notation3		index
/devel/WWW/2000/10/swap/notation3.py

$Id: notation3.html,v 1.41 2004/10/28 17:41:59 timbl Exp $
 
 
This module implements basic sources and sinks for RDF data.
It defines a stream interface for such data.
It has a command line interface, can work as a web query engine,
and has built in test(), all of which demosntrate how it is used.
 
To make a new RDF processor, subclass RDFSink.
 
See also:
 
Notation 3
http://www.w3.org/DesignIssues/Notation3
 
Closed World Machine - and RDF Processor
http;//www.w3.org/2000/10/swap/cwm
 
To DO: See also "@@" in comments
 
Internationlization:
- Decode incoming N3 file as unicode
- Encode outgoing file
- unicode \u  (??) escapes in parse
- unicode \u  (??) escapes in string output
 
Note currently unicode strings work in this code
but fail when they are output into the python debugger
interactive window.
 
______________________________________________
 
Module originally by Dan Connolly, includeing notation3
parser and RDF generator. TimBL added RDF stream model
and N3 generation, replaced stream model with use
of common store/formula API.
 
DWC:
oops... I'm not doing qname expansion as described
there (i.e. adding a # if it's not already there).
 
I allow unprefixed qnames, so not all barenames
are keywords.
 
---- hmmmm ... not expandable - a bit of a trap.
 
I haven't done quoting yet.
 
idea: migrate toward CSS notation?
 
idea: use notation3 for wiki record keeping.

 
Modules
            
RDFSink
codecs
diag
re
string
sys
types
uripath
urllib
 
Classes
            
RDFSink.RDFSink
ToN3
SinkParser
exceptions.SyntaxError(exceptions.StandardError)
BadSyntax
 
class BadSyntax(exceptions.SyntaxError)
       
  
Method resolution order:
BadSyntax
exceptions.SyntaxError
exceptions.StandardError
exceptions.Exception
Methods defined here:
__init__(self, uri, lines, str, i, why)
__str__(self)
Data and non-method functions defined here:
__doc__ = None
__module__ = 'notation3'
Data and non-method functions inherited from exceptions.SyntaxError:
filename = None
lineno = None
msg = ''
offset = None
print_file_and_line = None
text = None
Methods inherited from exceptions.Exception:
__getitem__(...)
 
class SinkParser
       
   Methods defined here:
__init__(self, store, openFormula=None, thisDoc='', baseURI=None, genPrefix='', metaURI=None, flags='', why=None)
 note: namespace names should *not* end in #;
the # will get added during qname processing
anonymousNode(self, ln)
Remember or generate a term for one of these _: anonymous nodes
bareWord(self, str, i, res)
        abc -> :abc
bind(self, qn, uri)
checkDot(self, str, i)
commaSeparatedList(self, str, j, res, what)
return value: -1 bad syntax; >1 new position in str
res has things found appended
directive(self, str, i)
directiveOrStatement(self, str, h)
endDoc(self)
Signal end of document and stop parsing. returns formula
feed(self, octets)
Feed an octet stream tothe parser
 
if BadSyntax is raised, the string
passed in the exception object is the
remainder after any statements have been parsed.
So if there is more data to feed to the
parser, it should be straightforward to recover.
formula(self)
here(self, i)
String generated from position in file
 
This is for repeatability when refering people to bnodes in a document.
This has diagnostic uses less formally, as it should point one to which 
bnode the arbitrary identifier actually is. It gives the
line and character number of the '[' charcacter or path character
which introduced the blank node.  The first blank node is boringly _L1C1.
It used to be used only for tracking, but for tests in general
it makes the canonical ordering of bnodes repeatable.
item(self, str, i, res)
loadBuf(self, buf)
Parses a buffer and returns its top level formula
loadStream(self, stream)
makeStatement(self, quadruple)
node(self, str, i, res, subjectAlready=None)
Parse the <node> production.
Space is now skipped once at the beginning
instead of in multipe calls to skipSpace().
nodeOrLiteral(self, str, i, res)
object(self, str, i, res)
objectList(self, str, i, res)
path(self, str, i, res)
Parse the path production.
prop(self, str, i, res)
property_list(self, str, i, subj)
Parse property list
Leaves the terminating punctuation in the buffer
qname(self, str, i, res)
xyz:def -> ('xyz', 'def')
If not in keywords and keywordsSet: def -> ('', 'def')
:def -> ('', 'def')
setKeywords(self, k)
Takes a list of strings
skipSpace(self, str, i)
Skip white space, newlines and comments.
return -1 if EOF, else position of first non-ws character
startDoc(self)
statement(self, str, i)
strconst(self, str, i, delim)
parse an N3 string constant delimited by delim.
return index, val
subject(self, str, i, res)
tok(self, tok, str, i)
Check for keyword.  Space must have been stripped on entry and
we must not be at end of file.
uEscape(self, str, i, startline)
uri_ref2(self, str, i, res)
Generate uri from n3 representation.
 
