a
_h.m L @ s
d Z ddlZddlZddlZddlZddlZg dZejZdZdZ dZ
dZ dZ G d d
d
e
Zdd lmZ d d
ZeddeddZdd ZeddeddZdd ZeddeddZd d Zed ded dZd d! Zed"d#ed$dZdd&d'Z ed(ee d)dZ d*d+ Z ed,ee d-dZ d.d/ Z ed0ee d1dZ!d2d3 Z"ed4e e"d5dZ#d6d7 Z$ed8e
e$d9dZ%d:d; Z&edd? Z(ed@e e(dAdZ)dBdC Z*edDe e*dEdZ+dFdG Z,edHe e,dIdZ-dJdK Z.edLee.dMdZ/dNdO Z0edPe e0dQdZ1dRdS Z2edTe e2dUdZ3dVdW Z4edXe e4dYdZ5dZd[ Z6d\d] Z7ed^ee6d_dZ8ed`ee7dadZ9dbdc Z:eddee:dedZ;dfdg ZZ> dkdl Z?edme e?dndZ@dodp ZAedqe
eAdrdZBG dsdt dte
ZCeCdueDdvdw ZEZFeCdxeDeGfdydwZHeCdzeGd{dwZIeCd|eJd}dwZKeCd~eLeMfddw ZNZOeCdeLddwZPeCdeQddwZReCdeMddwZSeCdeTdddwZUeCdeVddwZWeCdeXddwZYeCdeZddwZ[eCde\ddwZ]eCde\ddwZ^eCde
ddwZ_eCde
ddwZ`eCdeCddwZaeCdeCddwZbG dd de
ZcecZdeddde8g eHgdddedddeg eEgdddedddeg eEgdddedddeg eEgdddeddde9g eEgdddeddde@g eEgdddedddeBg eEgdddeddde g eNgdddeddde%g eNgdddeddde#g eNgdddeddde)g ePgdddeddde'g ePgdddeddde+g ePgdddeddde-g eRgdddeddddg e_gdddedddde_ge_gdddeddddg eUgdddeddddg eIgdddeddddg eIgdddeddde/g eSgdddeddde1g eSgdddeddde3g eSgdddeddde5g eSgdddeddde;g eKgdddeddde=g eKgdddeddddg eYgdddeddddeYe`geYgdddeddddeYeaebgeYgdddeddddeaebgeYgdddeddddg eWgdddeddddeaebgeWgdddedddde`geWgdddeddd de`e`geWgdddedddde`e`e`geWgdddeddddg e[gdddeddd deaebge[gdd
dedd d de[e`e`ge[gdd
dedddde[eaebge[gdddeddddg e]gdddedddde]eaebge]gdddeddddeaebge^gdddedddde`gg dd dedd d de`ge`e`gdd dedd d!dg eagdd"dedd#d$deaebgg dd%dedd&d'e8g e`gdd(dedd)d*eg e`gdd+dedd,d-eg e`gdd.dedd/d0e8g g dd1dedd2d3eg g dd4dedd5d6eg g dd7dedd8d9de`ge`gdd:dedd;dd?eg e`gdd@deddAdBeg e`gddCdeddDdEe!g e`gddFdeddGdHdeSeSge`gddIdeddJdKde`e`ge`gddLdeddMdNde`e`ge`gddOdeddPdQe!eaebge`gddRdeddSdTdeae`ebge`gddUdeddVdWde`e`ge`gddXdeddYdZde`e`e`ge`gdd[dedd\d]eg g dd^dedd_d`de`gg ddadeddbdceg g ddddeddedfe g e`gddgdeddhdide`ge`gddjdgDZe[di Zfi ZgeheeD ]n\ZiZjejjkefv rZejjejkdZejed eseejeeje
qdS ( ar "Executable documentation" for the pickle module.
Extensive comments about the pickle protocols and pickle-machine opcodes
can be found here. Some functions meant for external use:
genops(pickle)
Generate all the opcodes in a pickle, as (opcode, arg, position) triples.
dis(pickle, out=None, memo=None, indentlevel=4)
Print a symbolic disassembly of a pickle.
N)disgenopsoptimizec @ s e Zd ZdZdd ZdS )ArgumentDescriptornamenreaderdocc C s` t |tsJ || _t |tr8|dks<|tttttfv s>> import io
>>> read_uint1(io.BytesIO(b'\xff'))
255
r z'not enough data in stream to read uint1N)read
ValueErrorfdatar r r
read_uint1 s
r) uint1r# zOne-byte unsigned integer.r c C s0 | d}t|dkr$td|d S tddS )z
>>> import io
>>> read_uint2(io.BytesIO(b'\xff\x00'))
255
>>> read_uint2(io.BytesIO(b'\xff\xff'))
65535
z>> import io
>>> read_int4(io.BytesIO(b'\xff\x00\x00\x00'))
255
>>> read_int4(io.BytesIO(b'\x00\x00\x00\x80')) == -(2**31)
True
z>> import io
>>> read_uint4(io.BytesIO(b'\xff\x00\x00\x00'))
255
>>> read_uint4(io.BytesIO(b'\x00\x00\x00\x80')) == 2**31
True
r1 z>> import io
>>> read_uint8(io.BytesIO(b'\xff\x00\x00\x00\x00\x00\x00\x00'))
255
>>> read_uint8(io.BytesIO(b'\xff' * 8)) == 2**64-1
True
z>> import io
>>> read_stringnl(io.BytesIO(b"'abcd'\nefg\n"))
'abcd'
>>> read_stringnl(io.BytesIO(b"\n"))
Traceback (most recent call last):
...
ValueError: no string quotes around b''
>>> read_stringnl(io.BytesIO(b"\n"), stripquotes=False)
''
>>> read_stringnl(io.BytesIO(b"''\n"))
''
>>> read_stringnl(io.BytesIO(b'"abcd"'))
Traceback (most recent call last):
...
ValueError: no newline found when trying to read stringnl
Embedded escapes are undone in the result.
>>> read_stringnl(io.BytesIO(br"'a\n\\b\x00c\td'" + b"\n'e'"))
'a\n\\b\x00c\td'
z-no newline found when trying to read stringnlNr ) " 'z,strinq quote %r not found at both ends of %rr# zno string quotes around %rr ascii)readlineendswithr%
startswithcodecs
escape_decodedecode)r' rB stripquotesr( qr r r
read_stringnl; s"
rE stringnlzA newline-terminated string.
