Lexer diagnostics used a single `const char*` as their source location,
which only allowed pointing at a single character rather than a run of
characters. This limited the underlining of unrecognized tokens, and
also made escape sequence diagnostics particularly strange by stopping
at whatever point the lexer discovered the error and pointing only at
that location. For example: `"\w"` was reported at the `w`, and `"\x"`
at the closing `"`.
This change tries to always keep the primary location of a diagnostic
the point at which typing (inserting or deleting) might fix the error.
So for `"\x"`, `\x` is underlined, but the position is at the `"` as
that's where you would type a hex code.
The lexer's emitter now takes a special lexing source location that
represents either a single position _between_ two characters, a single
character, or a range of the source buffer. The lexer diagnostics are
updated to construct these for the character range as appropriate.
Parser diagnostics for a missing token were located at the _next_ token,
which is particularly confusing for terminators like missing `;`s. Now
they are located at a position after the previous token, with or without
a space depending on the token, but typically on the same line.
Diagnostics still sort by how far the parser has read.
The parser diagnostics for unclosed groups also now attaches a label to
the opening token when it is on a separate line. And the diagnostics for
mismatched indent in a multi-line string attaches a label to the closing
`'''` that establishes that indent.
There are likely more location improvements we should do here, but this
seemed like a reasonable incremental improvement and is what Claude
found when doing an initial audit. One example where I think we should
make further improvements is around suggesting raw identifiers when we
find an unexpected `Cpp` or `Core` or `base`. We do this consistently
now, but raw identifiers should be rare. A TODO is added for these
cases.
Assisted-by: Claude Code
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Diagnostic rendering was part of the data model: `Loc::FormatLocation`
and
`Loc::FormatSnippet` wrote plain text to a stream, so no capability,
color, or
terminal width could reach them.
A diagnostic is now one message with labels, contexts, and location
information
attached, and `Diagnostics::Renderer` draws it into a
`Terminal::Buffer`. The
design is in `toolchain/docs/diagnostics_rendering.md` (#7668),
authoring in
`toolchain/docs/diagnostics.md`.
Before:
```
example.carbon:6:3: error: 1 argument passed to function expecting 0 arguments
Run0(1);
^~~~~~~
example.carbon:3:1: note: calling function declared here
fn Run0() {}
^~~~~~~~~~~
```
After:
```
error: 1 argument passed to function expecting 0 arguments
╭─┤ example.carbon:6:3
┆
3 │ fn Run0() {}
· ─────┬─────
· ╰──┤ calling function declared here
┆
-> 6 │ Run0(1);
· ───────
╰────────────
```
`--no-diagnostic-snippets` gives a compact form, `example.carbon:6:3-9`.
File
tests use it; the rendering tests under `toolchain/diagnostics/testdata`
ask for
snippets.
C++ interop diagnostics are data now rather than text Clang drew:
`SemIR::ClangSourceLocId` names a `clang::SourceRange`, and a Clang
location is
marked as the point it names.
`StreamConsumer` moves to a `:renderer` target that `:emitter` does not
depend
on, keeping `common/terminal` out of `consumer.h`. Capabilities come
from the
file behind the error stream, so tests get plain ASCII.
`renderer_fuzzer` draws diagnostics built from fuzzer bytes with every
capability combination, since a shape the renderer doesn't expect must
never be
why the compiler dies while reporting a problem.
TODOs mark the follow-ups: line weights and crossing lines, wrapping
that keeps
a quoted type whole, and ordering diagnostics that share a byte offset.
Also: `IncorrectExtension`'s hint moves into its message, since its note
had no
location of its own, and `strip_noise` in `check_diagnostics.py` blanks
multi-line `#define`s instead of deleting them.
Assisted-by: Claude Code
Adds an algorithm to compute where to insert brackets to repair
bracketing mismatches during lexing. This takes indentation, as well as
a number of other cues, into account to predict where the brackets
should have gone. Detects when there is ambiguity between solutions and
makes no suggestion in that case. Reduces the problem by splitting on
properly bracketed top-level constructs, then uses a beam search to find
good candidate solutions quickly.
This includes both a fuzzer and an eval tool that can be used to
determine how well the algorithm fares against a given corpus of valid
Carbon code, by damaging it in various ways and seeing whether the
algorithm can correctly fix it. On all the eval modes, this algorithm
can correctly infer the positions for over 80% of lost brackets (and can
correctly restore 95+% of brackets in some modes), with low rates of
incorrect suggestions.
See added documentation for full details.
Assisted-by: Gemini via Antigravity, Claude via Claude Code
It's hard to track 2-space indent levels across large vertical gaps, and
it can be hard to suppress the instinct to read the dump as if it were
preorder. This change introduces a new dump mode that addresses both
problems by using box-drawing characters to explicitly represent the
parent-child edges of the tree. This new mode is the default, but the
old behavior remains available with
`--parse-dump-format=yaml-postorder`.
