[erlang-patches] Add lists:init/1 - got tired of implementing it

Hans Svensson hanssv@REDACTED
Thu Jan 23 10:50:56 CET 2014


And a few usages of this function introduced as well. Note, HiPE uses 
its own utiity function hipe_bb:butlast/1 which could/should be replaced 
by droplast as well. But that change looked a bit too invasive so I left 
it as an exercise ;-)

I made this branch on top of the branch with lists:droplast/1, I hope I 
got it right!

git fetch git://github.com/hanssv/otp.git using_lists_droplast

https://github.com/hanssv/otp/compare/add_droplast_to_lists...using_lists_droplast
https://github.com/hanssv/otp/compare/add_droplast_to_lists...using_lists_droplast.patch

Cheers,

/Hans

On 2014-01-22 23:14, Hans Svensson wrote:
> Finally got around to do the fixes. Screwed up my git branch, so the 
> lists:droplast/1 is now in a differently named branch:
>
> git fetch git://github.com/hanssv/otp.git add_droplast_to_lists
>
> https://github.com/hanssv/otp/compare/erlang:maint...add_droplast_to_lists 
>
> https://github.com/hanssv/otp/compare/erlang:maint...add_droplast_to_lists.patch 
>
>
> A separate patch with usage will follow as soon as I can find the time 
> for it!
>
> Cheers,
>
> Hans
>
> On 2014-01-16 11:26, Siri Hansen wrote:
>> Hi Hans!
>>
>> After many and long discussions we have now decided to re-open this
>> patch and approve it with some comments:
>>
>> 1. we would like the name of the function to be 'droplast'.
>> 2. we want the simple recursive implementation with no special handling
>> of bad arguments, i.e. we think it is ok with a function_clause error
>> for arguments that are not non-empty proper lists.
>>
>> Would you mind doing these adjustments in the patch?
>>
>> Best Regards
>> /siri
>>
>>
>> 2013/5/21 Hans Svensson <hanssv@REDACTED <mailto:hanssv@REDACTED>>
>>
>>         Hello Hans,
>>         I've got some feedback on your patch after review:
>>         We have decided to reject this patch as we don't see a general
>>         need for
>>         the new function.
>>
>>         Thanks for showing interest in contributing!
>>
>>         --
>>
>>         BR Fredrik Gustafsson
>>         Erlang OTP Team
>>
>>
>>     That is sad news, and we find it a bit strange since *you* (as in
>>     the OTP team) yourself seems use this particular function/pattern
>>     quite frequently. We did a quick search through the OTP code base
>>     and found quite a few places where this particular function is
>>     either implemented recursively and used, or just uses
>>     reverse(tl(reverse(List)))). Also interesting to see is the
>>     different names used: first, butlast, but_last and 
>> remove_last_element.
>>
>>     Code implementing the "init"-function:
>>     compiler
>>        v3_core:first/1
>>        v3_kernel:first/1
>>     dialyzer
>>        dialyzer_contracts:but_last/1
>>        dialyzer_gui:butlast/1
>>     hipe
>>        hipe_icode_type:butlast/1
>>     orber
>>        orber_interceptors:remove___last_element/1
>>
>>     Direct uses of reverse(tl(reverse(L))):
>>     sasl
>>        in systools_make:smart_guess/3
>>     wx_gen
>>        in wx_gen:erl_skip_opt2/4
>>     ssh
>>        in ssh_sftpd
>>
>>     Any further comments on this, it is really not a "new" function it
>>     is a missing function ;-)
>>
>>     Cheers,
>>
>>     Hans
>>
>>     _________________________________________________
>>     erlang-patches mailing list
>>     erlang-patches@REDACTED <mailto:erlang-patches@REDACTED>
>>     http://erlang.org/mailman/__listinfo/erlang-patches
>>     <http://erlang.org/mailman/listinfo/erlang-patches>
>>
>>
>




More information about the erlang-patches mailing list