This is a repr-style string, with embedded escapes, and
bracketing quotes.
c C s t | ddS )NF)rC )rE r' r r r read_stringnl_noescapet s rH stringnl_noescapea A newline-terminated string.
This is a str-style string, without embedded escapes,
or bracketing quotes. It should consist solely of
printable ASCII characters.
c C s dt | t | f S )zp
>>> import io
>>> read_stringnl_noescape_pair(io.BytesIO(b"Queue\nEmpty\njunk"))
'Queue Empty'
z%s %s)rH rG r r r read_stringnl_noescape_pair s rJ stringnl_noescape_paira A pair of newline-terminated strings.
These are str-style strings, without embedded
escapes, or bracketing quotes. They should
consist solely of printable ASCII characters.
The pair is returned as a single string, with
a single blank separating the two strings.
c C sL t | }|dksJ | |}t||kr4|dS td|t|f dS )z
>>> import io
>>> read_string1(io.BytesIO(b"\x00"))
''
>>> read_string1(io.BytesIO(b"\x03abcdef"))
'abc'
r latin-1z2expected %d bytes in a string1, but only %d remainN)r) r$ r- rB r% r' r
r( r r r read_string1 s
rN string1zA counted string.
The first argument is a 1-byte unsigned int giving the number
of bytes in the string, and the second argument is that many
bytes.
c C sT t | }|dk r td| | |}t||kr<|dS td|t|f dS )aP
>>> import io
>>> read_string4(io.BytesIO(b"\x00\x00\x00\x00abc"))
''
>>> read_string4(io.BytesIO(b"\x03\x00\x00\x00abcdef"))
'abc'
>>> read_string4(io.BytesIO(b"\x00\x00\x00\x03abcdef"))
Traceback (most recent call last):
...
ValueError: expected 50331648 bytes in a string4, but only 6 remain
r zstring4 byte count < 0: %drL z2expected %d bytes in a string4, but only %d remainN)r2 r% r$ r- rB rM r r r read_string4 s
rP string4zA counted string.
The first argument is a 4-byte little-endian signed int giving
the number of bytes in the string, and the second argument is
that many bytes.
c C sF t | }|dksJ | |}t||kr.|S td|t|f dS )z
>>> import io
>>> read_bytes1(io.BytesIO(b"\x00"))
b''
>>> read_bytes1(io.BytesIO(b"\x03abcdef"))
b'abc'
r z1expected %d bytes in a bytes1, but only %d remainN)r) r$ r- r% rM r r r read_bytes1 s
rR bytes1zA counted bytes string.
The first argument is a 1-byte unsigned int giving the number
of bytes, and the second argument is that many bytes.
c C s\ t | }|dksJ |tjkr*td| | |}t||krD|S td|t|f dS )aN
>>> import io
>>> read_bytes4(io.BytesIO(b"\x00\x00\x00\x00abc"))
b''
>>> read_bytes4(io.BytesIO(b"\x03\x00\x00\x00abcdef"))
b'abc'
>>> read_bytes4(io.BytesIO(b"\x00\x00\x00\x03abcdef"))
Traceback (most recent call last):
...
ValueError: expected 50331648 bytes in a bytes4, but only 6 remain
r z#bytes4 byte count > sys.maxsize: %dz1expected %d bytes in a bytes4, but only %d remainN)r4 sysmaxsizer% r$ r- rM r r r read_bytes4 s
rV bytes4zA counted bytes string.
The first argument is a 4-byte little-endian unsigned int giving
the number of bytes, and the second argument is that many bytes.
c C s\ t | }|dksJ |tjkr*td| | |}t||krD|S td|t|f dS )a
>>> import io, struct, sys
>>> read_bytes8(io.BytesIO(b"\x00\x00\x00\x00\x00\x00\x00\x00abc"))
b''
>>> read_bytes8(io.BytesIO(b"\x03\x00\x00\x00\x00\x00\x00\x00abcdef"))
b'abc'
>>> bigsize8 = struct.pack(">> read_bytes8(io.BytesIO(bigsize8 + b"abcdef")) #doctest: +ELLIPSIS
Traceback (most recent call last):
...
ValueError: expected ... bytes in a bytes8, but only 6 remain
r z#bytes8 byte count > sys.maxsize: %dz1expected %d bytes in a bytes8, but only %d remainN)r7 rT rU r% r$ r- rM r r r read_bytes8 s
rX bytes8zA counted bytes string.
The first argument is an 8-byte little-endian unsigned int giving
the number of bytes, and the second argument is that many bytes.
c C s` t | }|dksJ |tjkr*td| | |}t||krHt|S td|t|f dS )a
>>> import io, struct, sys
>>> read_bytearray8(io.BytesIO(b"\x00\x00\x00\x00\x00\x00\x00\x00abc"))
bytearray(b'')
>>> read_bytearray8(io.BytesIO(b"\x03\x00\x00\x00\x00\x00\x00\x00abcdef"))
bytearray(b'abc')
>>> bigsize8 = struct.pack(">> read_bytearray8(io.BytesIO(bigsize8 + b"abcdef")) #doctest: +ELLIPSIS
Traceback (most recent call last):
...
ValueError: expected ... bytes in a bytearray8, but only 6 remain
r z'bytearray8 byte count > sys.maxsize: %dz5expected %d bytes in a bytearray8, but only %d remainN)r7 rT rU r% r$ r- bytearrayrM r r r read_bytearray89 s
r[
bytearray8zA counted bytearray.
The first argument is an 8-byte little-endian unsigned int giving
the number of bytes, and the second argument is that many bytes.
c C s0 | }|dstd|dd }t|dS )zm
>>> import io
>>> read_unicodestringnl(io.BytesIO(b"abc\\uabcd\njunk")) == 'abc\uabcd'
True
r9 z4no newline found when trying to read unicodestringnlNr zraw-unicode-escape)r= r> r% r r& r r r read_unicodestringnl[ s
r] unicodestringnlzA newline-terminated Unicode string.
