from util import defaultdict
import codecs, datetime, grp, os, pwd, re, shutil, stat, string
import fcntl, select, pty, threading
import chardet
#import aes, dirutil, regex, highlight, lex, term
import dirutil, regex, highlight, lex, term
from point import Point
from subprocess import Popen, PIPE, STDOUT
from keyinput import MAP

# undo/redo stack constants
ACT_NONE = 'none'
ACT_NORM = 'norm'
ACT_UNDO = 'undo'
ACT_REDO = 'redo'

STACK_LIMIT = 1024

def hasher(data):
    try:
        import hashlib
        m = hashlib.md5(data)
    except:
        import md5
        m = md5.new(data)
    return m

class ReadOnlyError(Exception): pass
class FileChangedError(Exception): pass
class FileGoneError(Exception): pass

# used for undo/redo stacks when text will need to be added back
class AddMove(object):
    def __init__(self, buffer, p, lines):
        self.buffer  = buffer
        self.p       = p
        self.lines   = lines
        self.undo_id = buffer.undo_id
    def restore(self, act):
        assert act in (ACT_UNDO, ACT_REDO)
        self.buffer.insert_lines(self.p, self.lines, act)
    def getpos(self):
        return self.p

# used for undo/redo stacks when text will need to be removed
class DelMove(object):
    def __init__(self, buffer, p1, p2):
        self.buffer  = buffer
        self.p1      = p1
        self.p2      = p2
        self.undo_id = buffer.undo_id
    def restore(self, act):
        assert act in (ACT_UNDO, ACT_REDO)
        self.buffer.delete(self.p1, self.p2, act)
    def getpos(self):
        return self.p1

# abstract class
class Buffer(object):
    btype    = 'generic'
    modename = None
    mac_re   = re.compile('\r(?!\n)')
    unix_re  = re.compile('(?<!\r)\n')
    win_re   = re.compile('\r\n')
    def __init__(self, stack_limit=STACK_LIMIT):
        self.lines       = [""]
        self.windows     = []
        self.undo_id     = 1
        self.undo_stack  = []
        self.redo_stack  = []
        self.stack_limit = stack_limit
        self.nl          = '\n'
        self.modified    = False
        self.highlights  = {}
        self.settings    = {}
        self.indentlvl   = 4
        self.usetabs     = False
        self.writetabs   = False
        self.metadata    = {}

    def _detect_nl_type(self, data):
        mac_c = unix_c = win_c = 0
        l = len(data)
        i = 0
        while i < l:
            if data[i:i + 2] == '\r\n':
                win_c += 1
                i += 2
            else:
                if data[i] == '\n':
                    unix_c += 1
                elif data[i] == '\r':
                    mac_c += 1
                i += 1

        if (unix_c and mac_c) or (unix_c and win_c) or (mac_c and win_c):
            # warn the user?
            pass

        if unix_c >= win_c and unix_c >= mac_c:
            return '\n'
        elif mac_c >= win_c:
            return '\r'
        else:
            return '\r\n'

    # basic file operation stuff
    def _open_file_r(self, path):
        path = os.path.realpath(path)
        if not os.path.isfile(path):
            raise Exception("Path '%s' does not exist" % (path))
        if not os.access(path, os.R_OK):
            raise Exception("Path '%s' cannot be read" % (path))
        f = open(path, 'r')
        return f
    def _open_file_w(self, path):
        if os.path.isfile(path):
            raise Exception("Path '%s' already exists" % (path))
        d = os.path.dirname(path)
        if not os.access(d, os.R_OK):
            raise Exception("Dir '%s' cannot be read" % (path))
        if not os.access(d, os.W_OK):
            raise Exception("Dir '%s' cannot be written" % (path))
        f = open(path, 'w')
        return f
    def _temp_path(self, path):
        (dirname, basename) = os.path.split(path)
        return os.path.join(dirname, ".__%s__pmacs" % (basename))

