Software Público Brasileiro

#-----------------------------------------------------------------------------
# Name:        stcspellcheck.py
# Purpose:     Spell checking for the wx.StyledTextControl using pyenchant
#
# Author:      Rob McMullen
#
# Created:     2008
# RCS-ID:      $Id: stcspellcheck.py 69791 2011-11-19 22:02:59Z CJP $
# Copyright:   (c) 2008 Rob McMullen
# License:     wxWidgets
#-----------------------------------------------------------------------------
#
# Originally based on code from Luke-SDK, which includes the following
# copyright notice:
# 
# Copyright (c) 2007 Eur Ing Christopher Thoday, cthoday@mail.berlios.de.
#
# Permission to use, copy, modify and distribute this software and its
# documentation for any purpose and without fee is hereby granted, provided
# that the above copyright notice appear in all copies and that both
# the copyright notice and this permission notice appear in supporting
# documentation.
#
# Christopher Thoday makes no representations about the suitability of this
# software for any purpose.  It is provided "as is" without express or implied
# warranty.

"""Spell checking for the wx.StyledTextControl using pyenchant

This module was insipred by the spell check function from Christopher Thoday's
U{Luke SDK<http://luke-sdk.berlios.de/>}.

Spell checking is provided by the pyenchant library, which is an external
dependency not part of wxPython.  Packages are available for Mac, Unix, and
windows at U{http://pyenchant.sourceforge.net}

Currently provides:
 - spell checking of entire buffer, currently visible page, or selection region
 - user specified indicator number (0 - 2), style, and color
 - language can be changed on the fly
 - update the spelling as you type
 - check the document in either idle time or in a background thread

@author: Rob McMullen
@version: 1.2

Changelog::
    1.2:
        - Rewrote as a standalone class rather than a static mixin
    1.1:
        - Added helper function to use idle processing time to check document
        - Added word checking function for use in instant spell checking
    1.0:
        - First public release
"""

import os
import locale
import wx
import wx.stc

# Assume MacPorts install of Enchant
if wx.Platform == '__WXMAC__':
    if 'PYENCHANT_LIBRARY_PATH' not in os.environ:
        os.environ['PYENCHANT_LIBRARY_PATH'] = '/opt/local/lib/libenchant.dylib'

try:
    import enchant
except ImportError:
    # no big deal; support for enchant simply won't be included
    pass
except:
    # big deal; enchant is there but there's some error that is preventing
    # its import
    import traceback
    traceback.print_exc()

class STCSpellCheck(object):
    """Spell checking for use with wx.StyledTextControl.
    
    This shows spelling errors using the styling indicators (e.g.  the red
    squiggly underline) of the styled text control; I find this much more
    convenient than a dialog-box that makes you click through each mistake.
    
    The eventual goal of the module is to provide on-the-fly spell checking
    that will display errors as you type, and also will highlight errors
    during idle time or in a background thread.
    
    Spell checking is provided through the pyenchant module.  Without
    pyenchant, this object won't do anything useful, but it is still safe to
    be used.  It wraps all calls to pyenchant with try/except blocks to catch
    import errors, and any calls to the spell checking functions will return
    immediately.
    
    To use the spelling check, use one of the methods L{checkAll},
    L{checkCurrentPage}, or L{checkSelection}.  Clear the spelling
    indicators with L{clearAll}.
    """
    # Class attributes to act as default values
    _spelling_lang = None
    _spelling_dict = None
    
    def __init__(self, stc, *args, **kwargs):
        """Mixin must be initialized using this constructor.
        
        Keyword arguments are also available instead of calling the
        convenience functions.  For L{setIndicator}, use C{indicator},
        C{indicator_color}, and {indicator_style}; for L{setLanguage},
        use C{language}; and for L{setMinimumWordSize}, use
        C{min_word_size}.  See the descriptions of those methods for more info.
        """
        self.stc = stc
        self.setIndicator(kwargs.get('indicator', 2),
                          kwargs.get('indicator_color', "#FF0000"),
                          kwargs.get('indicator_style', wx.stc.STC_INDIC_SQUIGGLE))
        self.setMinimumWordSize(kwargs.get('min_word_size', 3))
        if 'language' in kwargs:
            # Don't set default language unless explicitly specified -- it
            # might have already been set through the class method
            self.setDefaultLanguage(kwargs['language'])
        if 'check_region' in kwargs:
            # optional function to specify if the region should be spell
            # checked.  Function should return True if the position should
            # be spell-checked; False if it doesn't make sense to spell check
            # that part of the document
            self._spell_check_region = kwargs['check_region']
        else:
            self._spell_check_region = lambda s: True
        self._spelling_debug = False
        
        self._spelling_last_idle_line = -1
        self.dirty_range_count_per_idle = 5
        
        self._no_update = False
        self._last_block = -1
        
        self.clearDirtyRanges()

    def setIndicator(self, indicator=None, color=None, style=None):
        """Set the indicator styling for misspelled words.
        
