A formal grammar for parsing GEDCOM 5.5x files

The following LALR(1) grammar can be used to parse GEDCOM 5.5x files generated by a wide variety of programs. Pointer strings, if used, must be manually trimmed to remove prefix and suffix characters ('@'). Likewise, embedded escape sequences in value strings must be manually extracted.

ALNUM:   // an alphanumeric character

[a-z_A-Z0-9]

 

WHITE:   // any whitespace character

[ \t\v\b\f\a]

 

ANY:    // any character except newline

[^\n]

 

POINTER:   // a pointer string

@ALNUM[^@]*@

 

BOM8:   // a UTF-8 byte order mark

\xEF\xBB\xBF

 

UBOM:   // a Unicode byte order mark

\xFF\xFE

 

NEWLINE:

\n

 

SPACE:

' '   // single space character (0x20)

 

TAG:

ALNUM+

 

LEVEL:

0|([1-9][0-9]*)

 

VALUE:

ANY+

%ignore%:

BOM8    // UTF-8 byte order mark

UBOM    // Unicode byte order mark

WHITE* // leading whitespace

 

file:

line

file line

line:

LEVEL SPACE TAG

LEVEL SPACE TAG line_value NEWLINE

LEVEL SPACE POINTER SPACE TAG

LEVEL SPACE POINTER SPACE TAG line_value NEWLINE

line_value:

VALUE

POINTER