    # undo/redo stack
    def _stack_trim(self, stack):
        if self.stack_limit:
            while len(stack) > self.stack_limit:
                stack.pop(0)
    def add_to_stack(self, move, act):
        if act == ACT_NONE:
            pass
        elif act == ACT_NORM:
            self.redo_stack = []
            self.undo_stack.append(move)
            self._stack_trim(self.undo_stack)
        elif act == ACT_UNDO:
            self.redo_stack.append(move)
            self._stack_trim(self.redo_stack)
        elif act == ACT_REDO:
            self.undo_stack.append(move)
            self._stack_trim(self.undo_stack)
        else:
            raise Exception("Invalid act: %d" % (act))
    def undo(self):
        if len(self.undo_stack):
            undo_id = self.undo_stack[-1].undo_id
            pos = None
            while self.undo_stack and self.undo_stack[-1].undo_id == undo_id:
                move = self.undo_stack.pop(-1)
                move.restore(ACT_UNDO)
                pos = move.getpos()
            return pos
        else:
            raise Exception("Nothing to Undo!")
    def redo(self):
        if len(self.redo_stack):
            undo_id = self.redo_stack[-1].undo_id
            pos = None
            while self.redo_stack and self.redo_stack[-1].undo_id == undo_id:
                move = self.redo_stack.pop(-1)
                move.restore(ACT_REDO)
                pos = move.getpos()
            return pos
        else:
            raise Exception("Nothing to Redo!")

    # window-buffer communication
    def add_window(self, w):
        if w not in self.windows:
            self.windows.append(w)
            modename = w.mode.name
            if modename not in self.highlights and w.mode.lexer is not None:
                self.highlights[modename] = highlight.Highlighter(w.mode.lexer)
                self.highlights[modename].highlight(self.lines)
            if modename not in self.settings:
                self.settings[modename] = {}

    def remove_window(self, w):
        if w in self.windows:
            self.windows.remove(w)
            if w.mode.name in self.highlights:
                for w2 in self.windows:
                    if w2.mode.name == w.mode.name:
                        return
                del self.highlights[w.mode.name]
    def _region_add(self, p1, p2, lines, act):
        move = DelMove(self, p1, p2)
        self.add_to_stack(move, act)
        for name in self.highlights:
            self.highlights[name].relex_add(self.lines, p1.y, p1.x, lines)
        for w in self.windows:
            w.region_added(p1, lines)
    def _region_del(self, p1, p2, lines, act):
        move = AddMove(self, p1, lines)
        self.add_to_stack(move, act)
        for name in self.highlights:
            self.highlights[name].relex_del(self.lines, p1.y, p1.x, p2.y, p2.x)
        for w in self.windows:
            w.region_removed(p1, p2)

    # internal validation
    def _validate_point(self, p):
        self._validate_xy(p.x, p.y)
    def _validate_xy(self, x, y):
        assert y >= 0 and y < len(self.lines), \
            "xy1: %d >= 0 and %d < %d" % (y, y, len(self.lines))
        assert x >= 0 and x <= len(self.lines[y]), \
            "xy2: %d >= 0 and %d <= %d" % (x, x, len(self.lines[y]))
    def _validate_y(self, y):
        assert y >= 0 and y < len(self.lines), \
            "y: %d >= 0 and %d < %d" % (y, y, len(self.lines))

    # deal with the actual logical document string
    def num_chars(self):
        n = 0
        for line in self.lines[:-1]:
            n += len(line) + 1
        n += len(self.lines[-1])
        return n
    def num_lines(self):
        return len(self.lines)
    def make_string(self):
        if self.writetabs:
            # reverse the "horrible hack" mentioned below when writing
            # out the string.
            lines2 = [s.replace('\t  \t', '\t') for s in self.lines]
        else:
            lines2 = self.lines
        return self.nl.join(self.lines)

    # methods to be overridden by subclasses
    def name(self):
        return "Generic"
    def close(self):
        pass
    def open(self):
        pass
    def changed(self):
        return self.modified
    def changed_on_disk(self):
        return False
    def reload(self):
        raise Exception("%s reload: Unimplemented" % (self.name()))
    def save_as(self, path, force=False):
        # check to see if the path exists, and if we're prepared to overwrite it
        # if yes to both, get its mode so we can preserve the path's permissions
        mode = None
        if os.path.exists(path):
            if force:
                mode = os.stat(path)[0]
            else:
                # XYZ
                raise Exception("oh no! %r already exists" % path)

        # create the string that we're going to write into the file
        data = self.write_filter(self.make_string())

        # create a safe temporary path to write to, and write out data to it
        temp_path = self._temp_path(path)
        f2 = self._open_file_w(temp_path)
        f2.write(data)
        f2.close()

        # move the temporary file to the actual path; maybe change permissions
        shutil.move(temp_path, path)
        if mode:
            os.chmod(path, mode)