This is in addition to finding a `where` on the RHS of another `where`.
Since a generic binding introduces `.Self`, any `where` expression that
isn't part of a facet type modifying the binding itself would introduce
an ambiguous `.Self`.
Add virtual parse nodes for let, var, and form bindings, which goes
before the type. This allows us to track if `where` appears in the
binding's type. We only need to look for an invalid `where` if any
appeared in the type. We combine these three nodes together into a
single node kind, which requires us to remove the name from it as a
child. We move it up to the Pattern node again, and rename the
PatternStart nodes to PatternTypeStart as they are now located in the
middle of the Pattern nodes, just before the type.
And we only need to thaw `.Self` in generic bindings. Non-generic
bindings can only have `.Self` through a `where` expression, since the
name is not provided otherwise to non-generic bindings. And `where`
expressions thaw their `.Self` independently. So the binding only needs
to thaw a `.Self` that it introduced, which is only for generic
bindings.
Carbon currently requires a comment to be the only non-whitespace on its
line. A `//` comment that follows other content on a line, called a
_trailing comment_, is a lexer error. This proposal removes that
restriction, allowing a comment to follow other content on a line.
Everything else about comments is unchanged: a comment still begins with
`//`, still requires whitespace after the `//`, and still runs to the
end of the line. Carbon continues to provide only line comments; no
block or intra-line comments are added.
Three observations motivate the change. First, trailing comments are
well suited to short _annotations_ attached to a specific entity or
value on a line. Second, the lexer design now makes it trivial to lex
trailing comments, and in fact requires extra logic and potentially cost
to reject them. Third, C++ code routinely uses trailing comments, so
allowing them lets Carbon carry the layout of migrated code over
directly, rather than reworking each comment to read well in a different
structure.
Implementation notes (beyond the proposal's design):
Keeping trailing comments cheap to lex required a few supporting
changes, all of which keep the cost off the lexer's hot path:
- The lexer already dispatches `//` to comment lexing wherever it
appears, so classifying a comment as trailing is a single O(1) check of
whether the `//` is the line's first non-whitespace (`start + indent`).
The hot comment path is otherwise unchanged.
- That check relies on each line's recorded indentation being its real
leading whitespace. Multi-line string literals previously recorded the
column where the literal opened for the lines they span; they now record
the true (closing-delimiter) indentation instead.
- Parser error recovery (`SkipPastLikelyEnd`) had relied on that
opening-column indentation to keep tokens following a multi-line string
literal attached to the same construct. It now reconstructs that
relationship directly by consulting the line on which the literal
opened, including when other tokens follow the closing delimiter (such
as `''' + "more"`). This is on the cold recovery path.
- `CommentData` records the trailing bit in the high bit of its length
field, keeping it at 8 bytes.
Assisted-by: Claude Code
If a keyword or a sized type literal (eg, `f2`) is used in a context
where we are confident that we are expecting an identifier -- either
before a `:` in a binding pattern or after a `.` in a member access or
designator -- then recover as if a raw identifier was used.
This appears to be a particular stumbling block for coding agents, so
seems worth paying special attention to.
Add a mechanism to the tokenized buffer to track additional tokens
synthesized for error recovery so that we can keep the lexed token
sequence immutable and still satisfy the invariants throughout the rest
of the toolchain for recovery tokens. Thanks to chandlerc for suggesting
this approach!
When parsing a pattern, if we encounter something that isn't pattern
syntax, try parsing as an expression instead. We only need one-token
lookahead to distinguish pattern syntax from expression syntax.
Track a precedence group through pattern parsing so that we can allow
different kinds of expressions in a top-level pattern (such as the
operand of `let`) and in a nested pattern (such as a subpattern of a
tuple pattern or within grouping parens). For example, we do not allow
`case if ...`, and for now I've chosen to also not allow logical or
relational operators at the top level of a pattern, so `case 1 + 1` is
OK, but `case 1 == 1` and `case true and false` require parentheses.
This decision should be ratified or revisited by a design proposal.
Very basic check support is also provided, only sufficient to form an
`ExprPattern` instruction and nothing beyond that. For now, all pattern
matching against an `ExprPattern` fails with a TODO error. To support
that, I've switched from calling `BeginSubpattern` in the parent handler
of a pattern and `EndSubpatternAs*` in the pattern handler itself to
calling both functions in parent handlers, with `EndSubpattern`
converting an expression into an expression pattern where needed.
Depends on #6976.
