...
|SongInfo|Title:"Speed the Plow"|Author:"Trad."|Lyricist:""|Copyright1:""|Copyright2:""
...
|Clef|Type:Treble
|Key|Signature:F#,C#
|TimeSig|Signature:4/4
|Bar|Style:MasterRepeatOpen
|Note|Dur:8th|Pos:-2|Opts:Stem=Up,Beam=First
|Note|Dur:8th|Pos:-1|Opts:Stem=Up,Beam
etc.
I am assuming I will have two textareas in a form, the data are copied from a clipboard to one of them and split into an array on newline characters. On the last few lines the letters A-G or a-g may or may not be preceded by _, =, or ^ and may or may not be followed by a comma or apostrophe then a (or not) number. The lack of a space after a letter indicates a grouping (beaming).
Does anyone have an example of some parsing code? Just something that looks at the input byte by byte building a token, putting a character back if it belongs to the next token and handling the end of the line and end of data. If someone has a parser for anything, I think I could adapt it.
Sounds confusing. Any parsing process is based on an algorithm. Set that algorithm for your particular case, and you have solved 90% of the problem. And the method is nothing but an equation, after all. What example do you need? Of an algorithm?
I am not sure what example to give... Maybe parsing a RGB value of a color to its Hexadecimal correspondent. Here's a simplified version of a color picker which output RGB to Hexa:
But this one uses a native method of JavaScript toString(base) which converts decimal numbers into any base, up to 36. Your case is strongly particular, thus you have to create your own "table of correspondence" between n pairs of values and your own algorithm which will be able reproduce that relationship even for n+1 pairs.
My only reference was to a compiler class I took back in 1994. The final product would look at a line such as
Code:
xrad= x * PI / 180.0;
to fetch the variables x and PI, multiply them, then divide by a constant in order to place the result in xrad, creating the commands to do each of these things. If reading a variable it continued until it found a character which could not be in a variable, or a space, or end-of-line. (xrad is different from x, so it has to check the byte(s) after the x to see if there is more to the name).
In this case c may stand alone or ^c'2/3 may be one token. A "|" means one thing by itself and "|:" does something different. If someone had some type of model I could use it and it would be much easier to modify later than if I had to do it from scratch.
Reply With Quote
Originally Posted by Fang
Bookmarks