        # the file has not been modified now
        self.modified = False
    def backup(self):
        '''backup path, and return the path to the temporary backup file'''
        tf, tpath = tempfile.mkstemp(prefix='pmc')
        tf.write(open(self.path, 'rb').read())
        tf.close()
        return tpath
    def readonly(self):
        return False
    def read_filter(self, data):
        return data
    def write_filter(self, data):
        return data

    # point retrieval
    def get_buffer_start(self):
        return Point(0, 0)
    def get_buffer_end(self):
        return Point(len(self.lines[-1]), len(self.lines) - 1)

    # data retrieval
    def get_sublines(self, p1, p2):
        self._validate_point(p1)
        self._validate_point(p2)
        assert p1 <= p2, "p1.x (%d) > p2.x (%d)" % (p1.x, p2.x)
        lines = []
        x = p1.x
        for i in xrange(p1.y, p2.y):
            lines.append(self.lines[i][x:])
            x = 0
        lines.append(self.lines[p2.y][x:p2.x])
        return lines
    def get_substring(self, p1, p2):
        lines = self.get_sublines(p1, p2)
        return '\n'.join(lines)

    # buffer set
    def set_lines(self, lines, force=False):
        if not force and self.readonly():
            raise Exception("set_data: buffer is readonly")
        start = self.get_buffer_start()
        end   = self.get_buffer_end()
        self.delete(start, end, force=force)
        self.insert_lines(start, lines, force=force)
        self.modified = True
    def set_data(self, data, force=False):
        lines = data.split('\n')
        self.set_lines(lines, force)

    # append into buffer
    def append_lines(self, lines, act=ACT_NORM, force=False):
        p = self.get_buffer_end()
        self.insert_lines(p, lines, act, force)
    def append_string(self, s, act=ACT_NORM, force=False):
        lines = s.split("\n")
        self.insert_lines(lines, act, force)

    # insertion into buffer
    def insert_lines(self, p, lines, act=ACT_NORM, force=False):
        llen = len(lines)
        assert llen > 0
        if not force and self.readonly():
            raise ReadOnlyError("buffer is read-only")
        p2 = p.vadd(len(lines[-1]), llen - 1)
        if llen > 1:
            self.lines.insert(p.y + 1, [])
            self.lines[p.y + 1] = lines[-1] + self.lines[p.y][p.x:]
            self.lines[p.y] = self.lines[p.y][:p.x] + lines[0]
            for i in xrange(1, llen - 1):
                self.lines.insert(p.y + i, lines[i])
        else:
            self.lines[p.y] = self.lines[p.y][:p.x] + lines[-1] + self.lines[p.y][p.x:]
        self._region_add(p, p2, lines, act)
        self.modified = True
    def insert_string(self, p, s, act=ACT_NORM, force=False):
        lines = s.split("\n")
        self.insert_lines(p, lines, act, force)
        
    # deletion from buffer
    def delete(self, p1, p2, act=ACT_NORM, force=False):
        """delete characters from p1 up to p2 from the buffer"""
        if not force and self.readonly():
            raise ReadOnlyError("buffer is read-only")
        self._validate_point(p1)
        self._validate_point(p2)
        if p1 == p2:
            return
        assert p1 < p2, "p1 %r > p2 %r" % (p1, p2)
        lines = self.get_sublines(p1, p2)
        line1 = self.lines[p1.y]
        line2 = self.lines[p2.y]
        self.lines[p1.y:p2.y+1] = ["%s%s" % (line1[:p1.x], line2[p2.x:])]
        self._region_del(p1, p2, lines, act)
        self.modified = True
    def delete_char(self, p, act=ACT_NORM, force=False):
        if p.x == len(self.lines[p.y]):
            p2 = Point(0, p.y + 1)
        else:
            p2 = Point(p.x + 1, p.y)
        self.delete(p, p2, act=act, force=force)
    def overwrite_char(self, p, c, act=ACT_NORM, force=False):
        self.delete_char(p, act=act, force=force)
        self.insert_string(p, c, act=act, force=force)
    def delete_line(self, y, act=ACT_NORM, force=False):
        p1 = Point(0, y)
        if y < len(self.lines) - 1:
            p2 = Point(0, y + 1)
        else:
            p2 = Point(len(self.lines[-1]), y)
        self.delete(p1, p2, act, force)

