^ bcC@s/dZdddgZdZdZdZdefdYZdd lmZ d Z ed d d dde ddZ dZ ed dd dde ddZ dZed dd ddeddZeedZed dd ededdZdZed dd ededdZd Zed d!d ededd"Zd#Zed d$d ededd%Zd&Zed d'd ededd(Zd)Zed d*d ededd+Zd,Zed d-d ededd.Zd/Zd0Z ed d1d ededd2Z!ed d3d ede dd4Z"d5Z#ed d6d ede#dd7Z$d8Z%ed d9d d:de%dd;Z&dd<l'm(Z(d=Z)ed d>d ede)dd?Z*d@Z+ed dAd ede+ddBZ,dCefdDYZ-e-d dEdFe.ddGZ/e-d dHdFe0ddIZ1e-d dJdFe.e0e2fddKZ3e-d dLdFe2fddMZ4e-d dNdFe5ddOZ6e-d dPdFe7ddQZ8e-d dRdFe9ddSZ:e-d dTdFe;dddUZ=e-d dVdFe>ddWZ?e-d dXdFe@ddYZAe-d dZdFeBdd[ZCe-d d\dFedd]ZDe-d d^dFe-dd_ZEe-d d`dFe-ddaZFdbefdcYZGeGZHeHd dddedfdge!dhgdie3gdjdkddleHd dmdedndgedhgdie/gdjdddoeHd dpdedqdge dhgdie/gdjdddreHd dsdedtdge dhgdie/gdjdddueHd dvdedwdge"dhgdie1gdjdkddxeHd dydedzdge*dhgdie1gdjddd{eHd d|ded}dge,dhgdie1gdjddd~eHd ddeddgedhgdie8gdjdkddeHd ddeddgedhgdie8gdjdddeHd ddeddgedhgdie8gdjdddeHd ddeddgddhgdie=gdjdkddeHd ddeddgddhgdie4gdjdddeHd ddeddgddhgdie4gdjdddeHd ddeddgedhgdie:gdjdkddeHd ddeddgedhgdie:gdjdddeHd ddeddge$dhgdie6gdjdkddeHd ddeddge&dhgdie6gdjdddeHd ddeddgddhgdieAgdjdddeHd ddeddgddheAeDgdieAgdjdkddeHd ddeddgddheAeEeFgdieAgdjdddeHd ddeddgddheEeFgdieAgdjdkddeHd ddeddgddhgdie?gdjdddeHd ddeddgddheEeFgdie?gdjdkddeHd ddeddgddheDgdie?gdjdddeHd ddeddgddheDeDgdie?gdjdddeHd ddeddgddheDeDeDgdie?gdjdddeHd ddeddgddhgdieCgdjdddeHd ddeddgddheEeFgdieCgdjdkddeHd ddeddgddheCeDeDgdieCgdjdkddeHd ddeddgddheCeEeFgdieCgdjdddeHd ddeddgddheDgdigdjdkddeHd ddeddgddheDgdieDeDgdjdkddeHd ddeddgddhgdieEgdjdkddeHd ddeddgddheEeFgdigdjdddeHd ddeddge!dhgdieDgdjdkddeHd ddeddge dhgdieDgdjdddeHd ddeddgedhgdieDgdjdddeHd ddeddge!dhgdigdjdkddeHd ddeddge dhgdigdjdddeHd ddeddgedhgdigdjdddeHd ddeddge dhgdieDgdjdddeHd ddeddge dhgdieDgdjdddeHd ddeddgedhgdieDgdjdddeHd ddeddgedhgdieDgdjdkddeHd ddeddgddheDeDgdieDgdjdkddeHd ddeddgddheDeDgdieDgdjdkddeHd ddeddgedheEeFgdieDgdjdkddeHd ddeddgddheEeDeFgdieDgdjdddeHd ddeddgddheDeDgdieDgdjdddeHd ddeddge dhgdigdjdddeHd ddeddgddheDgdigdjdkddeHd ddeddgedhgdieDgdjdkddeHd ddeddgddheDgdieDgdjdddg5ZI[HiZJiZKxeLeID]\ZMZNeNjOeJkrePd eNjOeJeNjOeMfneNjQeKkrNePd eNjQeKeNjQeMfneMeJeNjO>> import StringIO >>> read_uint1(StringIO.StringIO('\xff')) 255 is'not enough data in stream to read uint1N(treadtordt ValueError(tftdata((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_uint1s Rtuint1RiRRsOne-byte unsigned integer.cCsB|jd}t|dkr2td|dStddS(s >>> import StringIO >>> read_uint2(StringIO.StringIO('\xff\x00')) 255 >>> read_uint2(StringIO.StringIO('\xff\xff')) 65535 is>> import StringIO >>> read_int4(StringIO.StringIO('\xff\x00\x00\x00')) 255 >>> read_int4(StringIO.StringIO('\x00\x00\x00\x80')) == -(2**31) True is>> import StringIO >>> read_stringnl(StringIO.StringIO("'abcd'\nefg\n")) 'abcd' >>> read_stringnl(StringIO.StringIO("\n")) Traceback (most recent call last): ... ValueError: no string quotes around '' >>> read_stringnl(StringIO.StringIO("\n"), stripquotes=False) '' >>> read_stringnl(StringIO.StringIO("''\n")) '' >>> read_stringnl(StringIO.StringIO('"abcd"')) Traceback (most recent call last): ... ValueError: no newline found when trying to read stringnl Embedded escapes are undone in the result. >>> read_stringnl(StringIO.StringIO(r"'a\n\\b\x00c\td'" + "\n'e'")) 'a\n\\b\x00c\td' s s-no newline found when trying to read stringnlis'"s,strinq quote %r not found at both ends of %risno string quotes around %rt string_escape(treadlinetendswithRt startswithtdecode(RR't stripquotesRtq((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_stringnls     tstringnlsA newline-terminated string. This is a repr-style string, with embedded escapes, and bracketing quotes. cCst|dtdtS(NR'R((R*tFalse(R((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_stringnl_noescapeAststringnl_noescapesA newline-terminated string. This is a