Add support for switch_lang for Korean. (#981)

* switch korean letters to english

* revised according to recent changes

* Fix typo in tests/test_switch_lang.py

* Add a test case for coverage

* Change: Moved decomposing logic which changes command.script to get_new_command instead of match.

* Fix: changed unicode characters to unicode string for python2 compatibility.

* Fix: Modified to change request.

@ik1ne @yangkyeongmo

tests/rules/test_switch_lang.py

@@ -1,6 +1,7 @@
 # -*- encoding: utf-8 -*-
 
 import pytest
+
 from thefuck.rules import switch_lang
 from thefuck.types import Command
 
@@ -9,7 +10,8 @@ from thefuck.types import Command
     Command(u'фзе-пуе', 'command not found: фзе-пуе'),
     Command(u'λσ', 'command not found: λσ'),
     Command(u'שפא-עקא', 'command not found: שפא-עקא'),
-    Command(u'ךד', 'command not found: ךד')])
+    Command(u'ךד', 'command not found: ךד'),
+    Command(u'녀애 ㅣㄴ', 'command not found: 녀애 ㅣㄴ')])
 def test_match(command):
     assert switch_lang.match(command)
 
@@ -19,7 +21,8 @@ def test_match(command):
     Command(u'ls', 'command not found: ls'),
     Command(u'агсл', 'command not found: агсл'),
     Command(u'фзе-пуе', 'some info'),
-    Command(u'שפא-עקא', 'some info')])
+    Command(u'שפא-עקא', 'some info'),
+    Command(u'녀애 ㅣㄴ', 'some info')])
 def test_not_match(command):
     assert not switch_lang.match(command)
 
@@ -28,6 +31,9 @@ def test_not_match(command):
     (Command(u'фзе-пуе штыефдд мшь', ''), 'apt-get install vim'),
     (Command(u'λσ -λα', ''), 'ls -la'),
     (Command(u'שפא-עקא ןמדאשךך הןצ', ''), 'apt-get install vim'),
-    (Command(u'ךד -ךש', ''), 'ls -la')])
+    (Command(u'ךד -ךש', ''), 'ls -la'),
+    (Command(u'멧-ㅎㄷㅅ ㅑㅜㄴㅅ미ㅣ 퍄ㅡ', ''), 'apt-get install vim'),
+    (Command(u'ㅣㄴ -ㅣㅁ', ''), 'ls -la'),
+    (Command(u'ㅔㅁㅅ촤', ''), 'patchk'), ])
 def test_get_new_command(command, new_command):
     assert switch_lang.get_new_command(command) == new_command

thefuck/rules/switch_lang.py

@@ -5,13 +5,14 @@ target_layout = '''qwertyuiop[]asdfghjkl;'zxcvbnm,./QWERTYUIOP{}ASDFGHJKL:"ZXCVB
 # any new keyboard layout must be appended
 
 greek = u''';ςερτυθιοπ[]ασδφγηξκλ΄ζχψωβνμ,./:΅ΕΡΤΥΘΙΟΠ{}ΑΣΔΦΓΗΞΚΛ¨"ΖΧΨΩΒΝΜ<>?'''
+korean = u'''ㅂㅈㄷㄱㅅㅛㅕㅑㅐㅔ[]ㅁㄴㅇㄹㅎㅗㅓㅏㅣ;'ㅋㅌㅊㅍㅠㅜㅡ,./ㅃㅉㄸㄲㅆㅛㅕㅑㅒㅖ{}ㅁㄴㅇㄹㅎㅗㅓㅏㅣ:"ㅋㅌㅊㅍㅠㅜㅡ<>?'''
 
 source_layouts = [u'''йцукенгшщзхъфывапролджэячсмитьбю.ЙЦУКЕНГШЩЗХЪФЫВАПРОЛДЖЭЯЧСМИТЬБЮ,''',
                   u'''йцукенгшщзхїфівапролджєячсмитьбю.ЙЦУКЕНГШЩЗХЇФІВАПРОЛДЖЄЯЧСМИТЬБЮ,''',
                   u'''ضصثقفغعهخحجچشسیبلاتنمکگظطزرذدپو./ًٌٍَُِّْ][}{ؤئيإأآة»«:؛كٓژٰ‌ٔء><؟''',
                   u'''/'קראטוןםפ][שדגכעיחלךף,זסבהנמצתץ.QWERTYUIOP{}ASDFGHJKL:"ZXCVBNM<>?''',
-                  greek]
+                  greek,
-
+                  korean]
 
 source_to_target = {
     greek: {u';': "q", u'ς': "w", u'ε': "e", u'ρ': "r", u'τ': "t", u'υ': "y",
@@ -30,6 +31,19 @@ source_to_target = {
             u'Ώ': "V"},
 }
 