    # random
    def is_whitespace(self, y):
        return regex.whitespace.match(self.lines[y])
    def count_leading_whitespace(self, y):
        m = regex.leading_whitespace.match(self.lines[y])
        return m.end()
    def detect_indent_level(self, y1, y2):
        x = None
        for y in xrange(y1, y2):
            if self.is_whitespace(y):
                continue
            c = self.count_leading_whitespace(y)
            if x is None:
                x = c
            else:
                x = min(x, c)
        return x or 0

    # generic window functionality
    def forward(self, p):
        if p.x < len(self.lines[p.y]):
            return Point(p.x + 1, p.y)
        elif p.y < len(self.lines) - 1:
            return Point(0, p.y + 1)
        else:
            return p
    def backward(self, p):
        if p.x > 0:
            return Point(p.x - 1, p.y)
        elif p.y > 0:
            x = len(self.lines[p.y - 1])
            return Point(x, p.y - 1)
        else:
            return p
    def end_of_line(self, p):
        return Point(len(self.lines[p.y]), p.y)
    def start_of_line(self, p):
        return Point(0, p.y)
    def previous_line(self, p):
        if p.y > 0:
            return Point(p.x, p.y - 1)
        else:
            return p
    def next_line(self, p):
        if p.y < len(self.lines) - 1:
            return Point(p.x, p.y + 1)
        else:
            return p
    def left_delete(self, p):
        (x, y) = p.xy()
        if x > 0:
            self.delete_char(Point(x - 1, y))
        elif y > 0:
            x = len(self.lines[y - 1])
            self.delete_char(Point(x, y - 1))
    def right_delete(self, p):
        if (p.y < len(self.lines) - 1 or
            p.x < len(self.lines[-1])):
            self.delete_char(p)

class InterpreterPipeError(Exception):
    pass

class InterpreterBuffer(Buffer):
    _basename = 'Interpreter'
    def create_name(cls, parent):
        if hasattr(parent, 'path'):
            return '*%s:%s*' % (cls._basename, parent.name())
        else:
            return '*%s*' % cls._basename
    create_name = classmethod(create_name)
    btype    = 'interpreter'
    readre   = re.compile('^([A-Z]+):(.*)\n$')
    def __init__(self, parent, app):
        self.application = app
        if parent and hasattr(parent, 'path'):
            self.parent = parent
        else:
            self.parent = None
        Buffer.__init__(self)
        cmd = self.get_cmd()
        env = dict(os.environ)
        env.update(self.get_env())
        f = open('/dev/null', 'w')
        #f = open('pipe-errors', 'w')
        self.pipe = Popen(cmd, stdin=PIPE, stdout=PIPE, stderr=f, env=env)
        self.prompt = '***'
        self.clear()
        self.pipe_read()
        self._name = self.create_name(parent)
    def name(self):
        return self._name
    def get_env(self):
        return {}
    def get_cmd(self):
        raise Exception('unimplemented')
    def pipe_readline(self):
        if self.pipe.poll() is not None:
            raise InterpreterPipeError('broken pipe')
        line = self.pipe.stdout.readline()
        m = self.readre.match(line)
        if m:
            return (m.group(1), m.group(2))
        else:
            return (None, line.rstrip())
    def pipe_read(self):
        lines = []
        while True:
            (type_, value) = self.pipe_readline()
            if type_ == 'PROMPT':
                self.prompt = value.strip() + ' '
                break
            value.rstrip()
            if value:
                lines.append(value)
        if lines:
            output = '\n'.join(lines) + '\n'
            p = self.get_buffer_end()
            self.insert_string(p, output, force=True)
    def pipe_write(self, s):
        self.pipe.stdin.write("%s\n" % s)
        self.pipe.stdin.flush()
    def completions(self, word):
        self.pipe_write("COMPLETE:%s" % word)
        candidates = self.pipe_read_completions()
        self.pipe_read()
        return candidates
    def pipe_read_completions(self):
        try:
            (typ_, value) = self.pipe_readline()
            assert typ_ == 'COMPLETIONS', '%r %r' % (typ_, value)
            candidates = [x for x in value.split('|') if x]
            return candidates
        except:
            return []
    def clear(self):
        self.set_data('', force=True)
    def changed(self):
        return False
    def readonly(self):
        return True

