File objects for publish datasource #714
Conversation
| error = "Datasource item must have a name when passing a file object" | ||
| raise ValueError(error) | ||
|
|
||
| file_extension = 'tdsx' if file_is_compressed(file) else 'tds' |
There was a problem hiding this comment.
I think we might have a problem here -- you can directly publish a tde or hyper file, both of which are actually compressed file formats themselves (I'm not sure if they're both zip or some other format)
There was a problem hiding this comment.
This might not be an issue practically -- if they're loading a tde or hyper file in memory it might just fail on publish -- and given that they can't really update it in memory, there's little reason to do this.
I wonder if there is a way to warn the caller early though.
There was a problem hiding this comment.
I modified the file_peeking to return file_type instead of boolean value and extended it to identify .xml, .tde, and .hyper. I couldn't find official sources for .tde or .hyper file signatures but was able to deduce them by creating few different files and peeking their content.
Now if users provides '.tde' or '.hyper' file they are informed that these are not-supported file types. Additionally, they will be shown an error if they're providing file types we can't identify. Let me know what you think
Extend support to tde and hyper files
t8y8
left a comment
t8y8
left a comment
There was a problem hiding this comment.
Two small nits but approved.
The tests are excellent, thank you!
|
Made the changes pointed out. Should really learn to use pycodestyle locally. I've basically run into broken travis builds every time I've pushed something out 😅 It could be nice to force some local testing in the form of pre-push hooks or similar. |
2 participants
Extend support for publishing file objects to datasource.publish as discussed in #704. I followed the same approach as with workbooks. Let me know what you think :)