from test import support
from test.support import bigmemtest, _4G

import unittest
from io import BytesIO, DEFAULT_BUFFER_SIZE
import os
import pickle
import glob
import tempfile
import pathlib
import random
import shutil
import subprocess
import threading
from test.support import unlink
import _compression
import sys


# Skip tests if the bz2 module doesn't exist.
bz2 = support.import_module('bz2')
from bz2 import BZ2File, BZ2Compressor, BZ2Decompressor

has_cmdline_bunzip2 = None

def ext_decompress(data):
    global has_cmdline_bunzip2
    if has_cmdline_bunzip2 is None:
        has_cmdline_bunzip2 = bool(shutil.which('bunzip2'))
    if has_cmdline_bunzip2:
        return subprocess.check_output(['bunzip2'], input=data)
    else:
        return bz2.decompress(data)

class BaseTest(unittest.TestCase):
    "Base for other testcases."

    TEXT_LINES = [
        b'root:x:0:0:root:/root:/bin/bash\n',
        b'bin:x:1:1:bin:/bin:\n',
        b'daemon:x:2:2:daemon:/sbin:\n',
        b'adm:x:3:4:adm:/var/adm:\n',
        b'lp:x:4:7:lp:/var/spool/lpd:\n',
        b'sync:x:5:0:sync:/sbin:/bin/sync\n',
        b'shutdown:x:6:0:shutdown:/sbin:/sbin/shutdown\n',
        b'halt:x:7:0:halt:/sbin:/sbin/halt\n',
        b'mail:x:8:12:mail:/var/spool/mail:\n',
        b'news:x:9:13:news:/var/spool/news:\n',
        b'uucp:x:10:14:uucp:/var/spool/uucp:\n',
        b'operator:x:11:0:operator:/root:\n',
        b'games:x:12:100:games:/usr/games:\n',
        b'gopher:x:13:30:gopher:/usr/lib/gopher-data:\n',
        b'ftp:x:14:50:FTP User:/var/ftp:/bin/bash\n',
        b'nobody:x:65534:65534:Nobody:/home:\n',
        b'postfix:x:100:101:postfix:/var/spool/postfix:\n',
        b'niemeyer:x:500:500::/home/niemeyer:/bin/bash\n',
        b'postgres:x:101:102:PostgreSQL Server:/var/lib/pgsql:/bin/bash\n',
        b'mysql:x:102:103:MySQL server:/var/lib/mysql:/bin/bash\n',
        b'www:x:103:104::/var/www:/bin/false\n',
        ]
    TEXT = b''.join(TEXT_LINES)
    DATA = b'BZh91AY&SY.\xc8N\x18\x00\x01>_\x80\x00\x10@\x02\xff\xf0\x01\x07n\x00?\xe7\xff\xe00\x01\x99\xaa\x00\xc0\x03F\x86\x8c#&\x83F\x9a\x03\x06\xa6\xd0\xa6\x93M\x0fQ\xa7\xa8\x06\x804hh\x12$\x11\xa4i4\xf14S\xd2<Q\xb5\x0fH\xd3\xd4\xdd\xd5\x87\xbb\xf8\x94\r\x8f\xafI\x12\xe1\xc9\xf8/E\x00pu\x89\x12]\xc9\xbbDL\nQ\x0e\t1\x12\xdf\xa0\xc0\x97\xac2O9\x89\x13\x94\x0e\x1c7\x0ed\x95I\x0c\xaaJ\xa4\x18L\x10\x05#\x9c\xaf\xba\xbc/\x97\x8a#C\xc8\xe1\x8cW\xf9\xe2\xd0\xd6M\xa7\x8bXa<e\x84t\xcbL\xb3\xa7\xd9\xcd\xd1\xcb\x84.\xaf\xb3\xab\xab\xad`n}\xa0lh\tE,\x8eZ\x15\x17VH>\x88\xe5\xcd9gd6\x0b\n\xe9\x9b\xd5\x8a\x99\xf7\x08.K\x8ev\xfb\xf7xw\xbb\xdf\xa1\x92\xf1\xdd|/";\xa2\xba\x9f\xd5\xb1#A\xb6\xf6\xb3o\xc9\xc5y\\\xebO\xe7\x85\x9a\xbc\xb6f8\x952\xd5\xd7"%\x89>V,\xf7\xa6z\xe2\x9f\xa3\xdf\x11\x11"\xd6E)I\xa9\x13^\xca\xf3r\xd0\x03U\x922\xf26\xec\xb6\xed\x8b\xc3U\x13\x9d\xc5\x170\xa4\xfa^\x92\xacDF\x8a\x97\xd6\x19\xfe\xdd\xb8\xbd\x1a\x9a\x19\xa3\x80ankR\x8b\xe5\xd83]\xa9\xc6\x08\x82f\xf6\xb9"6l$\xb8j@\xc0\x8a\xb0l1..\xbak\x83ls\x15\xbc\xf4\xc1\x13\xbe\xf8E\xb8\x9d\r\xa8\x9dk\x84\xd3n\xfa\xacQ\x07\xb1%y\xaav\xb4\x08\xe0z\x1b\x16\xf5\x04\xe9\xcc\xb9\x08z\x1en7.G\xfc]\xc9\x14\xe1B@\xbb!8`'
    EMPTY_DATA = b'BZh9\x17rE8P\x90\x00\x00\x00\x00'
    BAD_DATA = b'this is not a valid bzip2 file'