class IperlBuffer(InterpreterBuffer):
    _basename = 'IPerl'
    btype     = 'iperl'
    modename  = 'iperl'
    def create_name(cls, parent):
        if parent and hasattr(parent, 'path'):
            if parent.path.endswith('.pm'):
                return '*%s:%s*' % (cls._basename, parent.name())
            else:
                raise Exception("not a perl module")
        else:
            return '*%s*' % cls._basename
    create_name = classmethod(create_name)
    def get_cmd(self):
        cmd = ['iperl', '-p']
        if self.parent:
            cmd.extend(['-r', self.parent.path])
        return cmd
    def get_env(self):
        lib = ':'.join(self.application.config.get('perl.libs', []))
        return {'PERL5LIB': lib}
    def readline_completions(self, x1, x2, line):
        self.pipe.stdin.write("READLINE:%d:%d:%s\n" % (x1, x2, line))
        self.pipe.stdin.flush()
        (typ_, value) = self.pipe_readline()
        assert typ_ == 'COMPLETIONS', '%r %r' % (typ_, value)
        candidates = [x for x in value.split('|') if x]
        self.pipe_read()
        return candidates

class IpythonBuffer(InterpreterBuffer):
    _basename = 'IPython'
    btype     = 'ipython'
    modename  = 'ipython'
    def get_cmd(self):
        if self.parent:
            return ('epython', '-p', '-r', self.parent.path)
        else:
            return ('epython', '-p')
    def get_env(self):
        return {'PYTHONPATH': self.application.config.get('python.lib', '.')}

class BinaryDataException(Exception):
    pass

class FileBuffer(Buffer):
    btype = 'file'
    def __init__(self, path, name=None):
        '''fb = FileBuffer(path)'''
        Buffer.__init__(self)
        self.path      = os.path.realpath(path)
        self.checksum  = None
        self.codec     = 'utf-8'
        self._name     = name or os.path.basename(path)
        self._readonly = os.path.exists(path) and not os.access(path, os.W_OK)

    def readonly(self):
        return self._readonly

    def _open_file_r(self, path=None):
        if path is None:
            path = self.path
        path = os.path.realpath(path)
        self.path = path
        if not os.path.isfile(path):
            raise Exception("Path '%s' does not exist" % (path))
        if not os.access(path, os.R_OK):
            raise Exception("Path '%s' cannot be read" % (path))
        f = open(path, 'r')
        return f
    def _open_file_w(self, path=None, preserve=True):
        if path is None:
            path = self.path
        if preserve and os.path.isfile(path):
            raise Exception("Path '%s' already exists" % (path))
        d = os.path.dirname(path)
        if not os.access(d, os.R_OK):
            raise Exception("Dir '%s' cannot be read" % (path))
        if not os.access(d, os.W_OK):
            raise Exception("Dir '%s' cannot be written" % (path))
        f = open(path, 'w')
        return f
    def _temp_path(self, path=None):
        if path is None:
            path = self.path
        (dirname, basename) = os.path.split(path)
        return os.path.join(dirname, ".__%s__pmacs" % (basename))

    # methods for dealing with the underlying resource, etc.
    def name(self):
        return self._name
    def path_exists(self):
        return os.path.exists(self.path)
    def store_checksum(self, data):
        self.checksum = hasher(data)

    def read(self):
        if self.path_exists():
            f = self._open_file_r()
            data = f.read()
            f.close()
            self.store_checksum(data)
            data = self.read_filter(data)
            codec = chardet.detect(data)['encoding'] or 'utf-8'
            self.codec = codec.lower()
        else:
            data = ''
            self.codec = 'utf-8'

        if self.codec == 'utf-8' and data.startswith(codecs.BOM_UTF8):
            self.codec = 'utf-8-sig'
        elif self.codec.startswith('utf-16'):
            self.codec = 'utf-16'

        data = data.decode(self.codec)

        if '\t' in data:
            self.writetabs = True
            self.usetabs   = True

            # OK, this is a horrible, horrible hack. because the "one unicode
            # char occupies one screen position" assumption is built into the
            # syntax highlighting and render strings at a pretty low level, we
            # have to "pad out" the tab to make it take up enough physical
            # space.
            data = data.replace('\t', '\t  \t')

        self.nl = self._detect_nl_type(data)

        if '\x00' in data[:8192]:
            raise BinaryDataException("binary files are not supported")

        return data

    def open(self):
        data = self.read()
        self.lines = data.split(self.nl)
    def reload(self):
        data = self.read()
        self.set_data(data)
    def changed_on_disk(self):
        assert self.checksum is not None
        f = open(self.path)
        data = f.read()
        f.close()
        m = hasher(data)
        return self.checksum.digest() != m.digest()
    def save(self, force=False):
        if self.readonly():
            raise ReadOnlyError("can't save read-only file")