str-style string, without embedded escapes, or bracketing quotes. It should consist solely of printable ASCII characters. cCsdt|t|fS(s| >>> import StringIO >>> read_stringnl_noescape_pair(StringIO.StringIO("Queue\nEmpty\njunk")) 'Queue Empty' s%s %s(R-(R((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_stringnl_noescape_pairOststringnl_noescape_pairsA 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. cCspt|}|dkr+td|n|j|}t||krP|Std|t|fdS(sh >>> import StringIO >>> read_string4(StringIO.StringIO("\x00\x00\x00\x00abc")) '' >>> read_string4(StringIO.StringIO("\x03\x00\x00\x00abcdef")) 'abc' >>> read_string4(StringIO.StringIO("\x00\x00\x00\x03abcdef")) Traceback (most recent call last): ... ValueError: expected 50331648 bytes in a string4, but only 6 remain isstring4 byte count < 0: %ds2expected %d bytes in a string4, but only %d remainN(R!RRR(RRR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_string4es  tstring4sA 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. cCsct|}|dkst|j|}t||krC|Std|t|fdS(s >>> import StringIO >>> read_string1(StringIO.StringIO("\x00")) '' >>> read_string1(StringIO.StringIO("\x03abcdef")) 'abc' is2expected %d bytes in a string1, but only %d remainN(RR RRR(RRR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_string1s tstring1sA 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. cCsA|j}|jds*tdn|d }t|dS(sq >>> import StringIO >>> read_unicodestringnl(StringIO.StringIO("abc\uabcd\njunk")) u'abc\uabcd' s s4no newline found when trying to read unicodestringnlisraw-unicode-escape(R$R%Rtunicode(RR((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_unicodestringnls   tunicodestringnlsA newline-terminated Unicode string. This is raw-unicode-escape encoded, so consists of printable ASCII characters, and may contain embedded escape sequences. cCsyt|}|dkr+td|n|j|}t||krYt|dStd|t|fdS(s >>> import StringIO >>> s = u'abcd\uabcd' >>> enc = s.encode('utf-8') >>> enc 'abcd\xea\xaf\x8d' >>> n = chr(len(enc)) + chr(0) * 3 # little-endian 4-byte length >>> t = read_unicodestring4(StringIO.StringIO(n + enc + 'junk')) >>> s == t True >>> read_unicodestring4(StringIO.StringIO(n + enc[:-1])) Traceback (most recent call last): ... ValueError: expected 7 bytes in a unicodestring4, but only 6 remain is!unicodestring4 byte count < 0: %dsutf-8s9expected %d bytes in a unicodestring4, but only %d remainN(R!RRRR5(RRR((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_unicodestring4s   tunicodestring4sAA 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. cCst|dtdt}|jdr:td|n|dkrJtS|dkrZtSyt|SWntk rt|SXdS(s >>> import StringIO >>> read_decimalnl_short(StringIO.StringIO("1234\n56")) 1234 >>> read_decimalnl_short(StringIO.StringIO("1234L\n56")) Traceback (most recent call last): ... ValueError: trailing 'L' not allowed in '1234L' R'R(tLstrailing 'L' not allowed in %rt00t01N(R*R,R%RtTrueR t OverflowErrorR (Rts((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_decimalnl_shorts    cCsDt|dtdt}|jds:td|nt|S(s >>> import StringIO >>> read_decimalnl_long(StringIO.StringIO("1234\n56")) Traceback (most recent call last): ... ValueError: trailing 'L' required in '1234' Someday the trailing 'L' will probably go away from this output. >>> read_decimalnl_long(StringIO.StringIO("1234L\n56")) 1234L >>> read_decimalnl_long(StringIO.StringIO("123456789012345678901234L\n6")) 123456789012345678901234L R'R(R:strailing 'L' required in %r(R*R,R%RR (RR?((s0/opt/alt/python27/lib64/python2.7/pickletools.pytread_decimalnl_longstdecimalnl_shortsA 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. tdecimalnl_longsA newline-terminated decimal integer literal. This has a trailing 'L', and can represent integers of any size. cCs"t|dtdt}t|S(s[ >>> import StringIO >>> read_floatnl(StringIO.StringIO("-1.25\n6")) -1.25 R'R((R*R,tfloat(RR?