py3: care for .iteritems

[fchapoton]

possibly duplicate with #25948 (which is stalled) ?

CC: @embray @tscrim

Component: python3

Author: Frédéric Chapoton

Branch/Commit: 1e90626

Reviewer: Travis Scrimshaw

Issue created by migration from https://trac.sagemath.org/ticket/26056

[fchapoton]

Commit: 2c8f1e2

[fchapoton]

New commits:

2c8f1e2

py3: some care for .iteritems

[fchapoton]

Branch: u/chapoton/26056

[tscrim]

Reviewer: Travis Scrimshaw

[tscrim]

Changed commit from 2c8f1e2 to 1bfad28

[tscrim]

Changed branch from u/chapoton/26056 to u/tscrim/26056

[tscrim]

comment:3

For free_module_element.pyx, the e is assumed to be a dict. So I can get a little bit of a speedup too:

sage: from sage.modules.free_module_element import FreeModuleElement_generic_sparse
sage: def S(R,n):
....:     return FreeModule(R, n, sparse=True)
....:
sage: %timeit FreeModuleElement_generic_sparse(S(RR,5), {0:-1, 2:2/3, 3:pi, 4:oo})
The slowest run took 15.53 times longer than the fastest. This could mean that an intermediate result is being cached.
100000 loops, best of 3: 16.4 µs per loop

vs before

sage: %timeit FreeModuleElement_generic_sparse(S(RR,5), {0:-1, 2:2/3, 3:pi, 4:oo})
The slowest run took 14.90 times longer than the fastest. This could mean that an intermediate result is being cached.
100000 loops, best of 3: 16.8 µs per loop

At the very least, it yields cleaner Cython code.

If my changes are good, then positive review.

---

New commits:

1bfad28

Making Cython know a little bit more information.

[fchapoton]

comment:4

ok, thanks. I have made one more commit that just fixes a wrong syntax inside one raise TypeError.

I allow myself to set to positive.

---

New commits:

13a2809

some minor details in raise

[fchapoton]

Changed commit from 1bfad28 to 13a2809

[fchapoton]

Changed branch from u/tscrim/26056 to public/ticket/26056

[vbraun]

comment:5

Merge conflict

[sagetrac-git]

Changed commit from 13a2809 to 1e90626

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

1e90626

Merge branch 'public/ticket/26056' in 8.4.b2

[fchapoton]

comment:7

Travis, I have doubts about the line

+        cdef dict entries_dict, e

as it seems that e has been renamed entries ?

EDIT: it's ok, in fact

[fchapoton]

comment:8

the rebase was trivial, I am setting back to positive

[vbraun]

Changed branch from public/ticket/26056 to 1e90626