Note that the RDF convention of directly concatenating
NS and local name is now used though I prefer inserting a '#'
to make the namesapces look more like what XML folks expect.
variable(self, str, i, res)
        ?abc -> variable(:abc)
verb(self, str, i, res)
 has _prop_
is _prop_ of
a
=
_prop_
>- prop ->
<- prop -<
_operator_
Data and non-method functions defined here:
__doc__ = None
__module__ = 'notation3'
 
class ToN3(RDFSink.RDFSink)
      Serializer output sink for N3
 
keeps track of most recent subject and predicate reuses them.
Adapted from Dan's ToRDFParser(Parser);
 
   Methods defined here:
__init__(self, write, base=None, genPrefix=None, noLists=0, quiet=0, flags='')
_endStatement(self)
_makeSubjPred(self, context, subj, pred)
_newline(self, extra=0)
bind(self, prefixString, uri)
 Just accepting a convention here
dummyClone(self)
retun a version of myself which will only count occurrences
endAnonymous(self, subject, verb)
endAnonymousNode(self, subj=None)
endBagNamed(self, subj)
endBagObject(self, pred, subj)
endBagSubject(self, subj)
endDoc(self, rootFormulaPair=None)
makeComment(self, str)
makeStatement(self, triple, why=None, aIsPossible=1)
representationOf(self, context, pair)
  Representation of a thing in the output stream
 
Regenerates genids if required.
Uses prefix dictionary to use qname syntax if possible.
setDefaultNamespace(self, uri)
startAnonymous(self, triple, isList=0)
startAnonymousNode(self, subj, isList=0)
startBagNamed(self, context, subj)
startBagObject(self, triple)
startBagSubject(self, context)
startDoc(self)
writeEncoded(self, str)
Write a possibly unicode string out to the output
Data and non-method functions defined here:
__doc__ = "Serializer output sink for N3\n \n keeps t... Adapted from Dan's ToRDFParser(Parser);\n "
__module__ = 'notation3'
flagDocumentation = 'Flags for N3 output are as follows:-\n \na ...f @forAll, and "this log:forAll" for "@forSome".\n'
Methods inherited from RDFSink.RDFSink:
checkNewId(self, uri)
The store can override this to raise an exception if the
id is not in fact new. This is useful because it is usfeul
to generate IDs with useful diagnostic ways but this lays them
open to possibly clashing in pathalogical cases.
countNamespace(self, namesp)
On output, count how many times each namespace is used
genId(self)
namespaceCounts(self)
newBlankNode(self, context, uri=None, why=None)
newExistential(self, context, uri=None, why=None)
newFormula(self, uri=None)
newLiteral(self, str, dt=None, lang=None)
newSymbol(self, uri)
newUniversal(self, context, uri=None, why=None)
reopen(self)
Un-End a document
 
If you have added stuff to a document, thought you were done, and
then want to add more, call this to get back into the sate that makeSatement
is again acceptable. Remember to end the document again when done.
setGenPrefix(self, genPrefix)
 
Functions
            
backslashUify(ustr)
Use URL encoding to return an ASCII string corresponding to the given unicode
dummy()
dummyWrite(x)
hexify(ustr)
Use URL encoding to return an ASCII string corresponding to the given unicode
stringToN3(str, singleLine=0, flags='')
#"
stripCR(str)
 
Data
             ADDED_HASH = '#'
ALL4 = (0, 1, 2, 3)
ANONYMOUS = 3
CONTEXT = 0
DAML_sameAs = (0, 'http://www.w3.org/2002/07/owl#sameAs')
DAML_sameAs_URI = 'http://www.w3.org/2002/07/owl#sameAs'
DECIMAL_DATATYPE = 'http://www.w3.org/2001/XMLSchema#decimal'
Escapes = {'"': '"', r'\': r'\', 'a': '\x07', 'b': '\x08', 'f': '\x0c', 'n': '\n', 'r': '\r', 't': '\t', 'v': '\x0b'}
FLOAT_DATATYPE = 'http://www.w3.org/2001/XMLSchema#double'
INTEGER_DATATYPE = 'http://www.w3.org/2001/XMLSchema#integer'
LITERAL = 2
LOG_implies_URI = 'http://www.w3.org/2000/10/swap/log#implies'
List_NS = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'
Logic_NS = 'http://www.w3.org/2000/10/swap/log#'
N3CommentCharacter = '#'
N3_Empty = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#Empty')
N3_List = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#List')
N3_first = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#first')
N3_forAll_URI = 'http://www.w3.org/2000/10/swap/log#forAll'
N3_forSome_URI = 'http://www.w3.org/2000/10/swap/log#forSome'
N3_nil = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#nil')
N3_rest = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#rest')
OBJ = 3
PARTS = (1, 2, 3)
PRED = 1
RDF_NS_URI = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'
RDF_spec = 'http://www.w3.org/TR/REC-rdf-syntax/'
RDF_type = (0, 'http://www.w3.org/1999/02/22-rdf-syntax-ns#type')
RDF_type_URI = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'
SUBJ = 2
SYMBOL = 0
__file__ = './notation3.pyc'
__name__ = 'notation3'
_notNameChars = '\t\r\n !"#$%&\'()*.,+/;<=>?@[\\]^`{|}~:'
_notQNameChars = '\t\r\n !"#$%&\'()*.,+/;<=>?@[\\]^`{|}~'
_rdfns = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'
digitstring = <_sre.SRE_Pattern object>
eof = <_sre.SRE_Pattern object>
eol = <_sre.SRE_Pattern object>
forbidden1 = <_sre.SRE_Pattern object>
forbidden2 = <_sre.SRE_Pattern object>
interesting = <_sre.SRE_Pattern object>
langcode = <_sre.SRE_Pattern object>
number_syntax = <_sre.SRE_Pattern object>
option_noregen = 0
parsesTo_URI = 'http://www.w3.org/2000/10/swap/log#parsesTo'
signed_integer = <_sre.SRE_Pattern object>
ws = <_sre.SRE_Pattern object>