((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_floatnl2stfloatnlsA 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). cCsB|jd}t|dkr2td|dStddS(s >>> import StringIO, struct >>> raw = struct.pack(">d", -1.25) >>> raw '\xbf\xf4\x00\x00\x00\x00\x00\x00' >>> read_float8(StringIO.StringIO(raw + "\n")) -1.25 is>dis(not enough data in stream to read float8N(RRRR(RR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_float8Hs tfloat8isAn 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 cPickle 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). (t decode_longcCsFt|}|j|}t||kr<tdnt|S(sT >>> import StringIO >>> read_long1(StringIO.StringIO("\x00")) 0L >>> read_long1(StringIO.StringIO("\x02\xff\x00")) 255L >>> read_long1(StringIO.StringIO("\x02\xff\x7f")) 32767L >>> read_long1(StringIO.StringIO("\x02\x00\xff")) -256L >>> read_long1(StringIO.StringIO("\x02\x00\x80")) -32768L s'not enough data in stream to read long1(RRRRRI(RRR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_long1ns  tlong1sA 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. cCset|}|dkr+td|n|j|}t||kr[tdnt|S(s >>> import StringIO >>> read_long4(StringIO.StringIO("\x02\x00\x00\x00\xff\x00")) 255L >>> read_long4(StringIO.StringIO("\x02\x00\x00\x00\xff\x7f")) 32767L >>> read_long4(StringIO.StringIO("\x02\x00\x00\x00\x00\xff")) -256L >>> read_long4(StringIO.StringIO("\x02\x00\x00\x00\x00\x80")) -32768L >>> read_long1(StringIO.StringIO("\x00\x00\x00\x00")) 0L islong4 byte count < 0: %ds'not enough data in stream to read long4(R!RRRRI(RRR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt read_long4s  tlong4sA 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 long 0L, although LONG1 should really be used then instead (and in any case where # of bytes < 256). t StackObjectcBs eZdZdZdZRS(RtobtypeRcCst|tst||_t|tsBt|tsBtt|trzx&|D]}t|tsXtqXWn||_t|tst||_dS(N(RR R RttypettupleROR(RRRORt contained((s0/opt/alt/python27/lib64/python2.7/pickletools.pyRs $  cCs|jS(N(R(R((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt__repr__s(RROR(RRRRRS(((s0/opt/alt/python27/lib64/python2.7/pickletools.pyRNs  R ROs3A short (as opposed to long) Python integer object.R s3A long (as opposed to short) Python integer object.t int_or_bools:A Python integer object (short or long), or a Python bool.tboolsA Python bool object.RDsA Python float object.R sA Python string object.R5sA Python Unicode string object.tNonesThe Python None object.RQsA Python tuple object.tlistsA Python list object.tdictsA Python dict object.tanysAny kind of object whatsoever.tmarks'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. t stackslicesLAn 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). t OpcodeInfocBseZdZdZRS( Rtcodetargt stack_beforet stack_aftertprotoRc Csht|tst||_t|ts3tt|dksKt||_|dksut|tsut||_t|t stx#|D]}t|t stqW||_ t|t stx#|D]}t|t stqW||_ t|t tfr7d|ko2dkns=t||_t|ts[t||_dS(Niii(RR R RRR]RVRR^RWRNR_R`R R RaR( RRR]R^R_R`RaRtx((s0/opt/alt/python27/lib64/python2.7/pickletools.pyRVs&  !     7 (RR]R^R_R`RaR(RRRR(((s0/opt/alt/python27/lib64/python2.7/pickletools.pyR\7stINTR]tIR^R_R`RaisPush 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. tBININTtJs1Push 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. tBININT1tKsPush a one-byte unsigned integer. This is a space optimization for pickling very small non-negative ints, in range(256). tBININT2tMsPush 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. tLONGR:sPush 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. tLONG1ss|Long integer using one-byte length. A more efficient encoding of a Python long; the long1 encoding says it all.tLONG4ss~Long integer