+'''Lists used for decomposing korean letters.'''
+HEAD_LIST = [u'ㄱ', u'ㄲ', u'ㄴ', u'ㄷ', u'ㄸ', u'ㄹ', u'ㅁ', u'ㅂ', u'ㅃ', u'ㅅ', u'ㅆ',
+             u'ㅇ', u'ㅈ', u'ㅉ', u'ㅊ', u'ㅋ', u'ㅌ', u'ㅍ', u'ㅎ']
+BODY_LIST = [u'ㅏ', u'ㅐ', u'ㅑ', u'ㅒ', u'ㅓ', u'ㅔ', u'ㅕ', u'ㅖ', u'ㅗ', u'ㅘ', u'ㅙ',
+             u'ㅚ', u'ㅛ', u'ㅜ', u'ㅝ', u'ㅞ', u'ㅟ', u'ㅠ', u'ㅡ', u'ㅢ', u'ㅣ']
+TAIL_LIST = [u' ', u'ㄱ', u'ㄲ', u'ㄳ', u'ㄴ', u'ㄵ', u'ㄶ', u'ㄷ', u'ㄹ', u'ㄺ', u'ㄻ',
+             u'ㄼ', u'ㄽ', u'ㄾ', u'ㄿ', u'ㅀ', u'ㅁ', u'ㅂ', u'ㅄ', u'ㅅ', u'ㅆ', u'ㅇ', u'ㅈ',
+             u'ㅊ', u'ㅋ', u'ㅌ', u'ㅍ', u'ㅎ']
+DOUBLE_LIST = [u'ㅘ', u'ㅙ', u'ㅚ', u'ㅝ', u'ㅞ', u'ㅟ', u'ㅢ', u'ㄳ', u'ㄵ', u'ㄶ', u'ㄺ',
+               u'ㄻ', u'ㄼ', u'ㄽ', u'ㄾ', u'ㅀ', u'ㅄ']
+DOUBLE_MOD_LIST = [u'ㅗㅏ', u'ㅗㅐ', u'ㅗㅣ', u'ㅜㅓ', u'ㅜㅔ', u'ㅜㅣ', u'ㅡㅣ', u'ㄱㅅ',
+                   u'ㄴㅈ', u'ㄴㅎ', u'ㄹㄱ', u'ㄹㅁ', u'ㄹㅂ', u'ㄹㅅ', u'ㄹㅌ', u'ㄹㅎ', u'ㅂㅅ']
+
 
 @memoize
 def _get_matched_layout(command):
@@ -50,7 +64,6 @@ def _get_matched_layout(command):
 def _switch(ch, layout):
     if ch in layout:
         return target_layout[layout.index(ch)]
-    else:
     return ch
 
 
@@ -63,9 +76,33 @@ def _switch_command(command, layout):
     return ''.join(_switch(ch, layout) for ch in command.script)
 
 
+def _decompose_korean(command):
+    def _change_double(ch):
+        if ch in DOUBLE_LIST:
+            return DOUBLE_MOD_LIST[DOUBLE_LIST.index(ch)]
+        return ch
+
+    hg_str = u''
+    for ch in command.script:
+        if u'가' <= ch <= u'힣':
+            ord_ch = ord(ch) - ord(u'가')
+            hd = ord_ch // 588
+            bd = (ord_ch - 588 * hd) // 28
+            tl = ord_ch - 588 * hd - 28 * bd
+            for ch in [HEAD_LIST[hd], BODY_LIST[bd], TAIL_LIST[tl]]:
+                if ch != ' ':
+                    hg_str += _change_double(ch)
+        else:
+            hg_str += _change_double(ch)
+    return hg_str
+
+
 def match(command):
     if 'not found' not in command.output:
         return False
+    if any(u'ㄱ' <= ch <= u'ㅎ' or u'ㅏ' <= ch <= u'ㅣ' or u'가' <= ch <= u'힣'
+            for ch in command.script):
+        return True
 
     matched_layout = _get_matched_layout(command)
     return (matched_layout and
@@ -73,5 +110,8 @@ def match(command):
 
 
 def get_new_command(command):
+    if any(u'ㄱ' <= ch <= u'ㅎ' or u'ㅏ' <= ch <= u'ㅣ' or u'가' <= ch <= u'힣'
+            for ch in command.script):
+        command.script = _decompose_korean(command)
     matched_layout = _get_matched_layout(command)
     return _switch_command(command, matched_layout)