        Set the indicator index to use, its color, and the visual style.
        
        @param indicator: indicator number (usually 0, 1, or 2, but may be fewer
        depending on the number of style bits you've chosen for the stc.)
        @param color: string indicating the color of the indicator (e.g.
        "#FF0000" for red)
        @param style: stc indicator style; one of the wx.stc.STC_INDIC_*
        constants (currently wx.stc.STC_INDIC_PLAIN, wx.stc.STC_INDIC_SQUIGGLE,
        wx.stc.STC_INDIC_TT, wx.stc.STC_INDIC_DIAGONAL,
        wx.stc.STC_INDIC_STRIKE, wx.stc.STC_INDIC_HIDDEN,
        wx.stc.STC_INDIC_BOX, wx.stc.STC_INDIC_ROUNDBOX)
        """
        indicators = {0: wx.stc.STC_INDIC0_MASK,
                      1: wx.stc.STC_INDIC1_MASK,
                      2: wx.stc.STC_INDIC2_MASK
                      }
        if indicator is not None:
            if indicator not in indicators:
                indicator = 0
            # The current view may have fewer than 3 indicators
            bitmax = 7 - self.stc.GetStyleBits()
            if indicator > bitmax:
                indicator = bitmax
            self._spelling_indicator = indicator
        self._spelling_indicator_mask = indicators[self._spelling_indicator]
        
        if color is not None:
            self._spelling_color = color
        self.stc.IndicatorSetForeground(self._spelling_indicator,
                                    self._spelling_color)
    
        if style is not None:
            if style > wx.stc.STC_INDIC_MAX:
                style = wx.stc.STC_INDIC_MAX
            self._spelling_style = style
        self.stc.IndicatorSetStyle(self._spelling_indicator,
                               self._spelling_style)
    
    @classmethod
    def getAvailableLanguages(cls):
        """Return a list of supported languages.
        
        Pyenchant supplies a list of its supported languages, so this is just
        a simple wrapper around its C{list_languages} function.  Each item in
        the list is a text string indicating the locale name, e.g.  en_US, ru,
        ru_RU, eo, es_ES, etc.
        
        @return: a list of text strings indicating the supported languages
        """
        try:
            return enchant.list_languages()
        except NameError:
            pass
        return []
    
    @classmethod
    def _getDict(cls, lang):
        try:
            d = enchant.Dict(lang)
        except:
            # Catch all exceptions, because if pyenchant isn't available, you
            # can't catch the enchant.DictNotFound error.
            d = None
        return d

    def setCheckRegion(self, func):
        """Set region checker callable
        @param func: def func(pos): return bool

        """
        self.clearAll()
        self._spell_check_region = func

    @classmethod
    def setDefaultLanguage(cls, lang):
        """Set the default language for spelling check.
        
        The string should be in language locale format, e.g.  en_US, ru, ru_RU,
        eo, es_ES, etc.  See L{getAvailableLanguages}.
        
        @param lang: text string indicating the language
        """
        cls._spelling_lang = lang
        cls._spelling_dict = cls._getDict(lang)

    @classmethod
    def getSpellingDictionary(cls):
        """Get the currently used spelling dictionary
        @return: enchant.Dict instance or None

        """
        return cls._spelling_dict

    def setLanguage(self, lang):
        """Set the language for spelling check for this class, if different than
        the default.
        
        The string should be in language locale format, e.g.  en_US, ru, ru_RU,
        eo, es_ES, etc.  See L{getAvailableLanguages}.
        
