They were a vestige of when Tree-sitter did sentential form-based
incremental parsing (as opposed to simply state matching). This was
elegant but not compatible with GLR as far as I could tell.
The previous approach to error recovery relied on special error-recovery
states in the parse table. For each token T, there was an error recovery
state in which the parser looked for *any* token that could follow T.
Unfortunately, sometimes the set of tokens that could follow T contained
conflicts. For example, in JS, the token '}' can be followed by the
open-ended 'template_chars' token, but also by ordinary tokens like
'identifier'. So with the old algorithm, when recovering from an
unexpected '}' token, the lexer had no way to distinguish identifiers
from template_chars.
This commit drops the error recovery states. Instead, when we encounter
an unexpected token T, we recover from the error by finding a previous
state S in the stack in which T would be valid, popping all of the nodes
after S, and wrapping them in an error.
This way, the lexer is always invoked in a normal parse state, in which
it is looking for a non-conflicting set of tokens. Eliminating the error
recovery states also shrinks the lex state machine significantly.
Signed-off-by: Rick Winfrey <rewinfrey@github.com>
The parser spends the majority of its time allocating and freeing trees and stack nodes.
Also, the memory footprint of the AST is a significant concern when using tree-sitter
with large files. This library is already unlikely to work very well with source files
larger than 4GB, so representing rows, columns, byte lengths and child indices as
unsigned 32 bit integers seems like the right choice.
* While generating the lex table, note which tokens can match the
same string. A token needs to be relexed when it has possible
homonyms in the current state.
* Also note which tokens can match substrings of each other tokens.
A token needs to be relexed when there are viable tokens that
could match longer strings in the current state and the next
token has been edited.
* Remove the logic for marking tokens as fragile on creation.
* Store the reusability/non-reusability of symbols off of individual
actions and onto the entire entry for the state & symbol.
Before, any syntax error would cause the lexer to create an error
leaf node. This could happen even with a valid input, if the parse
stack had split and one particular version of the parse stack
failed to parse.
Now, an error leaf node is only created when the lexer cannot understand
part of the input stream at all. When a normal syntax error occurs,
the lexer just returns a token that is outside of the expected token
set, and the parser handles the unexpected token.
* Use GLR stack-splitting to try all numbers of tokens to
discard until a repair is found.
* Check the validity of repairs by looking at the child trees,
rather than the statically-computed 'in-progress symbols' list
Previously, the lexer would operate in error-mode (ignoring any garbage input
until it found a valid token) if it was invoked in the 'error' state. Now that
the error state is deduped with other lexical states, the lexer might be invoked
in that state even when error-mode is not intended. This adds a third argument
to `ts_lex` that explicitly sets the error-mode.
This bug was unlikely to occur in any real grammars, but it caused the
node-tree-sitter-compiler test suite to fail for some grammars with only one
rule.