This is raw-unicode-escape encoded, so consists of
printable ASCII characters, and may contain embedded
escape sequences.
c C sN t | }|dksJ | |}t||kr6t|ddS td|t|f dS )a
>>> import io
>>> s = 'abcd\uabcd'
>>> enc = s.encode('utf-8')
>>> enc
b'abcd\xea\xaf\x8d'
>>> n = bytes([len(enc)]) # little-endian 1-byte length
>>> t = read_unicodestring1(io.BytesIO(n + enc + b'junk'))
>>> s == t
True
>>> read_unicodestring1(io.BytesIO(n + enc[:-1]))
Traceback (most recent call last):
...
ValueError: expected 7 bytes in a unicodestring1, but only 6 remain
r utf-8
surrogatepassz9expected %d bytes in a unicodestring1, but only %d remainN)r) r$ r- r r% rM r r r read_unicodestring1u s
ra unicodestring1aA A counted Unicode string.
The first argument is a 1-byte little-endian signed int
giving the number of bytes in the string, and the second
argument-- the UTF-8 encoding of the Unicode string --
contains that many bytes.
c C sd t | }|dksJ |tjkr*td| | |}t||krLt|ddS td|t|f dS )a
>>> import io
>>> s = 'abcd\uabcd'
>>> enc = s.encode('utf-8')
>>> enc
b'abcd\xea\xaf\x8d'
>>> n = bytes([len(enc), 0, 0, 0]) # little-endian 4-byte length
>>> t = read_unicodestring4(io.BytesIO(n + enc + b'junk'))
>>> s == t
True
>>> read_unicodestring4(io.BytesIO(n + enc[:-1]))
Traceback (most recent call last):
...
ValueError: expected 7 bytes in a unicodestring4, but only 6 remain
r z+unicodestring4 byte count > sys.maxsize: %dr_ r` z9expected %d bytes in a unicodestring4, but only %d remainN)r4 rT rU r% r$ r- r rM r r r read_unicodestring4 s
rc unicodestring4aA A counted Unicode string.
The first argument is a 4-byte little-endian signed int
giving the number of bytes in the string, and the second
argument-- the UTF-8 encoding of the Unicode string --
contains that many bytes.
c C sd t | }|dksJ |tjkr*td| | |}t||krLt|ddS td|t|f dS )a
>>> import io
>>> s = 'abcd\uabcd'
>>> enc = s.encode('utf-8')
>>> enc
b'abcd\xea\xaf\x8d'
>>> n = bytes([len(enc)]) + b'\0' * 7 # little-endian 8-byte length
>>> t = read_unicodestring8(io.BytesIO(n + enc + b'junk'))
>>> s == t
True
>>> read_unicodestring8(io.BytesIO(n + enc[:-1]))
Traceback (most recent call last):
...
ValueError: expected 7 bytes in a unicodestring8, but only 6 remain
r z+unicodestring8 byte count > sys.maxsize: %dr_ r` z9expected %d bytes in a unicodestring8, but only %d remainN)r7 rT rU r% r$ r- r rM r r r read_unicodestring8 s
re unicodestring8aB A counted Unicode string.
The first argument is an 8-byte little-endian signed int
giving the number of bytes in the string, and the second
argument-- the UTF-8 encoding of the Unicode string --
contains that many bytes.
c C s. t | ddd}|dkrdS |dkr&dS t|S )z
>>> import io
>>> read_decimalnl_short(io.BytesIO(b"1234\n56"))
1234
>>> read_decimalnl_short(io.BytesIO(b"1234L\n56"))
Traceback (most recent call last):
...
ValueError: invalid literal for int() with base 10: b'1234L'
FrB rC s 00s 01TrE r r' sr r r read_decimalnl_short s rk c C s2 t | ddd}|dd dkr*|dd }t|S )z
>>> import io
>>> read_decimalnl_long(io.BytesIO(b"1234L\n56"))
1234
>>> read_decimalnl_long(io.BytesIO(b"123456789012345678901234L\n6"))
123456789012345678901234
Frg r N Lrh ri r r r read_decimalnl_long s rm decimalnl_shorta A newline-terminated decimal integer literal.
This never has a trailing 'L', and the integer fit
in a short Python int on the box where the pickle
was written -- but there's no guarantee it will fit
in a short Python int on the box where the pickle
is read.
decimalnl_longzA newline-terminated decimal integer literal.
This has a trailing 'L', and can represent integers
of any size.
c C s t | ddd}t|S )zO
>>> import io
>>> read_floatnl(io.BytesIO(b"-1.25\n6"))
-1.25
Frg )rE floatri r r r read_floatnl- s rq floatnla A newline-terminated decimal floating literal.
In general this requires 17 significant digits for roundtrip
identity, and pickling then unpickling infinities, NaNs, and
minus zero doesn't work across boxes, or on some boxes even
on itself (e.g., Windows can't read the strings it produces
for infinities or NaNs).
c C s0 | d}t|dkr$td|d S tddS )z
>>> import io, struct
>>> raw = struct.pack(">d", -1.25)
>>> raw
b'\xbf\xf4\x00\x00\x00\x00\x00\x00'
>>> read_float8(io.BytesIO(raw + b"\n"))
-1.25
r6 z>dr z(not enough data in stream to read float8Nr, r& r r r read_float8C s
rs float8a An 8-byte binary representation of a float, big-endian.
The format is unique to Python, and shared with the struct
module (format string '>d') "in theory" (the struct and pickle
implementations don't share the code -- they should). It's
strongly related to the IEEE-754 double format, and, in normal
cases, is in fact identical to the big-endian 754 double format.
On other boxes the dynamic range is limited to that of a 754
double, and "add a half and chop" rounding is used to reduce
the precision to 53 bits. However, even on a 754 box,
infinities, NaNs, and minus zero may not be handled correctly
(may not survive roundtrip pickling intact).
) decode_longc C s. t | }| |}t||kr&tdt|S )a+
>>> import io
>>> read_long1(io.BytesIO(b"\x00"))
0
>>> read_long1(io.BytesIO(b"\x02\xff\x00"))
255
>>> read_long1(io.BytesIO(b"\x02\xff\x7f"))
32767
>>> read_long1(io.BytesIO(b"\x02\x00\xff"))
-256
>>> read_long1(io.BytesIO(b"\x02\x00\x80"))
-32768
z'not enough data in stream to read long1)r) r$ r- r% ru rM r r r
read_long1i s
rv long1a A binary long, little-endian, using 1-byte size.
This first reads one byte as an unsigned size, then reads that
many bytes and interprets them as a little-endian 2's-complement long.