        @param lang: text string indicating the language
        """
        # Note that this instance variable will shadow the class attribute
        self._spelling_lang = lang
        self._spelling_dict = self._getDict(lang)
    
    def hasDictionary(self):
        """Returns True if a dictionary is available to spell check the current
        language.
        """
        return self._spelling_dict is not None

    @classmethod
    def isEnchantOk(cls):
        """Returns True if enchant is available"""
        return 'enchant' in globals()

    @classmethod
    def reloadEnchant(cls, libpath=u''):
        """Try (re)loading the enchant module. Use to dynamically try to
        import enchant incase it could be loaded at the time of the import of
        this module.
        @keyword libpath: optionally specify path to libenchant
        @return: bool

        """
        try:
            if libpath and os.path.exists(libpath):
                os.environ['PYENCHANT_LIBRARY_PATH'] = libpath

            if cls.isEnchantOk():
                reload(enchant)
            else:
                mod = __import__('enchant', globals(), locals())
                globals()['enchant'] = mod
        except ImportError:
            return False
        else:
            return True

    def getLanguage(self):
        """Returns True if a dictionary is available to spell check the current
        language.
        """
        return self._spelling_lang
    
    def setMinimumWordSize(self, size):
        """Set the minimum word size that will be looked up in the dictionary.
        
        Words smaller than this size won't be spell checked.
        """
        self._spelling_word_size = size
    
    def clearAll(self):
        """Clear the stc of all spelling indicators."""
        self.stc.StartStyling(0, self._spelling_indicator_mask)
        self.stc.SetStyling(self.stc.GetLength(), 0)
    
    def checkRange(self, start, end):
        """Perform a spell check over a range of text in the document.
        
        This is the main spell checking routine -- it loops over the range
        of text using the L{findNextWord} method to break the text into
        words to check.  Misspelled words are highlighted using the current
        indicator.
        
        @param start: starting position
        @param end: last position to check
        """
        spell = self._spelling_dict
        if not spell:
            return
        
        # Remove any old spelling indicators
        mask = self._spelling_indicator_mask
        count = end - start
        if count <= 0:
            if self._spelling_debug:
                print("No need to check range: start=%d end=%d count=%d" % (start, end, count))
            return
        self.stc.StartStyling(start, mask)
        self.stc.SetStyling(count, 0)
        
        text = self.stc.GetTextRange(start, end) # note: returns unicode
        unicode_index = 0
        max_index = len(text)
        
        last_index = 0 # last character in text a valid raw byte position
        last_pos = start # raw byte position corresponding to last_index
        while unicode_index < max_index:
            start_index, end_index = self.findNextWord(text, unicode_index, max_index)
            if end_index >= 0:
                if end_index - start_index >= self._spelling_word_size:
                    if self._spelling_debug:
                        print("checking %s at text[%d:%d]" % (repr(text[start_index:end_index]), start_index, end_index))
                    if not spell.check(text[start_index:end_index]):
                        # Because unicode characters are stored as utf-8 in the
                        # stc and the positions in the stc correspond to the
                        # raw bytes, not the number of unicode characters, we
                        # have to find out the offset to the unicode chars in
                        # terms of raw bytes.
                        
                        # find the number of raw bytes from the last calculated
                        # styling position to the start of the word
                        last_pos += len(text[last_index:start_index].encode('utf-8'))
                        
                        # find the length of the word in raw bytes
                        raw_count = len(text[start_index:end_index].encode('utf-8'))
                        
                        if self._spell_check_region(last_pos):
                            if self._spelling_debug:
                                print("styling text[%d:%d] = (%d,%d) to %d" % (start_index, end_index, last_pos, last_pos + raw_count, mask))
                            self.stc.StartStyling(last_pos, mask)
                            self.stc.SetStyling(raw_count, mask)
                        elif self._spelling_debug:
                            print("not in valid spell check region.  styling position corresponding to text[%d:%d] = (%d,%d)" % (start_index, end_index, last_pos, last_pos + raw_count))
                        last_pos += raw_count
                        last_index = end_index
                unicode_index = end_index
            else:
                break

    def checkAll(self):
        """Perform a spell check on the entire document."""
        return self.checkRange(0, self.stc.GetLength())
    
    def checkSelection(self):
        """Perform a spell check on the currently selected region."""
        return self.checkRange(self.stc.GetSelectionStart(), self.stc.GetSelectionEnd())
    
    def checkLines(self, startline=-1, count=-1):
        """Perform a spell check on group of lines.
        
        Given the starting line, check the spelling on a block of lines.  If
        the number of lines in the block is not specified, use the number of
        currently visibile lines.
        