    # Some tests need more than one block of uncompressed data. Since one block
    # is at least 100,000 bytes, we gather some data dynamically and compress it.
    # Note that this assumes that compression works correctly, so we cannot
    # simply use the bigger test data for all tests.
    test_size = 0
    BIG_TEXT = bytearray(128*1024)
    for fname in glob.glob(os.path.join(glob.escape(os.path.dirname(__file__)), '*.py')):
        with open(fname, 'rb') as fh:
            test_size += fh.readinto(memoryview(BIG_TEXT)[test_size:])
        if test_size > 128*1024:
            break
    BIG_DATA = bz2.compress(BIG_TEXT, compresslevel=1)

    def setUp(self):
        fd, self.filename = tempfile.mkstemp()
        os.close(fd)

    def tearDown(self):
        unlink(self.filename)


class BZ2FileTest(BaseTest):
    "Test the BZ2File class."

    def createTempFile(self, streams=1, suffix=b""):
        with open(self.filename, "wb") as f:
            f.write(self.DATA * streams)
            f.write(suffix)

    def testBadArgs(self):
        self.assertRaises(TypeError, BZ2File, 123.456)
        self.assertRaises(ValueError, BZ2File, os.devnull, "z")
        self.assertRaises(ValueError, BZ2File, os.devnull, "rx")
        self.assertRaises(ValueError, BZ2File, os.devnull, "rbt")
        self.assertRaises(ValueError, BZ2File, os.devnull, compresslevel=0)
        self.assertRaises(ValueError, BZ2File, os.devnull, compresslevel=10)

    def testRead(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.read, float())
            self.assertEqual(bz2f.read(), self.TEXT)

    def testReadBadFile(self):
        self.createTempFile(streams=0, suffix=self.BAD_DATA)
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(OSError, bz2f.read)

    def testReadMultiStream(self):
        self.createTempFile(streams=5)
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.read, float())
            self.assertEqual(bz2f.read(), self.TEXT * 5)

    def testReadMonkeyMultiStream(self):
        # Test BZ2File.read() on a multi-stream archive where a stream
        # boundary coincides with the end of the raw read buffer.
        buffer_size = _compression.BUFFER_SIZE
        _compression.BUFFER_SIZE = len(self.DATA)
        try:
            self.createTempFile(streams=5)
            with BZ2File(self.filename) as bz2f:
                self.assertRaises(TypeError, bz2f.read, float())
                self.assertEqual(bz2f.read(), self.TEXT * 5)
        finally:
            _compression.BUFFER_SIZE = buffer_size

    def testReadTrailingJunk(self):
        self.createTempFile(suffix=self.BAD_DATA)
        with BZ2File(self.filename) as bz2f:
            self.assertEqual(bz2f.read(), self.TEXT)

    def testReadMultiStreamTrailingJunk(self):
        self.createTempFile(streams=5, suffix=self.BAD_DATA)
        with BZ2File(self.filename) as bz2f:
            self.assertEqual(bz2f.read(), self.TEXT * 5)

    def testRead0(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.read, float())
            self.assertEqual(bz2f.read(0), b"")

    def testReadChunk10(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            text = b''
            while True:
                str = bz2f.read(10)
                if not str:
                    break
                text += str
            self.assertEqual(text, self.TEXT)

    def testReadChunk10MultiStream(self):
        self.createTempFile(streams=5)
        with BZ2File(self.filename) as bz2f:
            text = b''
            while True:
                str = bz2f.read(10)
                if not str:
                    break
                text += str
            self.assertEqual(text, self.TEXT * 5)

    def testRead100(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertEqual(bz2f.read(100), self.TEXT[:100])

    def testPeek(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            pdata = bz2f.peek()
            self.assertNotEqual(len(pdata), 0)
            self.assertTrue(self.TEXT.startswith(pdata))
            self.assertEqual(bz2f.read(), self.TEXT)

    def testReadInto(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            n = 128
            b = bytearray(n)
            self.assertEqual(bz2f.readinto(b), n)
            self.assertEqual(b, self.TEXT[:n])
            n = len(self.TEXT) - n
            b = bytearray(len(self.TEXT))
            self.assertEqual(bz2f.readinto(b), n)
            self.assertEqual(b[:n], self.TEXT[-n:])

    def testReadLine(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.readline, None)
            for line in self.TEXT_LINES:
                self.assertEqual(bz2f.readline(), line)

    def testReadLineMultiStream(self):
        self.createTempFile(streams=5)
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.readline, None)
            for line in self.TEXT_LINES * 5:
                self.assertEqual(bz2f.readline(), line)

    def testReadLines(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.readlines, None)
            self.assertEqual(bz2f.readlines(), self.TEXT_LINES)

