Subject | Re: [Firebird-Architect] Re: the insert into t select * from t infinite |
---|---|
Author | Jim Starkey |
Post date | 2005-07-05T22:03:21Z |
Dmitry Yemanov wrote:
semantic analysis => optimization => code generation.
The parser's only job is to return a parse tree that reflects the
language string. Anything else is a mistake. Please believe me. I've
made it more times than I care to admit. A parse is -- and should be --
too stupid to have context to do effective tweaks.
--
Jim Starkey
Netfrastructure, Inc.
978 526-1376
>"Jim Starkey" <jas@...> wrote:Semantic analysis. Classical language processing steps: parse =>
>
>
>>Parsers don't do transformations. Parser parse. Period.
>>
>>
>
>What name should have PASS1 then, considering its logic? :-)
>
>
semantic analysis => optimization => code generation.
The parser's only job is to return a parse tree that reflects the
language string. Anything else is a mistake. Please believe me. I've
made it more times than I care to admit. A parse is -- and should be --
too stupid to have context to do effective tweaks.
>Nevermind. There should be a separate query rewrite code in the engine.This isn't the time.
>Currently it's splitted between PASS1, CMP and OPT. This should be cleaned
>up some time.
>
>
>
--
Jim Starkey
Netfrastructure, Inc.
978 526-1376