Rename reading methods, fix some encoding errors in error-processing paths and tests #412
Conversation
Mingun
added
enhancement
encoding
Codecov Report
@@ Coverage Diff @@
## master #412 +/- ##
==========================================
- Coverage 50.03% 49.58% -0.46%
==========================================
Files 22 22
Lines 13870 13967 +97
==========================================
- Hits 6940 6925 -15
- Misses 6930 7042 +112
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
tests/xmlrs_reader_tests.rs
Outdated
| from_utf8(a.key.as_ref()).unwrap(), | ||
| from_utf8(&*a.unescaped_value().unwrap()).unwrap() | ||
| decoder.decode(a.key.as_ref()).unwrap(), | ||
| decoder.decode(&*a.unescaped_value().unwrap()).unwrap() |
There was a problem hiding this comment.
This seems like it would be another instance of the incorrect "unescape, then decode" behavior mentioned, yes?
There was a problem hiding this comment.
Yeah, you right. Need to fix that too
|
In the future, something like https://crates.io/crates/buffered-reader might be nice - the internal buffer can dynamically resize which would allow avoiding the copy between the Unfortunately, at the moment, it is LGPL licensed which makes it a bit challenging to depend on. |
Also, Text events was incorrectly unescaped before decoding instead of correct decode, then unescape
…e number when test fail
- read_event => read_event_into - read_to_end => read_to_end_into - read_text => read_text_into - read_event_unbuffered => read_event - read_to_end_unbuffered => read_to_end - read_event_buffered => read_event_impl
…sRef<[u8]>` AsRef is too unsafe because you accidentally could pass wrong parameters Also fixes using incorrect encoding if `read_to_end` family of methods or `read_text` method would not find a corresponding end tag and the reader has a non-UTF-8 encoding
Co-authored-by: Daniel Alley <[email protected]>
This PR includes some of the stuff, improving code here and there, including:
fix incorrect encoding used when handle some errors
remove excess
BufReadconstraints which opens door for theasyncsupport fo the sameReaderstructsome minor improvements in tests
some documentation updates
rename core methods to better reflect their purpose:
read_eventread_event_intoread_to_endread_to_end_intoread_textread_text_intoread_event_unbufferedread_eventread_to_end_unbufferedread_to_end(
read_namespaced_eventis not renamed because I plan to remove it in the namespace PR)