Assisted-by: Gemini via Google Antigravity
This change ensures that a function signature always starts with an
`IdentifierNameMaybeBeforeSignature` node (renamed from
`IdentifierNameBeforeParams`), even in the case of function declarations
like `fn F -> T` that have no parameter list. As a consequence, this
ensures that we push new entries onto `pattern_block_stack` and
`full_pattern_stack` when we start processing the function signature.
Add a `Core.String` class to the prelude representing a string view, and
rename the `String` keyword to `str` and make it evaluate to
`Core.String`.
`Core.String` is represented as a pair of a pointer to a character
(actually, to the first character of a string, but we don't have a way
of modeling that yet) and a size (which should be pointer-width, but is
currently always a `u64` as we don't have a `usize` equivalent yet).
`Core.String` values are generated directly by the toolchain for string
literal expressions.
This follows the direction established at the recent summit, but the
design implemented here has not been through the proposal process yet.
toolchain/check/testdata/builtins/char/basics.carbon and
toolchain/lower/testdata/builtins/char.carbon are probably the most
interesting tests here. The parse tests is required because this adds a
new node kind, and we need coverage of it; but the attached info is
minor. There's a fair amount of test churn here because I'm adding the
Core.Char and Core.CharLiteral types as new singletons.
My intent here is that `CharId` is always a unicode code point, even
when the type is a `Char` and thus must be a single UTF-8 code unit
(single byte). This mainly means the stored value of a `CharValue` can
be printed internally without knowing the type.
---------
Co-authored-by: josh11b <15258583+josh11b@users.noreply.github.com>
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Adds minimal defaults for codegen so that those tests don't need to
specify the full command line.
Moves driver/testdata/compile to check/testdata/basics/raw_sem_ir, which
seems like it better reflects the focus. Removes the textual IR test
because we have plenty of those now.
Moves a multiline token diagnostic test to parse because (a) diagnostics
doesn't have other tests and (b) this is really testing the way that
parse structures the location.
Drops `--include-diagnostic-kind` from some driver tests that aren't
testing a diagnostic, so the flag felt a bit like noise (even before
this change, there are a few tests that don't specify it because they're
not intended to test an emitted diagnostic, just high-level diagnostic
behaviors).
Adds some comments to reflect my understanding of why a test exists,
when I was pausing to think about it.
Modifies the stdin test to do more dumps; which happens to expose we
currently misbehave.
In line with the proposal in #4682, this changes the array syntax to be
array(T, N). `array` is a builtin keyword which must be followed by
parens containing two expressions and a separating comma.
The array type expression is still fully builtin, it does not forward to
a Core.Array library type yet. It merely adds the `ArrayType`
instruction, as was done with the previous syntax.
Followup work will change the instruction to reference to Core.Array,
once the library type exists and can be used directly.
---------
Co-authored-by: zygoloid <richard@metafoo.co.uk>
Parameter lists need substantially different treatment than tuple
patterns in other contexts, so this change splits them into separate
parse node kinds.
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
This removes some churn when adding new diagnostic cases to test files
(where previous to this change the newly added newline would cause the
previous diagnostic CHECKs to be updated including changes to the line
number because the CHECK for the blank line meant an extra line between
CHECK and source line).
A few alternatives discussed here:
https://discord.com/channels/655572317891461132/655578254970716160/1329573358475673723
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
This change splits `NodeKind::IdentifierName` into separate node kinds
depending on whether the identifier is followed by parameters, and
similarly splits `NameQualifier` based on whether the qualifier has
parameters. This enables us to only push a pattern block when it's
actually needed, rather than "defensively" pushing one when it might be
needed.
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
Split StructComma into StructLiteralComma and StructTypeLiteralComma in
order to easily differentiate handling (remains the same in this PR).
Add "Literal" to StructField and StructTypeField because it feels
inconsistent versus the other non-shared things. StructFieldDesignator
remains shared between value literals and type literals.
Note I probably would've made StructFieldDesignator non-shared too, but
that'd require either a lookahead of 2 (to see the separator`) or a
writeback after parsing the separator, neither of which felt especially
crucial for this, when what I'm really trying to do is split type
literal handling a little further.
Also surround it in square brackets rather than parentheses. This
matches the format used by Clang and GCC, and means diagnostics will
still match the `file:line:col: error: ` pattern used by some IDE tools.
Before:
```console
fail_builtins.carbon:11:11: error(AliasRequiresNameRef): alias initializer must be a name reference
```
After:
```console
fail_builtins.carbon:11:11: error: alias initializer must be a name reference [AliasRequiresNameRef]
```
Also tighten up test regex to only match on `STDERR` lines that list a
file name.
This is to help identify which diagnostics we're actually using.
Note that driver/testdata still has tests which don't pass this flag,
and so continue to test the kind-less (default) behavior.
