6); *wptr++ = 0x80 | (cp & 0x3f); break;
- case 3: *wptr++ = 0xe0 | (cp >> 12); *wptr++ = 0x80 | ((cp >> 6) & 0x3f); *wptr++ = 0x80 | (cp & 0x3f); break;
- case 4: *wptr++ = 0xf0 | (cp >> 18); *wptr++ = 0x80 | ((cp >> 12) & 0x3f); *wptr++ = 0x80 | ((cp >> 6) & 0x3f); *wptr++ = 0x80 | (cp & 0x3f); break;
- default: throw std::runtime_error("Invalid UTF-8");
- }
- *wptr = 0;
- sc += T(wbuf, bytes);
- cp = 0;
- bytes = 1;
- }
- return sc;
- }
-
struct WordHash
{
std::size_t operator()(const epee::wipeable_string &s) const
{
- const epee::wipeable_string sc = utf8canonical(s);
+ const epee::wipeable_string sc = tools::utf8canonical(s, [](wint_t c) -> wint_t { return std::towlower(c); });
return epee::fnv::FNV1a(sc.data(), sc.size());
}
};
@@ -153,8 +87,8 @@ namespace Language {
bool operator()(const epee::wipeable_string &s0, const epee::wipeable_string &s1) const
{
- const epee::wipeable_string s0c = utf8canonical(s0);
- const epee::wipeable_string s1c = utf8canonical(s1);
+ const epee::wipeable_string s0c = tools::utf8canonical(s0, [](wint_t c) -> wint_t { return std::towlower(c); });
+ const epee::wipeable_string s1c = tools::utf8canonical(s1, [](wint_t c) -> wint_t { return std::towlower(c); });
return s0c == s1c;
}
};
|