using found-byte length. A more efficient encoding of a Python long; the long4 encoding says it all.tSTRINGtSsPush 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. t BINSTRINGtTsPush 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. tSHORT_BINSTRINGtUsPush 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. tNONEtNsPush None on the stack.tNEWTRUEssPush True onto the stack.tNEWFALSEssPush False onto the stack.tUNICODEtVsPush 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. t BINUNICODEtXsPush a Python Unicode string object. There are two arguments: the first is a 4-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. tFLOATtFsNewline-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. tBINFLOATtGsFloat 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). t EMPTY_LISTt]sPush an empty list.tAPPENDtasAppend an object to a list. Stack before: ... pylist anyobject Stack after: ... pylist+[anyobject] although pylist is really extended in-place. tAPPENDStesExtend a list by a slice of stack objects. Stack before: ... pylist markobject stackslice Stack after: ... pylist+stackslice although pylist is really extended in-place. tLISTtlssBuild 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'] t EMPTY_TUPLEt)sPush an empty tuple.tTUPLEttsvBuild 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') tTUPLE1ssBuild 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:]) tTUPLE2ssBuild 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:])] tTUPLE3ssBuild 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:])] t EMPTY_DICTt}sPush an empty dict.tDICTtdsBuild 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'} tSETITEMR?sAdd 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. tSETITEMStus\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. tPOPt0s<Discard the top stack item, shrinking the stack by one item.tDUPt2s=Push the top stack item onto the stack again, duplicating it.tMARKt(sPush markobject onto the stack. markobject is a unique object, used by other opcodes to identify a region of the stack containing a variable number of objects for them to work on. See markobject.doc for more detail. tPOP_MARKt1sPop all the stack objects at and above the topmost markobject. When an opcode using a variable number of stack objects is done, POP_MARK is used to remove those objects, and to remove the markobject that delimited their starting position on the stack. tGETtgsRead an object from the memo and push it on the stack. The index of the memo object to push is given by the newline-terminated decimal string following. BINGET and LONG_BINGET are space-optimized versions. tBINGETthsRead an object from the memo and push it on the stack. The index of the memo object to push is given by the 1-byte unsigned integer following. t LONG_BINGETtjsRead an object from the memo and push it on the stack. The index of the memo object to push is given by the 4-byte signed little-endian integer following. tPUTtpsStore the stack top into the memo. The stack is not popped. The index of the memo location to write into is given by the newline- terminated decimal string following. BINPUT and LONG_BINPUT are space-optimized versions. tBINPUTR)sStore the stack top into the memo. The stack is not popped. The index of the memo location to write into is given by the 1-byte unsigned integer following. t LONG_BINPUTtrsStore the stack top into the memo. The stack is not popped. The index of the memo location to write into is given by the 4-byte signed little-endian integer following. tEXT1ssExtension code. This code and the similar EXT2 and EXT4 allow using a registry of popular objects that are pickled by name, typically classes. It is envisioned that through a