If the size is 0, that's taken as a shortcut for the long 0L.
c C sB t | }|dk r td| | |}t||kr:tdt|S )ag
>>> import io
>>> read_long4(io.BytesIO(b"\x02\x00\x00\x00\xff\x00"))
255
>>> read_long4(io.BytesIO(b"\x02\x00\x00\x00\xff\x7f"))
32767
>>> read_long4(io.BytesIO(b"\x02\x00\x00\x00\x00\xff"))
-256
>>> read_long4(io.BytesIO(b"\x02\x00\x00\x00\x00\x80"))
-32768
>>> read_long1(io.BytesIO(b"\x00\x00\x00\x00"))
0
r zlong4 byte count < 0: %dz'not enough data in stream to read long4)r2 r% r$ r- ru rM r r r
read_long4 s
rx long4a A binary representation of a long, little-endian.
This first reads four bytes as a signed size (but requires the
size to be >= 0), then reads that many bytes and interprets them
as a little-endian 2's-complement long. If the size is 0, that's taken
as a shortcut for the int 0, although LONG1 should really be used
then instead (and in any case where # of bytes < 256).
c @ s e Zd ZdZdd Zdd ZdS ) StackObjectr obtyper c C sl t |tsJ || _t |ts,t |ts,J t |trN|D ]}t |ts:J q:|| _t |tsbJ || _d S N)r r r typetupler| r )r r r| r Z containedr r r r s
zStackObject.__init__c C s | j S r} r )r r r r __repr__ s zStackObject.__repr__N)r r r r! r r r r r r rz s
rz r zA Python integer object.r{ Z int_or_boolz#A Python integer or boolean object.boolzA Python boolean object.rp zA Python float object.Z bytes_or_strz*A Python bytes or (Unicode) string object.byteszA Python bytes object.rZ zA Python bytearray object.r z!A Python (Unicode) string object.NonezThe Python None object.r zA Python tuple object.listzA Python list object.dictzA Python dict object.setzA Python set object. frozensetzA Python frozenset object.bufferz A Python buffer-like object.anyz Any kind of object whatsoever.Zmarkaz 'The mark' is a unique object.
Opcodes that operate on a variable number of objects
generally don't embed the count of objects in the opcode,
or pull it off the stack. Instead the MARK opcode is used
to push a special marker object on the stack, and then
some other opcodes grab all the objects from the top of
the stack down to (but not including) the topmost marker
object.
stackslicea An object representing a contiguous slice of the stack.
This is used in conjunction with markobject, to represent all
of the stack following the topmost markobject. For example,
the POP_MARK opcode changes the stack from
[..., markobject, stackslice]
to
[...]
No matter how many object are on the stack after the topmost
markobject, POP_MARK gets rid of all of them (including the
topmost markobject too).
c @ s e Zd ZdZdd ZdS )
OpcodeInfor codearg stack_before stack_afterprotor c C s t |tsJ || _t |ts"J t|dks2J || _|d u sNt |tsNJ || _t |tsbJ |D ]}t |tsfJ qf|| _ t |tsJ |D ]}t |tsJ q|| _
t |t rd| krt j
ksn J || _t |tsJ || _d S )Nr# r )r r r r- r r
r r rz r r r pickleHIGHEST_PROTOCOLr r ) r r r r r r r r xr r r r d s& &zOpcodeInfo.__init__Nr r r r r r E s r INTIa Push an integer or bool.
The argument is a newline-terminated decimal literal string.
The intent may have been that this always fit in a short Python int,
but INT can be generated in pickles written on a 64-bit box that
require a Python long on a 32-bit box. The difference between this
and LONG then is that INT skips a trailing 'L', and produces a short
int whenever possible.
Another difference is due to that, when bool was introduced as a
distinct type in 2.3, builtin names True and False were also added to
2.2.2, mapping to ints 1 and 0. For compatibility in both directions,
True gets pickled as INT + "I01\n", and False as INT + "I00\n".
Leading zeroes are never produced for a genuine integer. The 2.3
(and later) unpicklers special-case these and return bool instead;
earlier unpicklers ignore the leading "0" and return the int.
r BININTJa1 Push a four-byte signed integer.
This handles the full range of Python (short) integers on a 32-bit
box, directly as binary bytes (1 for the opcode and 4 for the integer).
If the integer is non-negative and fits in 1 or 2 bytes, pickling via
BININT1 or BININT2 saves space.
BININT1KzPush a one-byte unsigned integer.
This is a space optimization for pickling very small non-negative ints,
in range(256).
BININT2MzPush a two-byte unsigned integer.
This is a space optimization for pickling small positive ints, in
range(256, 2**16). Integers in range(256) can also be pickled via
BININT2, but BININT1 instead saves a byte.
LONGLa Push a long integer.
The same as INT, except that the literal ends with 'L', and always
unpickles to a Python long. There doesn't seem a real purpose to the
trailing 'L'.
Note that LONG takes time quadratic in the number of digits when
unpickling (this is simply due to the nature of decimal->binary
conversion). Proto 2 added linear-time (in C; still quadratic-time
in Python) LONG1 and LONG4 opcodes.
LONG1 z|Long integer using one-byte length.
A more efficient encoding of a Python long; the long1 encoding
says it all.LONG4 z~Long integer using found-byte length.
A more efficient encoding of a Python long; the long4 encoding
says it all.STRINGSa Push a Python string object.
The argument is a repr-style string, with bracketing quote characters,
and perhaps embedded escapes. The argument extends until the next
newline character. These are usually decoded into a str instance
using the encoding given to the Unpickler constructor. or the default,
'ASCII'. If the encoding given was 'bytes' however, they will be
decoded as bytes object instead.
BINSTRINGTa Push a Python string object.
There are two arguments: the first is a 4-byte little-endian
signed int giving the number of bytes in the string, and the
second is that many bytes, which are taken literally as the string
content. These are usually decoded into a str instance using the
encoding given to the Unpickler constructor. or the default,
'ASCII'. If the encoding given was 'bytes' however, they will be
decoded as bytes object instead.
SHORT_BINSTRINGUa Push a Python string object.
There are two arguments: the first is a 1-byte unsigned int giving
the number of bytes in the string, and the second is that many
bytes, which are taken literally as the string content. These are
usually decoded into a str instance using the encoding given to
the Unpickler constructor. or the default, 'ASCII'. If the
encoding given was 'bytes' however, they will be decoded as bytes
object instead.
