Issue 31155: Encode set, frozenset, bytearray, and iterators as json arrays
Created on 2017-08-09 09:23 by javenoneal, last changed 2022-04-11 14:58 by admin. This issue is now closed.
| Pull Requests | |||
|---|---|---|---|
| URL | Status | Linked | Edit |
| PR 3034 | closed | javenoneal, 2017-08-09 09:23 | |
| Messages (4) | |||
|---|---|---|---|
| msg299981 - (view) | Author: Javen O'Neal (javenoneal) * | Date: 2017-08-09 09:23 | |
Currently json.dump, json.dumps, and json.JSONEncoder do not support writing iterator objects. Users can choose to pass the iterator to a tuple and write that as a json array, but this is wasteful if the iterator is large, requiring a large amount of memory to be temporarily allocated for the tuple prior to json encoding. The json module also does not support writing sets, frozensets, or bytearrays. |
|||
| msg315782 - (view) | Author: Ori Avtalion (salty-horse) * | Date: 2018-04-26 07:48 | |
If sets are to be supported, then so should dictionary views. Please also consider the range type. |
|||
| msg315789 - (view) | Author: Raymond Hettinger (rhettinger) * ![]() |
Date: 2018-04-26 12:21 | |
There is some virtue to the current design. Having a user explicitly cast to a list will avoid unintentional conversions. |
|||
| msg339597 - (view) | Author: Inada Naoki (methane) * ![]() |
Date: 2019-04-08 07:52 | |
I concur with Raymond. If you really need this feature, please discuss on python-dev ML. |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:58:49 | admin | set | github: 75338 |
| 2021-10-07 08:11:02 | sedrubal | set | nosy:
+ sedrubal |
| 2019-04-08 07:52:33 | methane | set | status: open -> closed nosy:
+ methane resolution: rejected |
| 2018-04-26 12:21:11 | rhettinger | set | nosy:
+ rhettinger messages: + msg315789 |
| 2018-04-26 07:48:09 | salty-horse | set | nosy:
+ salty-horse messages: + msg315782 |
| 2017-08-09 19:15:47 | rhettinger | set | assignee: bob.ippolito nosy: + bob.ippolito |
| 2017-08-09 09:23:16 | javenoneal | create | |