        @param startline: current line, or -1 to use the first visible line
        @param count: number of lines in the block, or -1 to use the number of
        lines visible on screen
        """
        if startline < 0:
            startline = self.stc.GetFirstVisibleLine()
        start = self.stc.PositionFromLine(startline)
        if count < 0:
            count = self.stc.LinesOnScreen()
        endline = startline + count
        if endline > self.stc.GetLineCount():
            endline = self.stc.GetLineCount() - 1
        end = self.stc.GetLineEndPosition(endline)
        if self._spelling_debug:
            print("Checking lines %d-%d, chars %d=%d" % (startline, endline, start, end))
        return self.checkRange(start, end)
    
    def checkCurrentPage(self):
        """Perform a spell check on the currently visible lines."""
        return self.checkLines()
    
    def findNextWord(self, utext, index, length):
        """Find the next valid word to check.
        
        Designed to be overridden in subclasses, this method takes a starting
        position in an array of text and returns a tuple indicating the next
        valid word in the string.
        
        @param utext: array of unicode chars
        @param i: starting index within the array to search
        @param length: length of the text
        @return: tuple indicating the word start and end indexes, or (-1, -1)
        indicating that the end of the array was reached and no word was found
        """
        while index < length:
            if utext[index].isalpha():
                end = index + 1
                while end < length and utext[end].isalpha():
                    end += 1
                return (index, end)
            index += 1
        return (-1, -1)
    
    def startIdleProcessing(self):
        """Initialize parameters needed for idle block spell checking.
        
        This must be called before the first call to L{processIdleBlock}
        or if you wish to restart the spell checking from the start
        of the document.  It initializes parameters needed by the
        L{processIdleBlock} in order to process the document during idle
        time.
        """
        self._spelling_last_idle_line = 0
        
    def processIdleBlock(self):
        """Process a block of lines during idle time.
        
        This method is designed to be called during idle processing and will
        spell check a small number of lines.  The next idle processing event
        will continue from where the previous call left off, and in this way
        over some number of idle events will spell check the entire document.
        
        Once the entire document is spell checked, a flag is set and
        further calls to this method will immediately return.  Calling
        L{startIdleProcessing} will cause the idle processing to start
        checking from the beginning of the document.
        """
        self.processDirtyRanges()
        if self._spelling_last_idle_line < 0:
            return
        if self._spelling_debug:
            print("Idle processing page starting at line %d" % self._spelling_last_idle_line)
        self.checkLines(self._spelling_last_idle_line)
        self._spelling_last_idle_line += self.stc.LinesOnScreen()
        if self._spelling_last_idle_line > self.stc.GetLineCount():
            self._spelling_last_idle_line = -1
            return False
        return True

    def processCurrentlyVisibleBlock(self):
        """Alternate method to check lines during idle time.
        
        This method is designed to be called during idle processing and will
        spell check the currently visible block of lines.  Once the visible
        block has been checked, repeatedly calling this method will have
        no effect until the line position changes (or in the less frequent
        occurrence when the number of lines on screen changes by resizing
        the window).
        """
        self.processDirtyRanges()

        self._spelling_last_idle_line = self.stc.GetFirstVisibleLine()
        curr_block = self._spelling_last_idle_line + self.stc.LinesOnScreen()
        if self._no_update or curr_block == self._last_block:
            return

        self.checkLines(self._spelling_last_idle_line)
        self._spelling_last_idle_line += self.stc.LinesOnScreen()
        self._last_block = self._spelling_last_idle_line
        return True

    def getSuggestions(self, word):
        """Get suggestion for the correct spelling of a word.
        
        @param word: word to check
        
        @return: list of suggestions, or an empty list if any of the following
        are true: there are no suggestions, the word is shorter than the
        minimum length, or the dictionary can't be found.
        """
        spell = self._spelling_dict
        if spell and len(word) >= self._spelling_word_size:
            words = spell.suggest(word)
            if self._spelling_debug:
                print("suggestions for %s: %s" % (word, words))
            return words
        return []
    
    def checkWord(self, pos=None, atend=False):
        """Check the word at the current or specified position.
        