BINBYTESB zPush a Python bytes object.
There are two arguments: the first is a 4-byte little-endian unsigned int
giving the number of bytes, and the second is that many bytes, which are
taken literally as the bytes content.
SHORT_BINBYTESCzPush a Python bytes object.
There are two arguments: the first is a 1-byte unsigned int giving
the number of bytes, and the second is that many bytes, which are taken
literally as the string content.
BINBYTES8 zPush a Python bytes object.
There are two arguments: the first is an 8-byte unsigned int giving
the number of bytes in the string, and the second is that many bytes,
which are taken literally as the string content.
BYTEARRAY8 zPush a Python bytearray object.
There are two arguments: the first is an 8-byte unsigned int giving
the number of bytes in the bytearray, and the second is that many bytes,
which are taken literally as the bytearray content.
NEXT_BUFFER z"Push an out-of-band buffer object.READONLY_BUFFER z,Make an out-of-band buffer object read-only.NONENzPush None on the stack.NEWTRUE zPush True onto the stack.NEWFALSE zPush False onto the stack.UNICODEVzPush a Python Unicode string object.
The argument is a raw-unicode-escape encoding of a Unicode string,
and so may contain embedded escape sequences. The argument extends
until the next newline character.
SHORT_BINUNICODE a Push a Python Unicode string object.
There are two arguments: the first is a 1-byte little-endian signed int
giving the number of bytes in the string. The second is that many
bytes, and is the UTF-8 encoding of the Unicode string.
BINUNICODEXa Push a Python Unicode string object.
There are two arguments: the first is a 4-byte little-endian unsigned int
giving the number of bytes in the string. The second is that many
bytes, and is the UTF-8 encoding of the Unicode string.
BINUNICODE8 a Push a Python Unicode string object.
There are two arguments: the first is an 8-byte little-endian signed int
giving the number of bytes in the string. The second is that many
bytes, and is the UTF-8 encoding of the Unicode string.
FLOATFa Newline-terminated decimal float literal.
The argument is repr(a_float), and in general requires 17 significant
digits for roundtrip conversion to be an identity (this is so for
IEEE-754 double precision values, which is what Python float maps to
on most boxes).
In general, FLOAT cannot be used to transport infinities, NaNs, or
minus zero across boxes (or even on a single box, if the platform C
library can't read the strings it produces for such things -- Windows
is like that), but may do less damage than BINFLOAT on boxes with
greater precision or dynamic range than IEEE-754 double.
BINFLOATGa Float stored in binary form, with 8 bytes of data.
This generally requires less than half the space of FLOAT encoding.
In general, BINFLOAT cannot be used to transport infinities, NaNs, or
minus zero, raises an exception if the exponent exceeds the range of
an IEEE-754 double, and retains no more than 53 bits of precision (if
there are more than that, "add a half and chop" rounding is used to
cut it back to 53 significant bits).
EMPTY_LIST]zPush an empty list.APPENDazAppend an object to a list.
Stack before: ... pylist anyobject
Stack after: ... pylist+[anyobject]
although pylist is really extended in-place.
APPENDSezExtend a list by a slice of stack objects.
Stack before: ... pylist markobject stackslice
Stack after: ... pylist+stackslice
although pylist is really extended in-place.
LISTlas Build a list out of the topmost stack slice, after markobject.
All the stack entries following the topmost markobject are placed into
a single Python list, which single list object replaces all of the
stack from the topmost markobject onward. For example,
Stack before: ... markobject 1 2 3 'abc'
Stack after: ... [1, 2, 3, 'abc']
EMPTY_TUPLE)zPush an empty tuple.TUPLEtav Build a tuple out of the topmost stack slice, after markobject.
All the stack entries following the topmost markobject are placed into
a single Python tuple, which single tuple object replaces all of the
stack from the topmost markobject onward. For example,
Stack before: ... markobject 1 2 3 'abc'
Stack after: ... (1, 2, 3, 'abc')
TUPLE1 zBuild a one-tuple out of the topmost item on the stack.
This code pops one value off the stack and pushes a tuple of
length 1 whose one item is that value back onto it. In other
words:
stack[-1] = tuple(stack[-1:])
TUPLE2 a Build a two-tuple out of the top two items on the stack.
This code pops two values off the stack and pushes a tuple of
length 2 whose items are those values back onto it. In other
words:
stack[-2:] = [tuple(stack[-2:])]
TUPLE3 a Build a three-tuple out of the top three items on the stack.
This code pops three values off the stack and pushes a tuple of
length 3 whose items are those values back onto it. In other
words:
stack[-3:] = [tuple(stack[-3:])]
EMPTY_DICT}zPush an empty dict.DICTda Build a dict out of the topmost stack slice, after markobject.
All the stack entries following the topmost markobject are placed into
a single Python dict, which single dict object replaces all of the
stack from the topmost markobject onward. The stack slice alternates
key, value, key, value, .... For example,
Stack before: ... markobject 1 2 3 'abc'
Stack after: ... {1: 2, 3: 'abc'}
SETITEMrj zAdd a key+value pair to an existing dict.
Stack before: ... pydict key value
Stack after: ... pydict
where pydict has been modified via pydict[key] = value.
SETITEMSua\ Add an arbitrary number of key+value pairs to an existing dict.
The slice of the stack following the topmost markobject is taken as
an alternating sequence of keys and values, added to the dict
immediately under the topmost markobject. Everything at and after the
topmost markobject is popped, leaving the mutated dict at the top
of the stack.
Stack before: ... pydict markobject key_1 value_1 ... key_n value_n
Stack after: ... pydict
where pydict has been modified via pydict[key_i] = value_i for i in
1, 2, ..., n, and in that order.
EMPTY_SET zPush an empty set.ADDITEMS a$ Add an arbitrary number of items to an existing set.
The slice of the stack following the topmost markobject is taken as
a sequence of items, added to the set immediately under the topmost
markobject. Everything at and after the topmost markobject is popped,
leaving the mutated set at the top of the stack.
Stack before: ... pyset markobject item_1 ... item_n
Stack after: ... pyset
where pyset has been modified via pyset.add(item_i) = item_i for i in
1, 2, ..., n, and in that order.
