➜
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2017-08-17 10:59 by ksqsf, last changed 2022-04-11 14:58 by admin. This issue is now closed.
| Messages (4) | |||
|---|---|---|---|
| msg300406 - (view) | Author: ksqsf (ksqsf) | Date: 2017-08-17 10:59 | |
It'll be nice if `update' can accept str directly, rather than str.encode first. One of the most common use cases is "m.update(s.encode())", so it shouldn't make a big difference to just use "m.update(s)"; after all, users who want to specify encoding explicitly can still do so. |
|||
| msg300409 - (view) | Author: Antoine Pitrou (pitrou) *  |
Date: 2017-08-17 12:24 | |
This will never happen as this is precisely the kind of confusion Python 3 was designed to avoid. |
|||
| msg300411 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * |
Date: 2017-08-17 12:32 | |
Explicit is better than implicit. It is easy to add ".encode()" in the user code if it needs this. Adding this feature has non-zero (and actually not a small) cost. Since update() is not just a method, but a part of a protocol, you need to update all implementations. On other hand, raising an error on string argument can help to catch an error if a string unexpectedly occurred instead of a bytes object. |
|||
| msg300459 - (view) | Author: Christian Heimes (christian.heimes) * |
Date: 2017-08-17 21:56 | |
Yes, what Antoine and Serhiy said. |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:58:50 | admin | set | github: 75408 |
| 2017-08-17 21:56:57 | christian.heimes | set | messages: + msg300459 |
| 2017-08-17 12:32:08 | serhiy.storchaka | set | nosy:
+ gregory.p.smith, serhiy.storchaka, christian.heimes messages: + msg300411 versions: + Python 3.7, - Python 3.6 |
| 2017-08-17 12:24:43 | pitrou | set | status: open -> closed nosy: + pitrou messages: + msg300409 resolution: rejected stage: resolved |
| 2017-08-17 10:59:35 | ksqsf | create | |