That can happen because of anyone sending the process a SIGCONT.
Fixes an issue where continuing a process launched by the shell from
the System Monitor would cause the shell to spin on waitpid().
Since this is always set to true on the non-default constructor and
subsequently never modified, it is somewhat pointless. Furthermore,
there are arguably no invalid relative paths.
This will cause a problem when `NonnullRefPtr<T>::operator T*` will be
declared as RETURNS_NONNULL. Clang emits a warning for this pointless
null check, which breaks CI.
This commit converts naked `new`s to `AK::try_make` and `AK::try_create`
wherever possible. If the called constructor is private, this can not be
done, so we instead now use the standard-defined and compiler-agnostic
`new (nothrow)`.
This is so they can find their associated resources and it's
the same behavior as in Lagom.
This also required changing some tests so that they could
write their resources in a writable location.
Some of the code assumed that chars were always signed while that is
not the case on ARM hosts.
Also, some of the code tried to use EOF (-1) in a way similar to what
fgetc() does, however instead of storing the characters in an int
variable a char was used.
While this seemed to work it also meant that character 0xFF would be
incorrectly seen as an end-of-file.
Careful reading of fgetc() reveals that fgetc() stores character
data in an int where valid characters are in the range of 0-255 and
the EOF value is explicitly outside of that range (usually -1).
And use them to highlight javascript in HTML source.
This commit also changes how TextDocumentSpan::data is interpreted,
as it used to be an opaque pointer, but everyone stuffed an enum value
inside it, which made the values not unique to each highlighter;
that field is now a u64 serial id.
The syntax highlighters don't need to change their ways of stuffing
token types into that field, but a highlighter that calls another
nested highlighter needs to register the nested types for use with
token pairs.
This changes the Shell syntax highlighter to conform to the now-fixed
rendering of syntax highlighting spans in GUI::TextEditor.
This also adds some debug output to make debugging easier.
This replaces ctype.h with CharacterType.h everywhere I could find
issues with narrowing conversions. While using it will probably make
sense almost everywhere in the future, the most critical places should
have been addressed.
We had two functions for doing mostly the same thing. Combine both
of them into String::find() and use that everywhere.
Also add some tests to cover basic behavior.
Follow-on to #7337. Been seeing other CI test failures that point to
these temp directories, so let's just move all of them to /tmp. I'm sure
someone will write ext2fs stress tests later :^)
Example:
/usr/Tests/Shell/control-structure-as-command.sh
Core::Socket: Failed to connect() to /tmp/portal/inspectables: ...
+ rm -rf shell-test 2>/dev/null
+ mkdir shell-test
Error: The action has timed out.
Since applications using Core::EventLoop no longer need to create a
socket in /tmp/rpc/, and also don't need to listen for incoming
connections on this socket, we can remove a whole bunch of pledges!
This patch adds a new flag called history_dirty to Line::Editor that is
set when history is added to but written. Applications can leverage
this flag to write history only when it changes. This patch adds an
example usage of this functionality to Shell, which will now only save
the history when it is dirty.
Instead of the previous only-escape-with-backslashes, extend the
escaping to one of:
- No escape
- Escape with backslash
- Escape with "\xhh" if control character that isn't easily represented
as \X
- Escape with "\uhhhhhhhh" if unicode character that is too big to
represent as "\xhh".
Fixes#6986.
This commit makes the shell:
- highlight executables in the current directory as invalid, unless an
explicit `./' is given (so, `./foo` isn't red, but `foo` is)
- not suggest executables in the current directory unless explicitly
requested (by prepending `./`)
- not attempt to run an executable in the current directory that has
been invoked as a program name and failed execvp().
Note that `./foo` is still executed because it's not invoked as
a name, but rather as a path.
Fixes the other half of #6774.
This bit of code was kept unmodified since it was first implemented,
and I'm not entirely convinced that it ever actually worked :P
This commit updates the code to use "modern" classes and constructs,
and fixes an issue where the shebang would still contain the '#!'
when it was passed to execvp().
Fixes#6774.
Otherwise we would end up trying to parse the same heredoc entry, if it
contained a sequence terminated by a newline.
e.g. `<<-x\n$({` would attempt to read a heredoc entry after `x`, and
then after `{` while inside the first heredoc entry.
To make this work, we can simply empty the instance vector and keep the
state on the stack.
Issue found through oss-fuzz:
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=33852
Closes#4283.
Heredocs are implemented in a way that makes them feel more like a
string (and not a weird redirection, a la bash).
There are two tunables, whether the string is dedented (`<<-` vs `<<~`)
and whether it allows interpolation (quoted key vs not).
To the familiar people, this is how Ruby handles them, and I feel is the
most elegant heredoc syntax.
Unlike the oddjob that is bash, heredocs are treated exactly as normal
strings, and can be used _anywhere_ where a string can be used.
They are *required* to appear in the same order as used after a newline
is seen when parsing the sequence that the heredoc is used in.
For instance:
```sh
echo <<-doc1 <<-doc2 | blah blah
contents for doc1
doc1
contents for doc2
doc2
```
The typical nice errors are also implemented :^)
Some nodes (such as heredocs) cannot be validated immediately, so the
entire tree will need to be revalidated if we don't allow mutating
syntax errors.
