[ibus-typing-booster] libm17n-direct: WIP: use libm17n directly (c432510)
by mfabian@fedoraproject.org
Repository : http://git.fedorahosted.org/git/?p=ibus-typing-booster.git
On branch : libm17n-direct
>---------------------------------------------------------------
commit c432510ed4d07ecde109d9babcb2a20a2e4f3a8e
Author: Mike FABIAN <mfabian(a)redhat.com>
Date: Mon Dec 7 23:57:11 2015 +0100
WIP: use libm17n directly
>---------------------------------------------------------------
ibus-typing-booster/engine/m17n_translit.py | 189 +++++++++++++++++++++++++++
1 files changed, 189 insertions(+), 0 deletions(-)
diff --git a/ibus-typing-booster/engine/m17n_translit.py b/ibus-typing-booster/engine/m17n_translit.py
new file mode 100644
index 0000000..a40b9c5
--- /dev/null
+++ b/ibus-typing-booster/engine/m17n_translit.py
@@ -0,0 +1,189 @@
+import ctypes
+
+class libm17n__MSymbolStruct(ctypes.Structure):
+ pass
+libm17n__MSymbol = ctypes.POINTER(libm17n__MSymbolStruct)
+class libm17n__MPlist(ctypes.Structure):
+ pass
+class libm17n__MConverter(ctypes.Structure):
+ pass
+class libm17n__MInputMethod(ctypes.Structure): # m17n.h
+ pass
+class libm17n__MInputContext(ctypes.Structure): # m17n.h
+ pass
+class libm17n__MText(ctypes.Structure): # m17n-core.h
+ pass
+libm17n__MSymbolStruct._fields_ = [('managing_key', ctypes.c_uint),
+ ('name', ctypes.c_char_p),
+ ('length', ctypes.c_int),
+ ('plist', libm17n__MPlist),
+ ('next', ctypes.POINTER(libm17n__MSymbolStruct))]
+
+libm17n__lib = None
+libm17n__msymbol = None
+libm17n__mplist = None
+libm17n__mconv_buffer_converter = None
+libm17n__mconv_reset_converter = None
+libm17n__mconv_rebind_buffer = None
+libm17n__mconv_encode = None
+libm17n__minput_open_im = None
+libm17n__minput_create_ic = None
+libm17n__minput_reset_ic = None
+libm17n__minput_filter = None
+libm17n__minput_lookup = None
+libm17n__mtext = None
+libm17n__mtext_len = None
+libm17n__Mcoding_utf_8 = None
+
+_utf8_converter = None
+
+def mtext_to_string(_mt):
+ libm17n__mconv_reset_converter(_utf8_converter)
+ # one Unicode character cannot have more than 6 UTF-8 bytes
+ # (actually not more than 4 ...)
+ buf = bytes(libm17n__mtext_len(_mt) * 6)
+ libm17n__mconv_rebind_buffer(
+ _utf8_converter, ctypes.c_char_p(buf), ctypes.c_int(10))
+ libm17n__mconv_encode(_utf8_converter, _mt)
+ buf = buf[0:buf.find(b'\x00')]
+ return buf.decode('utf-8')
+
+def _init():
+ global libm17n__lib
+ libm17n__lib = ctypes.CDLL('libm17n.so.0', mode = ctypes.RTLD_GLOBAL)
+ libm17n__lib.m17n_init()
+ global libm17n__mplist
+ libm17n__mplist = libm17n__lib.mplist
+ libm17n__mplist.argtypes = []
+ libm17n__mplist.restype = ctypes.POINTER(libm17n__MPlist)
+ global libm17n__mconv_buffer_converter
+ libm17n__mconv_buffer_converter = libm17n__lib.mconv_buffer_converter
+ libm17n__mconv_buffer_converter.argtypes = [
+ libm17n__MSymbol, ctypes.c_char_p, ctypes.c_int]
+ libm17n__mconv_buffer_converter.restype = ctypes.POINTER(
+ libm17n__MConverter)
+ global libm17n__mconv_reset_converter
+ libm17n__mconv_reset_converter = libm17n__lib.mconv_reset_converter
+ libm17n__mconv_reset_converter.argtypes = [
+ ctypes.POINTER(libm17n__MConverter)]
+ libm17n__mconv_reset_converter.restype = ctypes.c_int
+ global libm17n__mconv_rebind_buffer
+ libm17n__mconv_rebind_buffer = libm17n__lib.mconv_rebind_buffer
+ libm17n__mconv_rebind_buffer.argtypes = [
+ ctypes.POINTER(libm17n__MConverter), ctypes.c_char_p, ctypes.c_int]
+ libm17n__mconv_rebind_buffer.restype = ctypes.POINTER(libm17n__MConverter)
+ global libm17n__mconv_encode
+ libm17n__mconv_encode = libm17n__lib.mconv_encode
+ libm17n__mconv_encode.argtypes = [
+ ctypes.POINTER(libm17n__MConverter), ctypes.POINTER(libm17n__MText)]
+ libm17n__mconv_encode.restype = ctypes.c_int
+ global libm17n__msymbol
+ libm17n__msymbol = libm17n__lib.msymbol
+ libm17n__msymbol.argtypes = [ctypes.c_char_p]