This is a primarily automated change:
- Search & replace for capitalization
-
`(CARBON_DIAGNOSTIC\((?:\n\s+)?\w+,(?:\n\s+)?\s\w+,(?:\n\s+)?\s")([A-Z])`
- `$1\L$2`
- Search & replace for period
-
`(CARBON_DIAGNOSTIC\((?:\n\s+)?\w+,(?:\n\s+)?\s\w+,(?:\n\s+)?\s"(?:[^)]|\n)+)\.("[,)])`
- `$1$2`
- Limited search & replace for `ERROR: ` -> `error: ` in streamed things
- Leaving a TODO for command_line because there's more cleanup that can
be done there
- Modify diagnostic_consumer.cpp
- ERROR -> error
- WARNING -> warning
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
This has is a nice-to-have for me. Frequently I want to run a specific
test, and end up digging through output to be able to copy-paste the run
line. This uses TIP lines to inject the command into the file when using
AUTOUPDATE.
Note, one of the reasons I want this is because "bazel test
//toolchain/testing:file_test --test_output=all" has been regularly
exceeding bazel's output limit for me (workaround is either opening the
output file or specifying an obscure output limit flag), making it a
little harder for me to get the commands. However, frequently I'm adding
a file and want to iterate on it, so that's really the use case I have
in mind here.
The expected parse tree size is validated for valid parse trees, which
is possible even if lex encounters an error. In that case, we were
missing recovery tokens, resulting in a crash. AddToken should be the
only place we call `token_infos_.push_back`.
I've been thinking about this since we decided to add placeholders in
the parse tree. This allows a clearer division of work in check
handling, where we were doing work for ExprOpenParen that's only
necessary for tuples (splitting/renaming handle_paren.cpp accordingly).
The purpose of the newline is to make it clearer where a given
diagnostic begins and ends, particularly as the first message of a
diagnostic may not be the error.
This is a trivial code change, but ripples edits through test files.
This undoes parts of #3515 in order to allow PushGlobalInit to be called
when the initializer is called, instead of at the end of the binding
pattern. The current approach is fragile because supported patterns will
become more complex. We also will likely want similar support in `let`,
which puts the initializer first, so this offers a consistent approach
for both.
[Looking
back](https://discord.com/channels/655572317891461132/655578254970716160/1184237511766179840),
this is more or less the second option in that message, but using the
PeekNextIs to avoid vagueness about what's being popped first.
Note I'm putting in PeekNextIs for what I'm hoping will be a pretty
narrow use-case. I could've added depth arguments to the Peek functions,
but that would've rippled through a number of APIs and it's not clear to
me that this has generic utility. I mean, right now it could just be
PeekNextIsVariableInitializer, since it's only optional in that case.
Also stop supporting `var` with initializer inside `for`.
Resolves TODO in `handle_variable.cpp`
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
(Split out of #3410)
Consistently use `ParenExpr` solely for parenthesized single
expressions, and use more syntax-oriented terminology for states and
nodes that might represent either a `ParenExpr` or a tuple literal.
This is supporting a direction that all parse nodes should correspond to
a single token, allowing for reduced tokenized buffer access during
checking (it's still necessary for diagnostics, and some literals).
One of the justifications for a unified parse node was implementation
LOC: note this is slightly smaller, using macros to reduce some
duplication. While this does add more switching in HandleDeclScopeLoop,
that's offset by less explicit switching in the check handlers. Also, I
think the duplication in HandleDeclScopeLoop can be reduced by shifting
the flow there, which I'll do in a separate PR.
I was suggesting this because `FloatingPoint` is pretty long. `int` and
`float` should be familiar abbreviations. `unsigned` should be familiar
to developers too, but `UnsignedInt` still feels usefully clearer for
the additional chars.
This is an incremental improvement on our diagnostic messages that
simply underlines an entire token if the token is larger than 1 char
(else it points to the single char with a caret like it used to).
This removes a (very) hot branch in the lexer where we need to special
case when a token is the first token and can't look at its previous
token. It also seems like a generally nice change to the structure of
both the token buffer and parse tree as there are now bracketing
elements for both ends and we should be able to avoid similar branching
in the future.
Mostly mechanical updates to the lexer and parser code to handle this,
but also needed to special case the location information in the
autoupdate code. And then the usual large body of auto-updated tests.
No benchmark data for this change alone as in isolation and in the
current lexer structure it doesn't make a big difference. But this
branch was particularly difficult to handle when trying to update the
whitespace skipping code to be faster, and so I think it is worth
systematically avoiding the special case here.
This makes the difference between errors and lower-level diagnostics
visible to users, and aligns the toolchain's behavior with the
expectations in `driver_fuzzer.cpp`.
Chandler CarruthandRichard Smith
Geoff Romer
Dana Jansens
David BlaikieandJon Ross-Perkins
Brymer Meneses
Jacob Schneider