        @param pos: position of a character in the word (or at the start or end
        of the word), or None to use the current position
        @param atend: True if you know the cursor is at the end of the word
        """
        if pos is None:
            pos = self.stc.GetCurrentPos()
        if atend:
            end = pos
        else:
            end = self.stc.WordEndPosition(pos, True)
        start = self.stc.WordStartPosition(pos, True)
        if self._spelling_debug:
            print("%d-%d: %s" % (start, end, self.stc.GetTextRange(start, end)))
        self.checkRange(start, end)
    
    def addDirtyRange(self, start, end, lines_added=0, deleted=False):
        """Add a range of characters to a list of dirty regions that need to be
        updated when some idle time is available.
        
        """
        count = end - start
        if deleted:
            count = -count
        if start == self.current_dirty_end:
            self.current_dirty_end = end
        elif start >= self.current_dirty_start and start < self.current_dirty_end:
            self.current_dirty_end += count
        else:
            ranges = []
            if self.current_dirty_start >= 0:
                ranges.append((self.current_dirty_start, self.current_dirty_end))
            for range_start, range_end in self.dirty_ranges:
                if start < range_start:
                    range_start += count
                    range_end += count
                ranges.append((range_start, range_end))
            self.dirty_ranges = ranges
                
            self.current_dirty_start = start
            self.current_dirty_end = end
        
        # If there has been a change before the word that used to be under the
        # cursor, move the pointer so it matches the text
        if start < self.current_word_start:
            self.current_word_start += count
            self.current_word_end += count
        elif start <= self.current_word_end:
            self.current_word_end += count
            # Prevent nonsensical word end if lots of text have been deleted
            if self.current_word_end < self.current_word_start:
                #print("word start = %d, word end = %d" % (self.current_word_start, self.current_word_end))
                self.current_word_end = self.current_word_start
            
        if lines_added > 0:
            start = self.current_dirty_start
            line = self.stc.LineFromPosition(start)
            while True:
                line_end = self.stc.GetLineEndPosition(line)
                if line_end >= end:
                    #self.dirty_ranges.append((start, line_end))
                    if end > start:
                        self.current_dirty_start = start
                        self.current_dirty_end = end
                    else:
                        self.current_dirty_start = self.current_dirty_end = -1
                    break
                self.dirty_ranges.append((start, line_end))
                line += 1
                start = self.stc.PositionFromLine(line)
            
        if self._spelling_debug:
            print("event: %d-%d, current dirty range: %d-%d, older=%s" % (start, end, self.current_dirty_start, self.current_dirty_end, self.dirty_ranges))
    
    def clearDirtyRanges(self, ranges=None):
        """Throw away all dirty ranges
        
        """
        self.current_dirty_start = self.current_dirty_end = -1
        self.current_word_start = self.current_word_end = -1
        if ranges is not None:
            self.dirty_ranges = ranges
        else:
            self.dirty_ranges = []
    
    def processDirtyRanges(self):
        cursor = self.stc.GetCurrentPos()
        
        # Check that the cursor has moved off the current word and if so check
        # its spelling
        if self.current_word_start > 0:
            if cursor < self.current_word_start or cursor > self.current_word_end:
                self.checkRange(self.current_word_start, self.current_word_end)
                self.current_word_start = -1
        
        # Check spelling around the region currently being typed
        if self.current_dirty_start >= 0:
            range_start, range_end = self.processDirtyRange(self.current_dirty_start, self.current_dirty_end)
            
            # If the cursor is in the middle of a word, remove the spelling
            # markers
            if cursor >= range_start and cursor <= range_end:
                word_start = self.stc.WordStartPosition(cursor, True)
                word_end = self.stc.WordEndPosition(cursor, True)
                mask = self._spelling_indicator_mask
                self.stc.StartStyling(word_start, mask)
                self.stc.SetStyling(word_end - word_start, 0)
                
                if word_start != word_end:
                    self.current_word_start = word_start
                    self.current_word_end = word_end
                else:
                    self.current_word_start = -1
            self.current_dirty_start = self.current_dirty_end = -1
        
