Please find below an EBNF grammar, XML schema, and XSLT transformations for Wiki Creole, currently in version 1.0. Wiki Creole is the first (and only) community standard for wiki markup.
These specifications were taken from the following two technical reports:
Wiki Creole EBNF grammar
Wiki Creole XML schema definition
Wiki Creole to XHTML transformation
XML to Wiki Creole transformation
It is pretty likely that despite our test suites, some bugs remain. So, if you find some, please let us know!
10 responses so far ↓
1 marko // Jan 21, 2008 at 10:08 am
The EBNF links are broken: From the first one “wiki-creole/” must be removed, from the latter the quotation marks at the end.
2 Dirk Riehle // Jan 21, 2008 at 10:21 am
@marko: Fixed, thank you.
3 Martien // Feb 12, 2008 at 8:42 pm
Generating code using ANTLRWorks 1.1.7 results in errors in rule text_bolditalcontent.
I am a absolute novice on the subject and appreciate your help.
4 Martin // Feb 12, 2008 at 10:27 pm
@Martien
ANTLRWorks does not always show the same behavior as ANTLR does. I couldn’t use ANTLRWorks for development as the grammar size magnified. Further you’ll probably have to scale up the size of heap memory for code generation, e.g. use the JVM option -Xmx1024M.
5 Martien // Feb 13, 2008 at 12:14 pm
Thanks Martin
Problem solved when generating from the command line using org.antlr.tool.
6 V // May 2, 2008 at 6:28 pm
Im trying to implement wikicreole using ocamlyacc or menhir (an LR1 parser for OCaml). But Wikicreole’s EBNF is using some ANTLR specific features (I think only this input.LA thing) that makes difficult to adapt to another parser generator … Does somebody have an idea about that?
7 Martin // May 2, 2008 at 7:57 pm
V, first you have to consider that ANTLR generates LL(*) parsers, in contrast to your LR(1).
We had to use the explicit input.LA statements since ANTLR chooses a derivation too early, i.e. even if other derivations are not wrong in respect to the current look-ahead. It does not increase the look-ahead as long as necessary to identify the only applicable derivation. Thus using the wrong derivation will result in an exception, even though the grammar is correct.
The _grammar_ does not need these explicit queries to the look-ahead. But the _ANTLR grammar specification_ seems to need it auxiliary.
If I recall LR(1)-parsers correctly, you can’t rely on the grammar specification for your parser generator only, because the WikiCreole language specification needs more look-ahead than 1.
Nevertheless, good luck for your work!
8 Greg // May 15, 2008 at 11:42 pm
Hi guys. I’m interested in using this XML format as my intermediate (stored) stage. However, it seems that I’ll have to write the code (Java) to translate from Creole to ‘X-Creole’. I realise there’s an XSD, but would some of you have some example XML files for me to test with? Thanks!
9 Martin // May 16, 2008 at 5:13 am
Hi Greg, you are right. We do rely on some code to transform a representation of the wiki after parsing to the XML representation. I’m not sure whether you can transform the parse tree (or AST) to XML by ANTLR’s domain specific language. We implemented a Java DFS parse tree walker that generates the XML file simultaneously. Regards, and I mail some files to you.
10 Dirk Riehle // May 16, 2008 at 9:23 am
Greg, Martin: I had been wondering about this for a while—we knew that the tree walker is missing from our publications, but then you can’t really publish some Java code.
What you can do, of course, is create an open source project. So for that reason I created a SourceForge project called wikicreole a while back. Is there any interest on writing Tree Walker code for that project?
Leave a Comment