FROZENSET az Build a frozenset out of the topmost slice, after markobject.
All the stack entries following the topmost markobject are placed into
a single Python frozenset, which single frozenset object replaces all
of the stack from the topmost markobject onward. For example,
Stack before: ... markobject 1 2 3
Stack after: ... frozenset({1, 2, 3})
POP0z.
BUILDba Finish building an object, via __setstate__ or dict update.
Stack before: ... anyobject argument
Stack after: ... anyobject
where anyobject may have been mutated, as follows:
If the object has a __setstate__ method,
anyobject.__setstate__(argument)
is called.
Else the argument must be a dict, the object must have a __dict__, and
the object is updated via
anyobject.__dict__.update(argument)
INSTiaq Build a class instance.
This is the protocol 0 version of protocol 1's OBJ opcode.
INST is followed by two newline-terminated strings, giving a
module and class name, just as for the GLOBAL opcode (and see
GLOBAL for more details about that). self.find_class(module, name)
is used to get a class object.
In addition, all the objects on the stack following the topmost
markobject are gathered into a tuple and popped (along with the
topmost markobject), just as for the TUPLE opcode.
Now it gets complicated. If all of these are true:
+ The argtuple is empty (markobject was at the top of the stack
at the start).
+ The class object does not have a __getinitargs__ attribute.
then we want to create an old-style class instance without invoking
its __init__() method (pickle has waffled on this over the years; not
calling __init__() is current wisdom). In this case, an instance of
an old-style dummy class is created, and then we try to rebind its
__class__ attribute to the desired class object. If this succeeds,
the new instance object is pushed on the stack, and we're done.
Else (the argtuple is not empty, it's not an old-style class object,
or the class object does have a __getinitargs__ attribute), the code
first insists that the class object have a __safe_for_unpickling__
attribute. Unlike as for the __safe_for_unpickling__ check in REDUCE,
it doesn't matter whether this attribute has a true or false value, it
only matters whether it exists (XXX this is a bug). If
__safe_for_unpickling__ doesn't exist, UnpicklingError is raised.
Else (the class object does have a __safe_for_unpickling__ attr),
the class object obtained from INST's arguments is applied to the
argtuple obtained from the stack, and the resulting instance object
is pushed on the stack.
NOTE: checks for __safe_for_unpickling__ went away in Python 2.3.
NOTE: the distinction between old-style and new-style classes does
not make sense in Python 3.
OBJoa Build a class instance.
This is the protocol 1 version of protocol 0's INST opcode, and is
very much like it. The major difference is that the class object
is taken off the stack, allowing it to be retrieved from the memo
repeatedly if several instances of the same class are created. This
can be much more efficient (in both time and space) than repeatedly
embedding the module and class names in INST opcodes.
Unlike INST, OBJ takes no arguments from the opcode stream. Instead
the class object is taken off the stack, immediately above the
topmost markobject:
Stack before: ... markobject classobject stackslice
Stack after: ... new_instance_object
As for INST, the remainder of the stack above the markobject is
gathered into an argument tuple, and then the logic seems identical,
except that no __safe_for_unpickling__ check is done (XXX this is
a bug). See INST for the gory details.
NOTE: In Python 2.3, INST and OBJ are identical except for how they
get the class object. That was always the intent; the implementations
had diverged for accidental reasons.
NEWOBJ aL Build an object instance.
The stack before should be thought of as containing a class
object followed by an argument tuple (the tuple being the stack
top). Call these cls and args. They are popped off the stack,
and the value returned by cls.__new__(cls, *args) is pushed back
onto the stack.
NEWOBJ_EX au Build an object instance.
The stack before should be thought of as containing a class
object followed by an argument tuple and by a keyword argument dict
(the dict being the stack top). Call these cls and args. They are
popped off the stack, and the value returned by
cls.__new__(cls, *args, *kwargs) is pushed back onto the stack.
PROTO zProtocol version indicator.
For protocol 2 and above, a pickle must start with this opcode.
The argument is the protocol version, an int in range(2, 256).
STOP.zStop the unpickling machine.
Every pickle ends with this opcode. The object at the top of the stack
is popped, and that's the result of unpickling. The stack should be
empty then.
FRAME zIndicate the beginning of a new frame.
The unpickler may use this opcode to safely prefetch data from its
underlying stream.
PERSIDPa Push an object identified by a persistent ID.
The pickle module doesn't define what a persistent ID means. PERSID's
argument is a newline-terminated str-style (no embedded escapes, no
bracketing quote characters) string, which *is* "the persistent ID".
The unpickler passes this string to self.persistent_load(). Whatever
object that returns is pushed on the stack. There is no implementation
of persistent_load() in Python's unpickler: it must be supplied by an
unpickler subclass.
BINPERSIDQaX Push an object identified by a persistent ID.
Like PERSID, except the persistent ID is popped off the stack (instead
of being a string embedded in the opcode bytestream). The persistent
ID is passed to self.persistent_load(), and whatever object that
returns is pushed on the stack. See PERSID for more detail.
z%repeated name %r at indices %d and %dz%repeated code %r at indices %d and %dFc C s t }tjD ]}td|s0| rtd| qtt|}t|t rPt
|dkrf| rtd||f q| d}||v r| rtd||f || }|j |krt
d|||j f ||= qt
d||f q|r d g}| D ] \}}|d
|j |f qt
d |d S ) Nz[A-Z][A-Z0-9_]+$z0skipping %r: it doesn't look like an opcode namer# z5skipping %r: value %r doesn't look like a pickle coderL z+checking name %r w/ code %r for consistencyzBfor pickle code %r, pickle.py uses name %r but we're using name %rzPpickle.py appears to have a pickle opcode with name %r and code %r, but we don'tz=we appear to have pickle opcodes that pickle.py doesn't have:z name %r with code %r
)code2opcopyr __all__rematchprintgetattrr r r- rB r r% itemsappendjoin)verboser r Z
picklecoder msgr r r r assure_pickle_consistency sH
r+ c c s t | trt| } t| dr&| j}ndd }| }| d}t| d}|d u r|dkrht
dn t
d|d u rxd n||f |j d u rd }n |j | }|r|||| fV n |||fV |d
kr.|j
d ksJ qq.d S ) Ntellc S s d S r} r r r r r z_genops..r# rL r. z#pickle exhausted before seeing STOPz!at position %s, opcode %r unknownz .r )r bytes_typesioBytesIOhasattrr, r$ r getrB r% r r r )r(
yield_end_posZgetposposr opcoder r r r _genops s0
r8 c C s t | S )ax Generate all the opcodes in a pickle.