+ libm17n__msymbol.restype = libm17n__MSymbol
+ global libm17n__minput_open_im
+ libm17n__minput_open_im = libm17n__lib.minput_open_im
+ libm17n__minput_open_im.argtypes = [
+ libm17n__MSymbol, libm17n__MSymbol, ctypes.c_void_p]
+ libm17n__minput_open_im.restype = ctypes.POINTER(libm17n__MInputMethod)
+ global libm17n__minput_create_ic
+ libm17n__minput_create_ic = libm17n__lib.minput_create_ic
+ libm17n__minput_create_ic.argtypes = [
+ ctypes.POINTER(libm17n__MInputMethod), ctypes.c_void_p]
+ libm17n__minput_create_ic.restype = ctypes.POINTER(libm17n__MInputContext)
+ global libm17n__minput_reset_ic
+ libm17n__minput_reset_ic = libm17n__lib.minput_reset_ic
+ libm17n__minput_reset_ic.argtypes = [ctypes.POINTER(libm17n__MInputContext)]
+ global libm17n__minput_filter
+ libm17n__minput_filter = libm17n__lib.minput_filter
+ libm17n__minput_filter.argtypes = [
+ ctypes.POINTER(libm17n__MInputContext),
+ libm17n__MSymbol,
+ ctypes.c_void_p]
+ libm17n__minput_filter.restype = ctypes.c_int
+ global libm17n__minput_lookup
+ libm17n__minput_lookup = libm17n__lib.minput_lookup
+ libm17n__minput_lookup.argtypes = [
+ ctypes.POINTER(libm17n__MInputContext),
+ libm17n__MSymbol,
+ ctypes.c_void_p,
+ ctypes.POINTER(libm17n__MText)]
+ libm17n__minput_lookup.restype = ctypes.c_int
+ global libm17n__mtext
+ libm17n__mtext = libm17n__lib.mtext
+ libm17n__mtext.argtypes = []
+ libm17n__mtext.restype = ctypes.POINTER(libm17n__MText)
+ global libm17n__mtext_len
+ libm17n__mtext_len = libm17n__lib.mtext_len
+ libm17n__mtext_len.argtypes = [ctypes.POINTER(libm17n__MText)]
+ libm17n__mtext_len.restype = ctypes.c_int
+ global libm17n__Mcoding_utf_8
+ libm17n__Mcoding_utf_8 = libm17n__MSymbol.in_dll(
+ ctypes.pythonapi, 'Mcoding_utf_8')
+ global _utf8_converter
+ _utf8_converter = libm17n__mconv_buffer_converter(
+ libm17n__Mcoding_utf_8, ctypes.c_char_p(None), ctypes.c_int(0))
+
+def _del():
+ libm17n__lib.m17n_fini()
+
+class __ModuleInitializer:
+ def __init__(self):
+ _init()
+ return
+
+ def __del__(self):
+ return
+
+__module_init = __ModuleInitializer()
+
+
+class Transliterator:
+ def __init__(self, ime):
+ self._language = ime.split('-')[0]
+ self._name = '-'.join(ime.split('-')[1:])
+ self._im = libm17n__minput_open_im(
+ libm17n__msymbol(ctypes.c_char_p(self._language.encode('utf-8'))),
+ libm17n__msymbol(ctypes.c_char_p(self._name.encode('utf-8'))),
+ ctypes.c_void_p(None))
+ self._ic = libm17n__minput_create_ic(self._im, ctypes.c_void_p(None))
+
+ def transliterate(self, msymbol_list):
+ libm17n__minput_reset_ic(self._ic)
+ output = ''
+ msymbol_list.append('nil')
+ for symbol in msymbol_list:
+ _symbol = libm17n__msymbol(symbol.encode('utf-8'))
+ retval = libm17n__minput_filter(
+ self._ic, _symbol, ctypes.c_void_p(None))
+ if retval == 0:
+ _mt = libm17n__mtext()
+ retval = libm17n__minput_lookup(
+ self._ic, _symbol, ctypes.c_void_p(None), _mt)
+ if libm17n__mtext_len(_mt) > 0:
+ output += mtext_to_string(_mt)
+ if retval and symbol != 'nil':
+ output += symbol
+ return output
+
+if __name__ == "__main__":
+ trans_hi_inscript2 = Transliterator('hi-inscript2')
+ result = trans_hi_inscript2.transliterate(['4', 'G-4'])
+ print(result)
+ print(bytes(result, encoding = 'utf-8'))
+ trans_mr_itrans = Transliterator('mr-itrans')
+ result = trans_mr_itrans.transliterate(list('praviN'))
+ print(result)
+ print(bytes(result, encoding = 'utf-8'))
+ trans_ru_translit = Transliterator('ru-translit')
+ result = trans_ru_translit.transliterate(list('yo'))
+ print(result)
+ print(bytes(result, encoding = 'utf-8'))
+ result = trans_ru_translit.transliterate(list('y'))
+ print(result)
+ print(bytes(result, encoding = 'utf-8'))
+ result = trans_ru_translit.transliterate(list('yo y'))
+ print(result)
+ print(bytes(result, encoding = 'utf-8'))
+ import doctest
+ doctest.testmod()
8 years, 5 months