AfterShip / clickhouse-sql-parser / 27487335859

53%

Build Type

push

github

Committed by web-flow

Commit Message

Rename lexer state to match what it holds (#277)

l.lastToken and p.last() actually hold the *current* lookahead token -
the one the parser is looking at but has not consumed - not a
previously-consumed token. Likewise l.current held the input byte
offset, not a token. Rename for accuracy, no behavior change:

- lexerState.current (int)    -> offset
- lexerState.lastToken        -> current
- Parser.last()               -> cur()
- Parser.lastTokenKind()      -> curTokenKind()
- Parser.lastTokenString()    -> curTokenString()
- local `lastToken` vars      -> curToken (parser files)

The one genuinely-previous token - the local in Lexer.consumeToken used
to disambiguate unary +/- - is named prevToken.

Also give TokenKindDot its missing TokenKind type annotation.
---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

Coverage Stats

160 of 231 new or added lines in 8 files covered. (69.26%)

1 existing line in 1 file now uncovered.

8495 of 16105 relevant lines covered (52.75%)

2997.35 hits per line