global negotiation and registration process, third parties can set up a mapping between ints and object names. In order to guarantee pickle interchangeability, the extension code registry ought to be global, although a range of codes may be reserved for private use. EXT1 has a 1-byte integer argument. This is used to index into the extension registry, and the object at that index is pushed on the stack. tEXT2ssNExtension code. See EXT1. EXT2 has a two-byte integer argument. tEXT4ssOExtension code. See EXT1. EXT4 has a four-byte integer argument. tGLOBALtcsPush a global object (module.attr) on the stack. Two newline-terminated strings follow the GLOBAL opcode. The first is taken as a module name, and the second as a class name. The class object module.class is pushed on the stack. More accurately, the object returned by self.find_class(module, class) is pushed on the stack, so unpickling subclasses can override this form of lookup. tREDUCEtRsNPush an object built from a callable and an argument tuple. The opcode is named to remind of the __reduce__() method. Stack before: ... callable pytuple Stack after: ... callable(*pytuple) The callable and the argument tuple are the first two items returned by a __reduce__ method. Applying the callable to the argtuple is supposed to reproduce the original object, or at least get it started. If the __reduce__ method returns a 3-tuple, the last component is an argument to be passed to the object's __setstate__, and then the REDUCE opcode is followed by code to create setstate's argument, and then a BUILD opcode to apply __setstate__ to that argument. If type(callable) is not ClassType, REDUCE complains unless the callable has been registered with the copy_reg module's safe_constructors dict, or the callable has a magic '__safe_for_unpickling__' attribute with a true value. I'm not sure why it does this, but I've sure seen this complaint often enough when I didn't want to . tBUILDtbsFinish 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) This may raise RuntimeError in restricted execution mode (which disallows access to __dict__ directly); in that case, the object is updated instead via for k, v in argument.items(): anyobject[k] = v tINSTtis 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). + It's an old-style class object (the type of the class object is ClassType). + 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. In restricted execution mode it can fail (assignment to __class__ is disallowed), and I'm not really sure what happens then -- it looks like the code ends up calling the class object's __init__ anyway, via falling into the next case. 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; cPickle requires the attribute to be true). 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. tOBJtosBuild 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; cPickle does test __safe_for_unpickling__). 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. tNEWOBJssLBuild 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. tPROTOssProtocol 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). tSTOPt.sStop 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. tPERSIDtPsPush 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. t BINPERSIDtQsXPush 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. s%repeated name %r at indices %d and %ds%repeated code %r at indices %d and %dc Csddl}ddl}tj}x|jD]}|jd|s^|r.d|GHq.q.nt||}t|t st |dkr|r.d||fGHq.q.n||kr|rd||fGHn||}|j |kr t d|||j fn||=q.t d||fq.W|rd g}x4|j D]&\}}|j d |j |fqIWt d j|ndS( Nis[A-Z][A-Z0-9_]+$s0skipping %r: it doesn't look like an opcode nameis5skipping %r: value %r doesn't look like a pickle codes+checking name %r w/ code %r for