Remove divisible, copy documentation to is_divisible_by instead

[lgoettgens]

Resolves oscar-system/Oscar.jl#3178 by removing divisible and promoting is_divisible_by with a fixed docstring instead. (as suggested by @fieker in oscar-system/Oscar.jl#3178 (comment))

This function is already used throughout AbstractAlgebra. Thus I removed the duplicate implementation for (::Integer, ::Integer) here in Nemo.

cc @JohnAAbbott

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (34c280b) 84.04% compared to head (87ffc41) 84.09%.
Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1618      +/-   ##
==========================================
+ Coverage   84.04%   84.09%   +0.04%     
==========================================
  Files          93       93              
  Lines       36766    36761       -5     
==========================================
+ Hits        30901    30913      +12     
+ Misses       5865     5848      -17     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[fieker]

Shall we also remove divides? Or is the meaning there clear?

[lgoettgens]

I think divides additionally returns the quotient.
I'll put it into the renaming hackmd for further discussion. But that's for the future

[fieker]

On Thu, Jan 11, 2024 at 08:10:57AM -0800, Lars Göttgens wrote: I think divides additionally returns the quotient. I'll put it into the renaming hackmd for further discussion. But that's for the future

Thanks

…

-- Reply to this email directly or view it on GitHub: #1618 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>

[thofma]

divides is is_divisible_by with additional quotient. Not sure if this makes sense, we decided upon the names here: #862 (comment). It would be more logical to do is_divisible_by_with_quotient, but decided against it in #862 (comment).

[lgoettgens]

divides is is_divisible_by with additional quotient. Not sure if this makes sense, we decided upon the names here: #862 (comment). It would be more logical to do is_divisible_by_with_quotient, but decided against it in #862 (comment).

Thanks for the ref. I added it to the hackmd.

This PR is good to go then, right?

[lgoettgens]

Ah, CI is failing, but it seems to be some kind of GAP thing:

OscarInterface: Error During Test at /home/runner/.julia/packages/Oscar/1Moid/test/GAP/oscarinterface.jl:5
  Test threw exception
  Expression: GAP.Globals.TestDirectory(GAP.Globals.DirectoriesPackageLibrary(GAP.Obj("OscarInterface"), GAP.Obj("tst")))
  Error thrown by GAP: Error, UndefVarError: `divisible` not defined

summoning our GAP expert @fingolfin since I don't see this function used in the GAP tests at all...

[lgoettgens]

@fingolfin wanted to look at this in the afternoon

[fingolfin]

Spent the afternoon in renaming discussions... But now I had a look. And it is exactly as I suspected, the error comes from a call into Hecke. Here is condensed version of some of the code being run,

julia> K = GAP.Globals.K
Number field with defining polynomial x^4 + 4*x^3 - 100*x^2 - 208*x + 1744
  over rational field

julia> maximal_order(K)
ERROR: UndefVarError: `divisible` not defined
Stacktrace:
 [1] pmaximal_overorder_at(O::NfOrd, primes::Vector{ZZRingElem})
   @ Hecke ~/.julia/packages/Hecke/HCQ1K/src/NumFieldOrd/NfOrd/MaxOrd/MaxOrd.jl:113
 [2] new_maximal_order(O::NfOrd; index_divisors::Vector{ZZRingElem}, disc::ZZRingElem, ramified_primes::Vector{ZZRingElem})
   @ Hecke ~/.julia/packages/Hecke/HCQ1K/src/NumFieldOrd/NfOrd/MaxOrd/MaxOrd.jl:200
 [3] (::Hecke.var"#2069#2070"{Vector{ZZRingElem}, AnticNumberField})()
   @ Hecke ~/.julia/packages/Hecke/HCQ1K/src/NumFieldOrd/NfOrd/MaxOrd/MaxOrd.jl:44
 [4] get!(default::Hecke.var"#2069#2070"{Vector{ZZRingElem}, AnticNumberField}, h::Dict{Symbol, Any}, key::Symbol)
   @ Base ./dict.jl:479
 [5] get_attribute!(f::Function, G::AnticNumberField, attr::Symbol)
   @ AbstractAlgebra ~/.julia/packages/AbstractAlgebra/R29qD/src/Attributes.jl:230
 [6] MaximalOrder(K::AnticNumberField; discriminant::ZZRingElem, ramified_primes::Vector{ZZRingElem})
   @ Hecke ~/.julia/packages/Hecke/HCQ1K/src/NumFieldOrd/NfOrd/MaxOrd/MaxOrd.jl:42
 [7] MaximalOrder(K::AnticNumberField)
   @ Hecke ~/.julia/packages/Hecke/HCQ1K/src/NumFieldOrd/NfOrd/MaxOrd/MaxOrd.jl:41
 [8] top-level scope
   @ REPL[5]:1

[lgoettgens]

Ah okay, so Hecke doesn't like the deprecation somehow. Let me just mark this breaking then.