Permalink
Comparing changes
Choose two branches to see what’s changed or to start a new pull request.
If you need to, you can also or
learn more about diff comparisons.
Open a pull request
Create a new pull request by comparing changes across two branches. If you need to, you can also .
Learn more about diff comparisons here.
...
- 5 commits
- 34 files changed
- 1 contributor
Commits on Feb 18, 2017
-
String: fix inconsistency introduced by lchomp. Fixes #3729
Now we have: - chomp: removes trailing '\n' or \'r\n' - chomp(suffix): removes suffix (if suffix is '\n' it acts as `chomp()`) - rchop: removes last char - rchop(suffix): removes suffix - lchop: removes first char - lchop(prefix): removes prefix We remove `'lchomp` We also have: - strip: removes leading and trailing whitespace - strip(char): removes all leading and trailing occurrences of char - strip(string): removes all leading and trailing occurences of any char of string - lstrip, rstrip: similar to the above three, but only in one side of the string With this we have pretty much covered all ways to remove things from the front/end of a String in a consistent way. There's `chomp` that has no right/left version, but that came from Ruby and is very specific to removing a trailing newline. The chomp(arg) variant is kept because one can do `io.gets(delimiter, chomp: true)` and that would be equivalent to doying `io.gets(delimiter).chomp(delimiter)`.
-
String: make
to_slicereturn a copy of the underlying bytes. Add `t……o_unsafe_slice` for an unsafe version. The `to_slice` method was unsafe as it could lead to segfaults if one wrote to the returned bytes. For example: ``` string = "hello" string.to_slice[0] = 0_u8 ``` However, there's nothing obviously unsafe by looking at the previous snippet. This isn't good, because as a general rule unsafe methods should have "unsafe" in their name. To fix this, we make `to_slice` return a copy of the bytes, which can be safely mutated. We also add `to_unsafe_slice` which returns an unsafe view of the bytes, which might lead to unsafe code if one mutates the returned bytes. Because some methods invoked `to_slice` on String only for read-only behaviour without expecting a copy, this might now introduce allocations that didn't exist before. To fix that we introduce `Slice.unsafe_readonly(data)` that will invoke `to_unsafe_slice` on String and `to_slice` on everything else. This is used in, for example, Base64, Adler32, MD5, CRC32, etc. We also make `IO::Memory#to_slice` return a copy, for the same reasons, providing `IO::Memory#to_unsafe_slice` too.
-
Syntax: removed octal escapes inside chars. Inside strings, an octal …
…means a byte value, not a codepoint. Added hex escape inside strings (byte value).
-
String: use replacement char when traversing an invalid UTF-8 byte se…
…quence. - String methods now never raise InvalidByteSequenceError - Added String#valid_encoding? to determine if it's properly encoded - Added String#scrub to replace invalid byte sequences - String#inspect and String#dump don't raise, and use \xHH for invalid byte sequences Fixes #2159 Fixes #2130
This comparison is taking too long to generate.
Unfortunately it looks like we can’t render this comparison for you right now. It might be too big, or there might be something weird with your repository.
You can try running this command locally to see the comparison on your machine:
git diff 183fd6b5f62d...e276338875e5