[NEM] '''3Mar2005''': Here's a very simple parser for the [BibTeX] bibliography database format. It's rather braindead at the moment, and is really not much more than a tokeniser. To use it, call bibtex::parse with the contents of a database, and you'll get back a list of entries. Each entry is itself a list of three elements: type (e.g. article, thesis, etc), key (whatever key it was given in the database), and a final list of key/value pairs (in [dict] form) representing the record. It doesn't handle BibTeXs (or TeXs) insane file format fully, and doesn't handle @string macros etc. There may well be further restrictions on what it can cope with. It suffices for what I need it for though -- a quick and dirty hack to dump a bibtex db to a website. Anyway, enjoy!

 # bibtex.tcl --
 #
 #	A basic parser for BibTeX bibliography databases.
 #
 # Copyright (c) 2005 Neil Madden.
 # License: Tcl/BSD style.
 
 package require Tcl 8.4
 package provide bibtex 0.1
 
 namespace eval bibtex {
 
    proc parse {bibtex} {
	set ret [list]
	foreach block [lrange [split $bibtex @] 1 end] {
	    if {[regexp {([^\{]+)\{([^,]+),(.*)\}[^\}]*} $block -> type key rest]} {
		lappend ret [list [Tidy $type] [string trim $key] [ParseBlock $rest]]
	    } else {
		puts stderr "Skipping: $block"
	    }
	}
	return $ret
    }
    proc Tidy {str} {
	string tolower [string trim $str]
    }
 
    proc ParseBlock {block} {
	set ret [list]
	set index 0
	while {1} {
	    if {[regexp -start $index -indices -- {(\S+)[^=]*=(.*)} $block -> \
		key rest]} {
		
		foreach {ks ke} $key { break }
		set k [Tidy [string range $block $ks $ke]]
		foreach {rs re} $rest { break }
		foreach {v index} \
		    [ParseBibString $rs [string range $block $rs $re]] \
		    { break }
		lappend ret $k $v
	    } else { break }
	}
	return $ret
    }
 
    proc ParseBibString {index str} {
	set count 0
	set retstr ""
	set escape 0
	foreach char [split $str ""] {
	    incr index
	    if {$escape} {
		set escape 0
	    } else {
		if {$char eq "\{"} {
		    incr count
		    continue
		} elseif {$char eq "\}"} {
		    incr count -1
		    if {$count < 0} { incr index -1; break }
		    continue
		} elseif {$char eq ","} {
		    if {$count == 0} { break }
		} elseif {$char eq "\\"} {
		    set escape 1; continue
		} elseif {$char eq "\""} { continue }
		append retstr $char
	    }
	}
	regsub -all {\s+} $retstr { } retstr
	return [list [string trim $retstr] $index]
    }
 }
----
[schlenk] Well spoken, the file format is just insane...

----
[Category Application]
[Category Word and Text Processing]