github
html_url | issue_url | id | node_id | user | created_at | updated_at | author_association | body | reactions | issue | performed_via_github_app |
---|---|---|---|---|---|---|---|---|---|---|---|
https://github.com/simonw/sqlite-utils/issues/448#issuecomment-1297703307 | https://api.github.com/repos/simonw/sqlite-utils/issues/448 | 1297703307 | IC_kwDOCGYnMM5NWWGL | 167893 | 2022-10-31T21:23:51Z | 2022-10-31T21:27:32Z | CONTRIBUTOR | The Windows aspect is a red herring: OP's sample above produces the same error on Linux. (Though I don't know what's going on with the CI). The same error can also be obtained by passing an `io` from a file opened in non-binary mode (`'r'` as opposed to `'rb'`) to `rows_from_file()`. This is how I got here. The fix for my case is easy: open the file in mode `'rb'`. The analagous fix for OP's problem also works: use `BytesIO` in place of `StringIO`. Minimal test case (derived from [utils.py](https://github.com/simonw/sqlite-utils/blob/main/sqlite_utils/utils.py#L304)): ``` python import io from typing import cast #fp = io.StringIO("id,name\n1,Cleo") # error fp = io.BytesIO(bytes("id,name\n1,Cleo", encoding='utf-8')) # okay reader = io.BufferedReader(cast(io.RawIOBase, fp)) reader.peek(1) # exception thrown here ``` I see the signature of `rows_from_file()` correctly has `fp: BinaryIO` but I guess you'd need either a runtime type check for that (not all `io`s have `mode()`), or to catch the `AttributeError` on `peek()` to produce a better error for users. Neither option is ideal. Some thoughts on testing binary-ness of `io`s in this SO question: https://stackoverflow.com/questions/44584829/how-to-determine-if-file-is-opened-in-binary-or-text-mode | {"total_count": 2, "+1": 2, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0} | 1279144769 | |
https://github.com/simonw/sqlite-utils/pull/508#issuecomment-1297788531 | https://api.github.com/repos/simonw/sqlite-utils/issues/508 | 1297788531 | IC_kwDOCGYnMM5NWq5z | 7908073 | 2022-10-31T22:54:33Z | 2022-11-17T15:11:16Z | CONTRIBUTOR | Maybe this is actually a problem in the python sqlite bindings. Given [SQLITE's stance on this](https://www.sqlite.org/invalidutf.html) they should probably use `encode('utf-8', 'surrogatepass')`. As far as I understand the error here won't actually be resolved by this PR as-is. We would need to modify the data with `surrogateescape`... :/ or modify the sqlite3 module to use `surrogatepass` | {"total_count": 0, "+1": 0, "-1": 0, "laugh": 0, "hooray": 0, "confused": 0, "heart": 0, "rocket": 0, "eyes": 0} | 1430563092 |