diff --git a/src/calibre/ebooks/lit/lzxd.py b/src/calibre/ebooks/lit/lzxd.py
new file mode 100644
index 0000000000..a09daf012b
--- /dev/null
+++ b/src/calibre/ebooks/lit/lzxd.py
@@ -0,0 +1,138 @@
+import copy
+
+# some constants defined by the LZX specification
+MIN_MATCH = 2
+MAX_MATCH = 257
+NUM_CHARS = 256
+BLOCKTYPE_INVALID = 0  # also blocktypes 4-7 invalid
+BLOCKTYPE_VERBATIM = 1
+BLOCKTYPE_ALIGNED = 2
+BLOCKTYPE_UNCOMPRESSED = 3
+PRETREE_NUM_ELEMENTS = 20
+ALIGNED_NUM_ELEMENTS = 8  # aligned offset tree #elements
+NUM_PRIMARY_LENGTHS = 7  # this one missing from spec!
+NUM_SECONDARY_LENGTHS = 249  # length tree #elements
+
+# LZX huffman defines: tweak tablebits as desired
+PRETREE_MAXSYMBOLS = LZX_PRETREE_NUM_ELEMENTS
+PRETREE_TABLEBITS = 6
+MAINTREE_MAXSYMBOLS = LZX_NUM_CHARS + 50*8
+MAINTREE_TABLEBITS = 12
+LENGTH_MAXSYMBOLS = LZX_NUM_SECONDARY_LENGTHS+1
+LENGTH_TABLEBITS = 12
+ALIGNED_MAXSYMBOLS = LZX_ALIGNED_NUM_ELEMENTS
+ALIGNED_TABLEBITS = 7
+LENTABLE_SAFETY = 64  # table decoding overruns are allowed
+
+FRAME_SIZE = 32768  # the size of a frame in LZX
+
+
+class BitReader(object):
+    def __init__(self, data):
+        self.data, self.pos, self.nbits = \
+            data + "\x00\x00\x00\x00", 0, len(data) * 8
+        
+    def peek(self, n):
+        r, g = 0, 0
+        while g < n:
+            r = (r << 8) | ord(self.data[(self.pos + g) >> 3])
+            g = g + 8 - ((self.pos + g) & 7)
+        return (r >> (g - n)) & ((1 << n) - 1)
+    
+    def remove(self, n):
+        self.pos += n
+        return self.pos <= self.nbits
+    
+    def left(self):
+        return self.nbits - self.pos
+
+    def read(self, n):
+        val = self.peek(n)
+        self.remove(n)
+        return val
+
+class LzxError(Exception):
+    pass
+
+POSITION_BASE = [0]*51
+EXTRA_BITS = [0]*51
+
+def _static_init():
+    j = 0
+    for i in xrange(0, 51, 2):
+        EXTRA_BITS[i] = j
+        EXTRA_BITS[i + 1] = j
+        if i != 0 or j < 17): j += 1
+    j = 0
+    for i in xrange(0, 51, 1):
+        POSITION_BASE[i] = j
+        j += 1 << extra_bits[i]
+_static_init()
+
+class LzxDecompressor(object):
+    def __init__(self, window_bits, reset_interval=0x7fff):
+        # LZX supports window sizes of 2^15 (32Kb) through 2^21 (2Mb)
+        if window_bits < 15 or window_bits > 21:
+            raise LzxError("Invalid window size")
+        
+        self.window_size = 1 << window_bits
+        self.window_posn = 0
+        self.frame_posn = 0
+        self.frame = 0
+        self.reset_interval = reset_interval
+        self.intel_filesize = 0
+        self.intel_curpos = 0
+        
+        # window bits:    15  16  17  18  19  20  21
+        # position slots: 30  32  34  36  38  42  50 
+        self.posn_solts = 50 if window_bits == 21 \
+            else 42 if window_bits == 20 else window_bits << 1
+        self.intel_started = 0
+        self.input_end = 0
+
+        # huffman code lengths
+        self.PRETREE_len = [0] * (PRETREE_MAXSYMBOLS + LENTABLE_SAFETY)
+        self.MAINTREE_len = [0] * (MAINTREE_MAXSYMBOLS + LENTABLE_SAFETY)
+        self.LENGTH_len = [0] * (LENGTH_MAXSYMBOLS + LENTABLE_SAFETY)
+        self.ALIGNED_len = [0] * (ALIGNED_MAXSYMBOLS + LENTABLE_SAFETY)
+
+        # huffman decoding tables
+        self.PRETREE_table = \
+            [0] * ((1 << PRETREE_TABLEBITS) + (PRETREE_MAXSYMBOLS * 2))
+        self.MAINTREE_table = \
+            [0] * ((1 << MAINTREE_TABLEBITS) + (MAINTREE_MAXSYMBOLS * 2))
+        self.LENGTH_table = \
+            [0] * ((1 << LENGTH_TABLEBITS) + (LENGTH_MAXSYMBOLS * 2))
+        self.ALIGNED_table = \
+            [0] * ((1 << ALIGNED_TABLEBITS) + (ALIGNED_MAXSYMBOLS * 2))
+
+        self.o_buf = self.i_buf = ''
+        
+        self._reset_state()
+
+    def _reset_state(self):
+        self.R0 = 1
+        self.R1 = 1
+        self.R2 = 1
+        self.header_read = 0
+        self.block_remaining = 0
+        self.block_type = BLOCKTYPE_INVALID
+
+        # initialise tables to 0 (because deltas will be applied to them)
+        for i in xrange(MAINTREE_MAXSYMBOLS): self.MAINTREE_len[i] = 0
+        for i in xrange(LENGTH_MAXSYMBOLS): self.LENGTH_len[i] = 0
+
+    def decompress(self, data, out_bytes):
+        return ''.join(self._decompress(data, out_bytes))
+        
+    def _decompress(self, data, out_bytes):
+        # easy answers
+        if out_bytes < 0:
+            raise LzxError('Negative desired output bytes')
+
+        # Initialize input and output
+        input = BitReader(data)
+        output = []
+        
+        
+