Feat: gzip/decompressed response content handling #584
Conversation
| In order to more closely resemble the `requests` interface, where a raw | ||
| and deflated content could be accessed at once, this class lazily reads the | ||
| stream in `transport.Response` so both return forms can be used | ||
| ''' |
There was a problem hiding this comment.
Use double quotes for docstring. Does this need to be fleshed out more?
There was a problem hiding this comment.
Resolved, changed to double quotes with blacken.
| self._raw_content = None | ||
|
|
||
| def _is_compressed(self): | ||
| # The gzip and deflate transfer-encodings are automatically decoded for you. |
There was a problem hiding this comment.
This comment is a bit unclear, maybe rephrase.
There was a problem hiding this comment.
possibly move this into the comment up top as this is the point of this added class.
There was a problem hiding this comment.
Fleshed out this comment in the latest version to explain reasoning.
| @@ -230,8 +280,16 @@ async def request( | |||
| transmitted. The timout error will be raised after such | |||
| request completes. | |||
| """ | |||
| # Headers come in as bytes which isn't expected behavior (resumable) | |||
There was a problem hiding this comment.
What do you mean by "isn't expected behavior"? Can you clarify to explain why the lines below are necessary?
There was a problem hiding this comment.
This isn't expected as most headers are of a str type. It is odd that some come through as bytes. This may be a bandaid, or may be entirely necessary :(
There was a problem hiding this comment.
Added more explanation to why we are doing this. Inconsistencies in the header value types for some cases in the resumable library, which is why we have this check.
Co-authored-by: Christopher Wilcox <[email protected]>
No description provided.