        # Process a chunk of dirty ranges
        needed = min(len(self.dirty_ranges), self.dirty_range_count_per_idle)
        ranges = self.dirty_ranges[0:needed]
        self.dirty_ranges = self.dirty_ranges[needed:]
        for start, end in ranges:
            if self._spelling_debug:
                print("processing %d-%d" % (start, end))
            self.processDirtyRange(start, end)
    
    def processDirtyRange(self, start, end):
        range_start = self.stc.WordStartPosition(start, True)
        range_end = self.stc.WordEndPosition(end, True)
        if self._spelling_debug:
            print("processing dirty range %d-%d (modified from %d-%d): %s" % (range_start, range_end, start, end, repr(self.stc.GetTextRange(range_start, range_end))))
        self.checkRange(range_start, range_end)
        return range_start, range_end


if __name__ == "__main__":
    import sys
    try:
        import enchant
    except:
        print("pyenchant not available, so spelling correction won't work.")
        print("Get pyenchant from http://pyenchant.sourceforge.net")
    
    class TestSTC(wx.stc.StyledTextCtrl):
        def __init__(self, *args, **kwargs):
            wx.stc.StyledTextCtrl.__init__(self, *args, **kwargs)
            self.spell = STCSpellCheck(self, language="en_US")
            self.SetMarginType(0, wx.stc.STC_MARGIN_NUMBER)
            self.SetMarginWidth(0, 32)
            self.Bind(wx.stc.EVT_STC_MODIFIED, self.OnModified)
            self.Bind(wx.EVT_IDLE, self.OnIdle)
            self.modified_count = 0
            self.idle_count = 0

        def OnModified(self, evt):
            # NOTE: on really big insertions, evt.GetText can cause a
            # MemoryError on MSW, so I've commented this dprint out.
            mod = evt.GetModificationType()
            if mod & wx.stc.STC_MOD_INSERTTEXT or mod & wx.stc.STC_MOD_DELETETEXT:
                #print("(%s) at %d: text=%s len=%d" % (self.transModType(evt.GetModificationType()),evt.GetPosition(), repr(evt.GetText()), evt.GetLength()))
                pos = evt.GetPosition()
                last = pos + evt.GetLength()
                self.spell.addDirtyRange(pos, last, evt.GetLinesAdded(), mod & wx.stc.STC_MOD_DELETETEXT)
                #self.modified_count += 1
                #if self.modified_count > 10:
                #    wx.CallAfter(self.spell.processDirtyRanges)
                #    self.modified_count = 0
            evt.Skip()
        
        def OnIdle(self, evt):
            #print("Idle")
            self.idle_count += 1
            if self.idle_count > 10:
                self.spell.processIdleBlock()
                self.idle_count = 0
            
        def transModType(self, modType):
            st = ""
            table = [(wx.stc.STC_MOD_INSERTTEXT, "InsertText"),
                     (wx.stc.STC_MOD_DELETETEXT, "DeleteText"),
                     (wx.stc.STC_MOD_CHANGESTYLE, "ChangeStyle"),
                     (wx.stc.STC_MOD_CHANGEFOLD, "ChangeFold"),
                     (wx.stc.STC_PERFORMED_USER, "UserFlag"),
                     (wx.stc.STC_PERFORMED_UNDO, "Undo"),
                     (wx.stc.STC_PERFORMED_REDO, "Redo"),
                     (wx.stc.STC_LASTSTEPINUNDOREDO, "Last-Undo/Redo"),
                     (wx.stc.STC_MOD_CHANGEMARKER, "ChangeMarker"),
                     (wx.stc.STC_MOD_BEFOREINSERT, "B4-Insert"),
                     (wx.stc.STC_MOD_BEFOREDELETE, "B4-Delete")
                     ]

            for flag,text in table:
                if flag & modType:
                    st = st + text + " "

            if not st:
                st = 'UNKNOWN'

            return st

    class Frame(wx.Frame):
        def __init__(self, *args, **kwargs):
            super(self.__class__, self).__init__(*args, **kwargs)

            self.stc = TestSTC(self, -1)

            self.CreateStatusBar()
            menubar = wx.MenuBar()
            self.SetMenuBar(menubar)  # Adding the MenuBar to the Frame content.
            menu = wx.Menu()
            menubar.Append(menu, "File")
            self.menuAdd(menu, "Open", "Open File", self.OnOpenFile)
            self.menuAdd(menu, "Quit", "Exit the pragram", self.OnQuit)
            menu = wx.Menu()
            menubar.Append(menu, "Edit")
            self.menuAdd(menu, "Check All", "Spell check the entire document", self.OnCheckAll)
            self.menuAdd(menu, "Check Current Page", "Spell check the currently visible page", self.OnCheckPage)
            self.menuAdd(menu, "Check Selection", "Spell check the selected region", self.OnCheckSelection)
            menu.AppendSeparator()
            self.menuAdd(menu, "Clear Spelling", "Remove spelling correction indicators", self.OnClearSpelling)
            menu = wx.Menu()
            menubar.Append(menu, "Language")
            langs = self.stc.spell.getAvailableLanguages()
            self.lang_id = {}
            for lang in langs:
                id = wx.NewId()
                self.lang_id[id] = lang
                self.menuAdd(menu, lang, "Change dictionary to %s" % lang, self.OnChangeLanguage, id=id)