    def testReadLinesMultiStream(self):
        self.createTempFile(streams=5)
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.readlines, None)
            self.assertEqual(bz2f.readlines(), self.TEXT_LINES * 5)

    def testIterator(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertEqual(list(iter(bz2f)), self.TEXT_LINES)

    def testIteratorMultiStream(self):
        self.createTempFile(streams=5)
        with BZ2File(self.filename) as bz2f:
            self.assertEqual(list(iter(bz2f)), self.TEXT_LINES * 5)

    def testClosedIteratorDeadlock(self):
        # Issue #3309: Iteration on a closed BZ2File should release the lock.
        self.createTempFile()
        bz2f = BZ2File(self.filename)
        bz2f.close()
        self.assertRaises(ValueError, next, bz2f)
        # This call will deadlock if the above call failed to release the lock.
        self.assertRaises(ValueError, bz2f.readlines)

    def testWrite(self):
        with BZ2File(self.filename, "w") as bz2f:
            self.assertRaises(TypeError, bz2f.write)
            bz2f.write(self.TEXT)
        with open(self.filename, 'rb') as f:
            self.assertEqual(ext_decompress(f.read()), self.TEXT)

    def testWriteChunks10(self):
        with BZ2File(self.filename, "w") as bz2f:
            n = 0
            while True:
                str = self.TEXT[n*10:(n+1)*10]
                if not str:
                    break
                bz2f.write(str)
                n += 1
        with open(self.filename, 'rb') as f:
            self.assertEqual(ext_decompress(f.read()), self.TEXT)

    def testWriteNonDefaultCompressLevel(self):
        expected = bz2.compress(self.TEXT, compresslevel=5)
        with BZ2File(self.filename, "w", compresslevel=5) as bz2f:
            bz2f.write(self.TEXT)
        with open(self.filename, "rb") as f:
            self.assertEqual(f.read(), expected)

    def testWriteLines(self):
        with BZ2File(self.filename, "w") as bz2f:
            self.assertRaises(TypeError, bz2f.writelines)
            bz2f.writelines(self.TEXT_LINES)
        # Issue #1535500: Calling writelines() on a closed BZ2File
        # should raise an exception.
        self.assertRaises(ValueError, bz2f.writelines, ["a"])
        with open(self.filename, 'rb') as f:
            self.assertEqual(ext_decompress(f.read()), self.TEXT)

    def testWriteMethodsOnReadOnlyFile(self):
        with BZ2File(self.filename, "w") as bz2f:
            bz2f.write(b"abc")

        with BZ2File(self.filename, "r") as bz2f:
            self.assertRaises(OSError, bz2f.write, b"a")
            self.assertRaises(OSError, bz2f.writelines, [b"a"])

    def testAppend(self):
        with BZ2File(self.filename, "w") as bz2f:
            self.assertRaises(TypeError, bz2f.write)
            bz2f.write(self.TEXT)
        with BZ2File(self.filename, "a") as bz2f:
            self.assertRaises(TypeError, bz2f.write)
            bz2f.write(self.TEXT)
        with open(self.filename, 'rb') as f:
            self.assertEqual(ext_decompress(f.read()), self.TEXT * 2)

    def testSeekForward(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.seek)
            bz2f.seek(150)
            self.assertEqual(bz2f.read(), self.TEXT[150:])

    def testSeekForwardAcrossStreams(self):
        self.createTempFile(streams=2)
        with BZ2File(self.filename) as bz2f:
            self.assertRaises(TypeError, bz2f.seek)
            bz2f.seek(len(self.TEXT) + 150)
            self.assertEqual(bz2f.read(), self.TEXT[150:])

    def testSeekBackwards(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            bz2f.read(500)
            bz2f.seek(-150, 1)
            self.assertEqual(bz2f.read(), self.TEXT[500-150:])

    def testSeekBackwardsAcrossStreams(self):
        self.createTempFile(streams=2)
        with BZ2File(self.filename) as bz2f:
            readto = len(self.TEXT) + 100
            while readto > 0:
                readto -= len(bz2f.read(readto))
            bz2f.seek(-150, 1)
            self.assertEqual(bz2f.read(), self.TEXT[100-150:] + self.TEXT)

    def testSeekBackwardsFromEnd(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            bz2f.seek(-150, 2)
            self.assertEqual(bz2f.read(), self.TEXT[len(self.TEXT)-150:])

    def testSeekBackwardsFromEndAcrossStreams(self):
        self.createTempFile(streams=2)
        with BZ2File(self.filename) as bz2f:
            bz2f.seek(-1000, 2)
            self.assertEqual(bz2f.read(), (self.TEXT * 2)[-1000:])

    def testSeekPostEnd(self):
        self.createTempFile()
        with BZ2File(self.filename) as bz2f:
            bz2f.seek(150000)
            self.assertEqual(bz2f.tell(), len(self.TEXT))
            self.assertEqual(bz2f.read(), b"")

    def testSeekPostEndMultiStream(self):
        self.createTempFile(streams=5)