from __future__ import absolute_import, division, unicode_literals

from six import text_type, binary_type
from six.moves import http_client, urllib

import codecs
import re

import webencodings

from .constants import EOF, spaceCharacters, asciiLetters, asciiUppercase
from .constants import ReparseException
from . import _utils

from io import StringIO

try:
    from io import BytesIO
except ImportError:
    BytesIO = StringIO

# Non-unicode versions of constants for use in the pre-parser
spaceCharactersBytes = frozenset([item.encode("ascii") for item in spaceCharacters])
asciiLettersBytes = frozenset([item.encode("ascii") for item in asciiLetters])
asciiUppercaseBytes = frozenset([item.encode("ascii") for item in asciiUppercase])
spacesAngleBrackets = spaceCharactersBytes | frozenset([b">", b" Normalized stream from source
        for use by html5lib.

        source can be either a file-object, local filename or a string.

        The optional encoding parameter must be a string that indicates
        the encoding.  If specified, that encoding will be used,
        regardless of any BOM or later declaration (such as in a meta
        element)

        """

        if not _utils.supports_lone_surrogates:
            # Such platforms will have already checked for such
            # surrogate errors, so no need to do this checking.
            self.reportCharacterErrors = None
        elif len("\U0010FFFF") == 1:
            self.reportCharacterErrors = self.characterErrorsUCS4
        else:
            self.reportCharacterErrors = self.characterErrorsUCS2

        # List of where new lines occur
        self.newLines = [0]

        self.charEncoding = (lookupEncoding("utf-8"), "certain")
        self.dataStream = self.openStream(source)

        self.reset()

    def reset(self):
        self.chunk = ""
        self.chunkSize = 0
        self.chunkOffset = 0
        self.errors = []

        # number of (complete) lines in previous chunks
        self.prevNumLines = 0
        # number of columns in the last line of the previous chunk
        self.prevNumCols = 0

        # Deal with CR LF and surrogates split over chunk boundaries
        self._bufferedCharacter = None

    def openStream(self, source):
        """Produces a file object from source.

        source can be either a file object, local filename or a string.

        """
        # Already a file object
        if hasattr(source, 'read'):
            stream = source
        else:
            stream = StringIO(source)

        return stream

    def _position(self, offset):
        chunk = self.chunk
        nLines = chunk.count('\n', 0, offset)
        positionLine = self.prevNumLines + nLines
        lastLinePos = chunk.rfind('\n', 0, offset)
        if lastLinePos == -1:
            positionColumn = self.prevNumCols + offset
        else:
            positionColumn = offset - (lastLinePos + 1)
        return (positionLine, positionColumn)

    def position(self):
        """Returns (line, col) of the current position in the stream."""
        line, col = self._position(self.chunkOffset)
        return (line + 1, col)

    def char(self):
        """ Read one character from the stream or queue if available. Return
            EOF when EOF is reached.
        """
        # Read a new chunk from the input stream if necessary
        if self.chunkOffset >= self.chunkSize:
            if not self.readChunk():
                return EOF

        chunkOffset = self.chunkOffset
        char = self.chunk[chunkOffset]
        self.chunkOffset = chunkOffset + 1

        return char

    def readChunk(self, chunkSize=None):
        if chunkSize is None:
            chunkSize = self._defaultChunkSize

        self.prevNumLines, self.prevNumCols = self._position(self.chunkSize)

        self.chunk = ""
        self.chunkSize = 0
        self.chunkOffset = 0

        data = self.dataStream.read(chunkSize)

        # Deal with CR LF and surrogates broken across chunks
        if self._bufferedCharacter:
            data = self._bufferedCharacter + data
            self._bufferedCharacter = None
        elif not data:
            # We have no more data, bye-bye stream
            return False

        if len(data) > 1:
            lastv = ord(data[-1])
            if lastv == 0x0D or 0xD800