        def loadFile(self, filename):
            fh = open(filename)
            self.stc.SetText(fh.read())
            self.stc.spell.clearDirtyRanges()
            self.stc.spell.checkCurrentPage()
        
        def loadSample(self, paragraphs=10):
            lorem_ipsum = u"""\
Lorem ipsum dolor sit amet, consectetuer adipiscing elit.  Vivamus mattis
commodo sem.  Phasellus scelerisque tellus id lorem.  Nulla facilisi.
Suspendisse potenti.  Fusce velit odio, scelerisque vel, consequat nec,
dapibus sit amet, tortor.  Vivamus eu turpis.  Nam eget dolor.  Integer
at elit.  Praesent mauris.  Nullam non nulla at nulla tincidunt malesuada.
Phasellus id ante.  Sed mauris.  Integer volutpat nisi non diam.  Etiam
elementum.  Pellentesque interdum justo eu risus.  Cum sociis natoque
penatibus et magnis dis parturient montes, nascetur ridiculus mus.  Nunc
semper.  In semper enim ut odio.  Nulla varius leo commodo elit.  Quisque
condimentum, nisl eget elementum laoreet, mauris turpis elementum felis, ut
accumsan nisl velit et mi.

And some Russian: \u041f\u0438\u0442\u043e\u043d - \u043b\u0443\u0447\u0448\u0438\u0439 \u044f\u0437\u044b\u043a \u043f\u0440\u043e\u0433\u0440\u0430\u043c\u043c\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f!

"""
            self.stc.ClearAll()
            for i in range(paragraphs):
                self.stc.AppendText(lorem_ipsum)
            # Call the spell check after the text has had a chance to be
            # displayed and the window resized to the correct size.
            self.stc.spell.clearDirtyRanges()
            wx.CallAfter(self.stc.spell.checkCurrentPage)

        def menuAdd(self, menu, name, desc, fcn, id=-1, kind=wx.ITEM_NORMAL):
            if id == -1:
                id = wx.NewId()
            a = wx.MenuItem(menu, id, name, desc, kind)
            menu.AppendItem(a)
            wx.EVT_MENU(self, id, fcn)
            menu.SetHelpString(id, desc)
        
        def OnOpenFile(self, evt):
            dlg = wx.FileDialog(self, "Choose a text file",
                               defaultDir = "",
                               defaultFile = "",
                               wildcard = "*")
            if dlg.ShowModal() == wx.ID_OK:
                print("Opening %s" % dlg.GetPath())
                self.loadFile(dlg.GetPath())
            dlg.Destroy()
        
        def OnQuit(self, evt):
            self.Close(True)
        
        def OnCheckAll(self, evt):
            self.stc.spell.checkAll()
        
        def OnCheckPage(self, evt):
            self.stc.spell.checkCurrentPage()
        
        def OnCheckSelection(self, evt):
            self.stc.spell.checkSelection()
        
        def OnClearSpelling(self, evt):
            self.stc.spell.clearAll()
        
        def OnChangeLanguage(self, evt):
            id = evt.GetId()
            normalized = locale.normalize(self.lang_id[id])
            try:
                locale.setlocale(locale.LC_ALL, normalized)
                print("Changing locale %s, dictionary set to %s" % (normalized, self.lang_id[id]))
            except locale.Error:
                print("Can't set python locale to %s; dictionary set to %s" % (normalized, self.lang_id[id]))
            self.stc.spell.setLanguage(self.lang_id[id])
            self.stc.spell.clearAll()
            self.stc.spell.checkCurrentPage()

    app = wx.App(False)
    frame = Frame(None, size=(600, -1))
    need_sample = True
    if len(sys.argv) > 1:
        if not sys.argv[-1].startswith("-"):
            frame.loadFile(sys.argv[-1])
            need_sample = False
    if need_sample:
        frame.loadSample()
    if '-d' in sys.argv:
        frame.stc.spell._spelling_debug = True
    frame.Show()
    app.MainLoop()