I'm somewhat unsure about the AST structure here.
VariadicPlaceholder is not a general expression. Maybe Arg->expr
should be Expr|VariadicPlaceholder? Or possibly the call arguments
should be an array of Arg|VariadicPlaceholder?
With the introduction of intersection types, PHP now lexes the
token '&' either as T_AMPERSAND_(NOT_)FOLLOWED_BY_VAR_OR_VARARG.
This completely breaks parsing of any code containing '&'.
Fix this by canonicalizing to the new token format (unconditionally,
independent of emulation) and adjusting the parser to use the two
new tokens.
This doesn't add actual support for intersection types yet.
Empty productions are supposed to be assigned the start attributes
of the lookahead token. Currently, this happens by assigning above
the current stack position when the token it read.
This fails in a situation where we first reduce an empty production
higher up in the stack, and then again reduce an empty production
lower in the stack, without consuming the lookahead token in the
meantime.
Fix this by moving the assignment into the reduction phase. We
also need to do this for error productions, which are effectively
empty.
Adds support for PHP 8 attributes, represented using `AttrGroup` nodes
containing `Attribute` nodes. The `attrGroup` subnode is added to all
nodes that can have attributes.
This is still missing FPPP support.
Co-authored-by: Nikita Popov <nikita.ppv@gmail.com>
Similar to the previous commit: list() syntactically accepts any
expression and non-variables are compile-time errors. The special
case of ($a) ends up being legal by accident.
Partial fix for #653. PHP 7 allows expressions inside isset(), but
rejects non-variables in the compiler. A side-effect of this is that
isset(($x)) is allowed, though this is not intentional.
The token stream should cover all characters in the original code,
insert a dummy token for missing illegal characters. We should
really be doing this in token_get_all() as well.
Previously zero-length nop nodes used the lookahead start attributes
and current end attributes. This choice ends up being somewhat weird,
because the end attributes will be the at the last non-whitespace,
non-comment token, which might be quite far back. More problematically,
we may not have encountered any non-discarded token if we're at the
start of the file, in which case we will have no end attributes to
assign.
Change things to use a canonical "zero-length" node representation,
where the end position (token & file) will be exactly one before the
start position.
Fixes#589.
