[with patch, positive review] def copy -- they should all be def __copy__

[was]

There are many instances of copy methods in SAGE. They should all be copy, which is what gets called by the standard copy module (part of the standard Python library).

sage: search_sage('def copy')

matrix/sparse_matrix.py:    def copy(self):
modules/free_module_element.py:    def copy(self):
plot/graph.py:    def copy(self, name):
rings/finite_field_element.py:    def copy(self):
rings/fraction_field_element.py:    def copy(self):
rings/laurent_series_ring_element.py:    def copy(self):
rings/padic.py:    def copy(self):
rings/polynomial_element.py:    def copy(self):
rings/polynomial_element.py:    def copy(self):
rings/polynomial_element.py:    def copy(self):
rings/polynomial_element.py:    def copy(self):
rings/power_series_ring_element.py:    def copy(self):
rings/power_series_ring_element.py:    def copy(self):
rings/quotient_ring_element.py:    def copy(self):
libs/pari/functional.py:def copy(self): return pari(self).copy()
modular/modsym/manin_symbols.py:    def copy(self):
server/server1/server1.py:    def copyfile(self, source, outputfile): 
matrix/dense_matrix_pyx.pyx:    def copy(self):
matrix/matrix_generic.pyx:    def copy(self):
matrix/matrix_integer_dense.pyx:    def copy(self):
matrix/matrix_modn_dense.pyx:    def copy(Matrix_modn_dense self):
matrix/matrix_modn_sparse.pyx:    def copy(self):
matrix/matrix_rational_dense.pyx:    def copy(self):
matrix/sparse_matrix_pyx.pyx:    def copy(self):
matrix/sparse_matrix_pyx.pyx:    def copy(self):
matrix/sparse_matrix_pyx.pyx:    def copy(self):
rings/integer.pyx:    def copy(self):
rings/integer_mod.pyx:    def copy(IntegerMod_gmp self):
rings/integer_mod.pyx:    def copy(IntegerMod_int self):
rings/integer_mod.pyx:    def copy(IntegerMod_int64 self):
rings/mpc.pyx:    def copy(self):
rings/polynomial_pyx.pyx:    def copy(self):
rings/polynomial_pyx.pyx:    def copy(self):
rings/rational.pyx:    def copy(self):
rings/real_double.pyx:    def copy(self):
rings/real_mpfr.pyx:    def copy(self):
rings/sparse_poly.pyx:    def copy(self):
rings/sparse_poly.pyx:    def copy(self):
libs/linbox/finite_field_givaro.pyx:    def copy(self):
libs/ntl/ntl.pyx:    def copy(self):
libs/ntl/ntl.pyx:    def copy(self):
libs/ntl/ntl.pyx:    def copy(ntl_GF2E self):
libs/pari/_py_pari_orig.pyx:    def copy(gen self):
libs/pari/gen.pyx:    def copy(gen self):

[was]

Fixed for sage-1.4

[was]

I reverted my changes in my local copy -- there were some nontrivial issues with make such a global change.

[mabshoff]

The number has decreased significantly:

libs/ntl/ntl_ZZX.pyx:    def copy(self):
libs/ntl/ntl_ZZ_pEX.pyx:    def copy(self):
libs/ntl/ntl_ZZ_pX.pyx:    def copy(self):
libs/pari/gen.pyx:    def copy(gen self):
matrix/matrix0.pyx:    def copy(self):
modules/free_module_element.pyx:    def copy(self):
rings/padics/padic_ZZ_pX_CA_element.pyx:    def copy(self):
rings/padics/padic_ZZ_pX_CR_element.pyx:    def copy(self):
rings/padics/padic_ZZ_pX_FM_element.pyx:    def copy(self):
rings/padics/padic_capped_absolute_element.pyx:    def copy(pAdicCappedAbsoluteElement self):
rings/padics/padic_capped_relative_element.pyx:    def copy(self):
rings/padics/padic_fixed_mod_element.pyx:    def copy(self):
rings/polynomial/polynomial_pyx.pyx:    def copy(self):
rings/polynomial/polynomial_pyx.pyx:    def copy(self):
rings/laurent_series_ring_element.pyx:    def copy(self):
rings/power_series_poly.pyx:    def copy(self):
rings/rational.pyx:    def copy(self):
rings/sparse_poly.pyx:    def copy(self):
rings/sparse_poly.pyx:    def copy(self):

It would be nice if someone could take an axe to the remaining ones.

Cheers,

Michael

[AlexGhitza]

In sage-4.1, I am seeing

libs/pari/gen.pyx:1051:    def copy(gen self):
libs/ntl/ntl_ZZ_pEX.pyx:197:    def copy(self):
libs/ntl/ntl_ZZ_pX.pyx:189:    def copy(self):
libs/ntl/ntl_ZZX.pyx:152:    def copy(self):
graphs/graph.py:780:    def copy(self, implementation='networkx', sparse=None):
combinat/matrices/latin.py:321:    def copy(self):
matrix/matrix0.pyx:115:    def copy(self):
modular/modsym/manin_symbols.py:1667:    def copy(self):
rings/laurent_series_ring_element.pyx:943:    def copy(self):
rings/finite_field_givaro.pyx:1118:    cdef FiniteField_givaro copy
rings/finite_field_element.py:396:    def copy(self):
rings/fraction_field_element.pyx:179:    def copy(self):
rings/power_series_poly.pyx:575:    def copy(self):
rings/rational.pyx:558:    def copy(self):
rings/padics/padic_fixed_mod_element.pyx:661:    def copy(self):
rings/padics/padic_ZZ_pX_CA_element.pyx:1567:    def copy(self):
rings/padics/padic_ZZ_pX_CR_element.pyx:2193:    def copy(self):
rings/padics/padic_ZZ_pX_FM_element.pyx:840:    def copy(self):
rings/padics/padic_capped_relative_element.pyx:1453:    def copy(self):
rings/padics/padic_capped_absolute_element.pyx:810:    def copy(pAdicCappedAbsoluteElement self):
rings/polynomial/polynomial_element_generic.py:866:    def copy(self):
rings/polynomial/padics/polynomial_padic_capped_relative_dense.py:726:    def copy(self):
groups/perm_gps/partn_ref/refinement_python.pyx:145:    def copy(self):
modules/free_module_element.pyx:513:    def copy(self):
databases/database.py:575:    def copy(self):
databases/database.py:1003:    def copy(self):
databases/database.py:1546:    def copy(self):

[AlexGhitza]

Doing all of these in one shot is pretty nasty. I've attached a patch that fixes most of the instances, and I have opened two follow-up tickets: #6521 for matrix0.pyx and #6522 for graph.py.

[davidloeffler]

Patch looks fine to me -- I am running doctests at the moment -- but it looks like you missed one in sage/rings/polynomial/polynomial_element_generic.py. I'll do a mini-patch and see if that breaks anything :-)

[davidloeffler]

apply over previous patch

[davidloeffler]

I have uploaded a one-liner patch to deal with the def copy in polynomial_element_generic that got missed out. All doctests pass, both on sage.math and on my machine, so let's get this one in.

[mvngu]

Merged both patches.

[kcrisman]

Just an FYI to those involved on this ticket: there may be a new ticket coming to put in deprecation warnings for these. See the discussion at #6521.