consistencysBfor pickle code %r, pickle.py uses name %r but we're using name %rsPpickle.py appears to have a pickle opcode with name %r and code %r, but we don'ts=we appear to have pickle opcodes that pickle.py doesn't have:s name %r with code %rs (tpickletretcode2optcopyt__all__tmatchtgetattrRR RRRtitemstappendtjoin( tverboseRRRRt picklecodeRtmsgR]((s0/opt/alt/python27/lib64/python2.7/pickletools.pytassure_pickle_consistencys>  "    ccs8ddl}t|tr-|j|}nt|drH|j}n d}xtr3|}|jd}tj |}|dkr|dkrt dqt d|dkrd p||fn|j dkrd}n|j j |}|||fV|d krT|jd ks,tPqTqTWdS( szGenerate 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 string object, it's wrapped in a StringIO 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. iNttellcSsdS(N(RV(((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt+tiRs#pickle exhausted before seeing STOPs!at position %s, opcode %r unknowns RR(t cStringIORR tStringIOthasattrRR=RRtgetRVRR^RRR (RRtgetpostposR]topcodeR^((s0/opt/alt/python27/lib64/python2.7/pickletools.pyR s.         c Cst}g}d}xt|D]x\}}}|dk r\|j|||fd}nd|jkr{||}}q"d|jkr"|j|q"q"Wg}d} xI|D]A\}} } ||kr| n| } |j|| | !| } qW|j|| dj|S(s7Optimize a pickle string by removing unused PUT opcodesRRiRN(tsetRVRRRtaddR( RtgetstputstprevposRR^RtprevargR?RtstarttstopR((s0/opt/alt/python27/lib64/python2.7/pickletools.pyRDs&    cCsg}|dkri}nd}g}d|}d}xt|D]q\} } } | dk rp|d| Indt| jdd!|t|| jf} t|| j}| j} | j }t| }d}t | ks| jdkr|r|dt krt |kst t | kr>| dt ks>t n|r|j }|dkred}n d|}x|dt k r|j qrW|j y| jt }Wqtk r| jdkst d }qXqd }}n| jdkr_| dk s t | |kr&d| }q|s5d}q|dt krNd}q|d|| >> import pickle >>> x = [1, 2, (3, 4), {'abc': u"def"}] >>> pkl = pickle.dumps(x, 0) >>> dis(pkl) 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: S STRING 'abc' 37: p PUT 3 40: V UNICODE u'def' 45: p PUT 4 48: s SETITEM 49: a APPEND 50: . STOP highest protocol among opcodes = 0 Try again with a "binary" pickle. >>> pkl = pickle.dumps(x, 1) >>> dis(pkl) 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: U SHORT_BINSTRING 'abc' 24: q BINPUT 3 26: X BINUNICODE u'def' 34: q BINPUT 4 36: s SETITEM 37: e APPENDS (MARK at 3) 38: . 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: ( MARK 6: i INST 'pickletools _Example' (MARK at 5) 28: p PUT 1 31: ( MARK 32: d DICT (MARK at 31) 33: p PUT 2 36: S STRING 'value' 45: p PUT 3 48: I INT 42 52: s SETITEM 53: b BUILD 54: a APPEND 55: g GET 1 58: a APPEND 59: . STOP highest protocol among opcodes = 0 >>> dis(pickle.dumps(x, 1)) 0: ] EMPTY_LIST 1: q BINPUT 0 3: ( MARK 4: ( MARK 5: c GLOBAL 'pickletools _Example' 27: q BINPUT 1 29: o OBJ (MARK at 4) 30: q BINPUT 2 32: } EMPTY_DICT 33: q BINPUT 3 35: U SHORT_BINSTRING 'value' 42: q BINPUT 4 44: K BININT1 42 46: s SETITEM 47: b BUILD 48: h BINGET 2 50: e APPENDS (MARK at 3) 51: . 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 sM >>> import pickle >>> from StringIO import StringIO >>> f = StringIO() >>> p = pickle.Pickler(f, 2) >>> x = [1, 2, 3] >>> p.dump(x) >>> p.dump(x) >>> f.seek(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 tdisassembler_testtdisassembler_memo_testcCsddl}|jS(Ni(tdoctestttestmod(R((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt_tests t__main__N((^t__doc__RR RRtobjectRtstructRRRRRR R!R"R=R*R+R-R.R/R0R1R2R3R4R6R7R8R9R@RARBRCRERFRGRHRRIRJRKRLRMRNR tpyintR tpylongRUtpyinteger_or_booltpyboolRDtpyfloatR tpystringR5t pyunicodeRPRVtpynoneRQtpytupleRWtpylistRXtpydictt anyobjectRR[R\Rdtopcodestname2itcode2it enumerateRRRRR]RR,RRRRRt _dis_testt _memo_testt__test__RR(((s0/opt/alt/python27/lib64/python2.7/pickletools.pyt sh"      1                                         ;                                                                                                           1                     & 9