Re: error: parse error

[Mats Bengtsson]

> While editing lilypond input files, I often (although with more practice not 
> quite as much) have errors.  The most of the time results in the following:
> 
> error: parse error
> 
> Not only is this completely redudant (if something bad happened after 
> lilypond said it was parsing, it must obviously be a parse error), it 
> provides little suggestions for the cause of the error.  The most common 
> cause of this is any of ( ) \( \) ~ \< \> \! < or > being in the wrong order. 
>  Usually changing the order of these fixes the problem, although forgetting a 
> > can be extremely difficult to pin down.
> 
> One possible solution to this problem would be to improve the emacs mode.  
> For example, after using color syntax highlighting and automatic indentation 
> for Java programs, I eliminate a lot of the standard typographical errors 
> (oops, forgot that semi-colon) before I even go to compile the code.  It 
> would be spiff if the lilypond mode could somehow know which symbols belong 
> where and maybe even prevent you from adding extra \property \revert 
> statements.
> 
> Peter Dobratz

Agreed! The automatic matching of {}, [], <> and () in Emacs
provides some help in the debugging, but fails to understand
some correct code, such as "[c d ( e f ] g a b ) c".
Of course, the best solution would be a smarter parser that
could point out the error.

  /Mats