'pickle' is a file-like object, or string, containing the pickle.
Each opcode in the pickle is generated, from the current pickle position,
stopping after a STOP opcode is delivered. A triple is generated for
each opcode:
opcode, arg, pos
opcode is an OpcodeInfo record, describing the current opcode.
If the opcode has an argument embedded in the pickle, arg is its decoded
value, as a Python object. If the opcode doesn't have an argument, arg
is None.
If the pickle has a tell() method, pos was the value of pickle.tell()
before reading the current opcode. If the pickle is a bytes object,
it's wrapped in a BytesIO object, and the latter's tell() result is
used. Else (the pickle doesn't have a tell(), and it's not obvious how
to query its current position) pos is None.
)r8 )r r r r r s r c C s d}d}t }i }g }d}d}t| ddD ]\}} }
} d|jv rZ|| ||| f q*|jdkrt|} || ||| f q*d|jv rq*d|jv r|j|kr|j}d || < ||| f q*|jd
kr| |kr| }|
dkr| |
| }n||
| f q*||
| f q*~t }
|
| t
|
|}|d krF|j
d} |D ]\}} d }||u r| |vrrqN|| }| || < | d
7 } n6||u r||| }n | || }t||j jk}|j j|d |r|j | n
| | qN|j |
S )z7Optimize a pickle string by removing unused PUT opcodesr r r r. T)r5 r r Nr r1 Fr# )force)r r8 r addr' r- r r1 r2 writer _Picklerframer
start_framingputr4 _FRAME_SIZE_TARGET commit_frame
file_write end_framinggetvalue)r r? r4 ZoldidsZnewidsopcodesr Z protoheaderr7 r r6 Zend_posidxoutZpickleropZ framelessr( r r r r sl
r c C sl g }|du ri }d}g }d| }d} |}
t | D ]\} } }
|
durVtd|
d|d dt| jdd |t| | jf }t|| j}| j}| j }t|}d}t
|v s| jdkrv|rv|d t
u rvt
|vsJ t
|v r|d t u sJ |rn| }|du r
d }nd
| }|d t
ur,| q| z|
t
}W n( tyj | jdksbJ d }Y n0 nd } }| jd
v r| jdkrt|}d| }n| dusJ | }||v rd| } n,|sd} n |d t
u rd} n |d ||< n<| jdv r0| |v r(t|dks J || g}nd| } | dus@|r|ddt| j 7 }| durp|dt| 7 }|r|d| 7 }|r|d|
t| 7 }t|}
|
dkr|}
|d| jddd 7 }t||d | rt| t||k rtd|t|f |r|| d= t
|v r>> import pickle
>>> x = [1, 2, (3, 4), {b'abc': "def"}]
>>> pkl0 = pickle.dumps(x, 0)
>>> dis(pkl0)
0: ( MARK
1: l LIST (MARK at 0)
2: p PUT 0
5: I INT 1
8: a APPEND
9: I INT 2
12: a APPEND
13: ( MARK
14: I INT 3
17: I INT 4
20: t TUPLE (MARK at 13)
21: p PUT 1
24: a APPEND
25: ( MARK
26: d DICT (MARK at 25)
27: p PUT 2
30: c GLOBAL '_codecs encode'
46: p PUT 3
49: ( MARK
50: V UNICODE 'abc'
55: p PUT 4
58: V UNICODE 'latin1'
66: p PUT 5
69: t TUPLE (MARK at 49)
70: p PUT 6
73: R REDUCE
74: p PUT 7
77: V UNICODE 'def'
82: p PUT 8
85: s SETITEM
86: a APPEND
87: . STOP
highest protocol among opcodes = 0
Try again with a "binary" pickle.
>>> pkl1 = pickle.dumps(x, 1)
>>> dis(pkl1)
0: ] EMPTY_LIST
1: q BINPUT 0
3: ( MARK
4: K BININT1 1
6: K BININT1 2
8: ( MARK
9: K BININT1 3
11: K BININT1 4
13: t TUPLE (MARK at 8)
14: q BINPUT 1
16: } EMPTY_DICT
17: q BINPUT 2
19: c GLOBAL '_codecs encode'
35: q BINPUT 3
37: ( MARK
38: X BINUNICODE 'abc'
46: q BINPUT 4
48: X BINUNICODE 'latin1'
59: q BINPUT 5
61: t TUPLE (MARK at 37)
62: q BINPUT 6
64: R REDUCE
65: q BINPUT 7
67: X BINUNICODE 'def'
75: q BINPUT 8
77: s SETITEM
78: e APPENDS (MARK at 3)
79: . STOP
highest protocol among opcodes = 1
Exercise the INST/OBJ/BUILD family.
>>> import pickletools
>>> dis(pickle.dumps(pickletools.dis, 0))
0: c GLOBAL 'pickletools dis'
17: p PUT 0
20: . STOP
highest protocol among opcodes = 0
>>> from pickletools import _Example
>>> x = [_Example(42)] * 2
>>> dis(pickle.dumps(x, 0))
0: ( MARK
1: l LIST (MARK at 0)
2: p PUT 0
5: c GLOBAL 'copy_reg _reconstructor'
30: p PUT 1
33: ( MARK
34: c GLOBAL 'pickletools _Example'
56: p PUT 2
59: c GLOBAL '__builtin__ object'
79: p PUT 3
82: N NONE
83: t TUPLE (MARK at 33)
84: p PUT 4
87: R REDUCE
88: p PUT 5
91: ( MARK
92: d DICT (MARK at 91)
93: p PUT 6
96: V UNICODE 'value'
103: p PUT 7
106: I INT 42
110: s SETITEM
111: b BUILD
112: a APPEND
113: g GET 5
116: a APPEND
117: . STOP
highest protocol among opcodes = 0
>>> dis(pickle.dumps(x, 1))
0: ] EMPTY_LIST
1: q BINPUT 0
3: ( MARK
4: c GLOBAL 'copy_reg _reconstructor'
29: q BINPUT 1
31: ( MARK
32: c GLOBAL 'pickletools _Example'
54: q BINPUT 2
56: c GLOBAL '__builtin__ object'
76: q BINPUT 3
78: N NONE
79: t TUPLE (MARK at 31)
80: q BINPUT 4
82: R REDUCE
83: q BINPUT 5
85: } EMPTY_DICT
86: q BINPUT 6
88: X BINUNICODE 'value'
98: q BINPUT 7
100: K BININT1 42
102: s SETITEM
103: b BUILD
104: h BINGET 5
106: e APPENDS (MARK at 3)
107: . STOP
highest protocol among opcodes = 1
Try "the canonical" recursive-object test.