        if self.checksum is not None and force is False:
            # the file already existed and we took a checksum so make sure it's
            # still the same right now
            if not self.path_exists():
                raise FileGoneError("oh no! %r disappeared!" % self.path)
            if self.changed_on_disk():
                raise FileChangedError("oh no! %r has changed on-disk!" % self.path)

        exists = os.path.exists(self.path)
        if exists:
            temp_path = self._temp_path()
            shutil.copyfile(self.path, temp_path)


        try:
            data = self.write_filter(self.make_string().encode(self.codec))
            f2   = self._open_file_w(self.path, preserve=False)
            f2.write(data)
            f2.close()
        except NameError, e:
            if exists: shutil.copyfile(temp_path, self.path)
            raise e
        else:
            self.store_checksum(data)
            self.modified = False
        if exists: os.unlink(temp_path)
    def save_as(self, path):
        self.path = path
        self.save()

class Binary32Buffer(FileBuffer):
    btype     = 'bin32file'
    grouppad  = 2
    groupsize = 8
    numgroups = 2
    bytepad   = 1
    data      = None
    wordsize  = 4
    def __init__(self, path, name=None):
        '''fb = FileBuffer(path)'''
        FileBuffer.__init__(self, path, name)
    def _detect_nl_type(self, data):
        return '\n'
    def cursorx_to_datax(self, cy, cx):
        bytespace = 2 + self.bytepad
        groupspace = bytespace * self.groupsize - self.bytepad + self.grouppad

        groupmod = (cx + self.grouppad) % groupspace
        if groupmod < self.grouppad:
            return None

        groupdiv = (cx + 2) // groupspace
        if groupdiv >= self.numgroups:
            return None

        bytemod = (cx + self.bytepad - groupdiv) % bytespace
        if bytemod == 0:
            return None
        
        bytediv = ((cx + self.bytepad) % groupspace) // bytespace
        ix = self.groupsize * groupdiv  + bytediv
        if ix < len(self.rawdata[cy]):
            return ix
        else:
            return None

    def datax_to_cursorx(self, ix):
        groupsize = (((2 + self.bytepad) * self.groupsize) + self.grouppad)
        maxsize = groupsize * self.numgroups - self.grouppad
        if ix < maxsize:
            return (ix // self.groupsize) * (self.grouppad - 1) + ix * (2 + self.bytepad)
        else:
            return None
    def datax_to_cursory(self, ix):
        return ix // (self.groupsize * self.numgroups)

    def get_address(self, cy, ix):
        return (cy * self.numgroups * self.groupsize) + ix

    def overwrite_char(self, p, c, act=ACT_NORM, force=False):
        ix = self.cursorx_to_datax(p.y, p.x)
        if ix is None:
            return
        Buffer.overwrite_char(self, p, c, act, force)
        cx = self.datax_to_cursorx(ix)
        c = chr(int(self.lines[p.y][cx:cx + 2], 16))
        rawline = self.rawdata[p.y]
        self.rawdata[p.y] = rawline[0:ix] + c + rawline[ix + 1:]

    def read_filter(self, data):
        bytepad      = ' ' * self.bytepad
        grouppad     = ' ' * self.grouppad
        self.rawdata = []
        lines        = []
        i            = 0
        while i < len(data):
            self.rawdata.append(data[i:i + self.numgroups * self.groupsize])
            j = 0
            groups = []
            while j < self.numgroups * self.groupsize and i + j < len(data):
                bytes = []
                for c in data[i + j:i + j + self.groupsize]:
                    bytes.append(string.hexdigits[ord(c) / 16] + string.hexdigits[ord(c) % 16])
                groups.append(bytepad.join(bytes))
                j += self.groupsize
            lines.append(grouppad.join(groups))
            i += self.numgroups * self.groupsize
        if not self.rawdata:
            self.rawdata = ['']
        return '\n'.join(lines)
    def write_filter(self, data):
        return ''.join(self.rawdata)

# log is another singleton
log = None
class LogBuffer(Buffer):
    btype = 'log'
    def __new__(cls, *args, **kwargs):
        global log
        if log is None:
            log = object.__new__(LogBuffer, *args, **kwargs)
        return log

    def __init__(self): Buffer.__init__(self)
    def clear(self): log.set_data('', force=True)
    def name(self): return '*Log*'
    def changed(self): return False
    def close(self): global log; log = None
    def readonly(self): return True