>>> L = []
>>> T = L,
>>> L.append(T)
>>> L[0] is T
True
>>> T[0] is L
True
>>> L[0][0] is L
True
>>> T[0][0] is T
True
>>> dis(pickle.dumps(L, 0))
0: ( MARK
1: l LIST (MARK at 0)
2: p PUT 0
5: ( MARK
6: g GET 0
9: t TUPLE (MARK at 5)
10: p PUT 1
13: a APPEND
14: . STOP
highest protocol among opcodes = 0
>>> dis(pickle.dumps(L, 1))
0: ] EMPTY_LIST
1: q BINPUT 0
3: ( MARK
4: h BINGET 0
6: t TUPLE (MARK at 3)
7: q BINPUT 1
9: a APPEND
10: . STOP
highest protocol among opcodes = 1
Note that, in the protocol 0 pickle of the recursive tuple, the disassembler
has to emulate the stack in order to realize that the POP opcode at 16 gets
rid of the MARK at 0.
>>> dis(pickle.dumps(T, 0))
0: ( MARK
1: ( MARK
2: l LIST (MARK at 1)
3: p PUT 0
6: ( MARK
7: g GET 0
10: t TUPLE (MARK at 6)
11: p PUT 1
14: a APPEND
15: 0 POP
16: 0 POP (MARK at 0)
17: g GET 1
20: . STOP
highest protocol among opcodes = 0
>>> dis(pickle.dumps(T, 1))
0: ( MARK
1: ] EMPTY_LIST
2: q BINPUT 0
4: ( MARK
5: h BINGET 0
7: t TUPLE (MARK at 4)
8: q BINPUT 1
10: a APPEND
11: 1 POP_MARK (MARK at 0)
12: h BINGET 1
14: . STOP
highest protocol among opcodes = 1
Try protocol 2.
>>> dis(pickle.dumps(L, 2))
0: \x80 PROTO 2
2: ] EMPTY_LIST
3: q BINPUT 0
5: h BINGET 0
7: \x85 TUPLE1
8: q BINPUT 1
10: a APPEND
11: . STOP
highest protocol among opcodes = 2
>>> dis(pickle.dumps(T, 2))
0: \x80 PROTO 2
2: ] EMPTY_LIST
3: q BINPUT 0
5: h BINGET 0
7: \x85 TUPLE1
8: q BINPUT 1
10: a APPEND
11: 0 POP
12: h BINGET 1
14: . STOP
highest protocol among opcodes = 2
Try protocol 3 with annotations:
>>> dis(pickle.dumps(T, 3), annotate=1)
0: \x80 PROTO 3 Protocol version indicator.
2: ] EMPTY_LIST Push an empty list.
3: q BINPUT 0 Store the stack top into the memo. The stack is not popped.
5: h BINGET 0 Read an object from the memo and push it on the stack.
7: \x85 TUPLE1 Build a one-tuple out of the topmost item on the stack.
8: q BINPUT 1 Store the stack top into the memo. The stack is not popped.
10: a APPEND Append an object to a list.
11: 0 POP Discard the top stack item, shrinking the stack by one item.
12: h BINGET 1 Read an object from the memo and push it on the stack.
14: . STOP Stop the unpickling machine.
highest protocol among opcodes = 2
a=
>>> import pickle
>>> import io
>>> f = io.BytesIO()
>>> p = pickle.Pickler(f, 2)
>>> x = [1, 2, 3]
>>> p.dump(x)
>>> p.dump(x)
>>> f.seek(0)
0
>>> memo = {}
>>> dis(f, memo=memo)
0: \x80 PROTO 2
2: ] EMPTY_LIST
3: q BINPUT 0
5: ( MARK
6: K BININT1 1
8: K BININT1 2
10: K BININT1 3
12: e APPENDS (MARK at 5)
13: . STOP
highest protocol among opcodes = 2
>>> dis(f, memo=memo)
14: \x80 PROTO 2
16: h BINGET 0
18: . STOP
highest protocol among opcodes = 2
)Zdisassembler_testZdisassembler_memo_testc C s dd l } | S r )doctesttestmod)r] r r r _test s r_ __main__z$disassemble one or more pickle files) description pickle_filebr*zthe pickle file)r~ nargshelpz-oz--outputwz+the file where the output should be written)defaultr~ rf z-mz--memo
store_truez#preserve memo between disassemblies)actionrf z-lz
--indentlevelz8the number of blanks by which to indent a new MARK levelz-az
--annotatez2annotate each line with a short opcode descriptionz-pz
--preamblez==> {name} <==zMif more than one pickle file is specified, print this before each disassembly)rh rf z-tz--testzrun self-test suitez-vz)run verbosely; only affects self-test run r r )TT)F)F)NNr1 r )__doc__r@ r1 r r" rT r! r0 r r r r r objectr
structr" r. r) r* r/ r0 r2 r3 r4 r5 r7 r8 rE rF rH rI rJ rK rN rO rP rQ rR rS rV rW rX rY r[ r\ r] r^ ra rb rc rd re rf rk rm rn ro rq rr rs rt ru rv rw rx ry rz r ZpyintZpylongr Zpyinteger_or_boolZpyboolrp Zpyfloatr r Zpybytes_or_strZpystringZpybytesrZ Z pybytearrayZ pyunicoder~ Zpynoner Zpytupler Zpylistr Zpydictr ZpysetZ pyfrozensetZpybufferZ anyobjectrP r r r rE Zname2iZcode2i enumerater
r r r% r r r+ r8 r r r r[ Z _dis_testZ
_memo_testZ__test__r_ r argparseArgumentParserparser add_argumentFileTypestdout
parse_argsargstestrW rb
print_helpr- outputrV rU r' Zpreambleformatr; r r r r s< $ /
;
2
& C