#!/usr/bin/python
#
# 
# Open source. W3C licence.
#
"""
See http://www.w3.org/2002/12/cal/vcard-notes

See Namespace <http://www.w3.org/2006/vcard/ns>
and python iCalendar (vcard?) implementtions

<http://www.schooltool.org/products/schooltool-calendar>

<http://www.w3.org/2002/12/cal/icslex.py>

<http://www.ietf.org/rfc/rfc2425.txt>
    dc:title <A MIME Content-Type for Directory Information>;
    s:comment "Defines text/directory . the line-folding
     and basic record type structure".
<http://www.ietf.org/rfc/rfc2426.txt>
    dc:title "vCard MIME Directory Profile".

<http://www.w3.org/TR/vcard-rdf> dc:title "Representing vCard Objects in RDF/XML".

<http://lesscode.org/projects/kid/> an project used by 
<http://www.w3.org/2002/12/cal/vcardin.py> written by danc
"""

import sys, string, re, os

# export PYTHONPATH=$PYTHONPATH:/devel/WWW/2002/12/cal
import icslex # from http://www.w3.org/2002/12/cal/icslex.py

import notation3 # http://www.w3.org/2000/10/swap/notation3.py

from notation3 import stringToN3

lineLength = 76 # A few clear of 80, as used by AB base64 wrapping

singleTextField = [ 'fn',  'title', 'bday', 'description', 'note',
            'x-abuid','x-abadr', 'x-aim', 'x-abrelatednames', 'x-abshowas' ]

kludgeTags = [ 'v:x-ablabel', 'v:x-abadr']  # These have sideeffects

# From http://www.ietf.org/rfc/rfc2426.txt  adr-type
typeFields = {
    'adr':   [ "dom" , "intl" , "postal" , "parcel" , "home"
                , "work" , "pref" ], # , iana-type / x-name

    'bday': [ 'date' ],
    
    'tel': ["home" , "work", "pref", "voice", "fax", "msg",
                "cell", "pager",  "bbs", "modem", "car", "isdn",
                "video", "pcs" 
                , 'main'      # @@ Added for AB
                ],
    'email': [ "internet", "x400"
                    ,'home', 'work'    # @@ Added as they turn up in AB
                ],
    'x-aim': ['home', 'work'    # @@ Added as they turn up in AB
                ],
    'url':  [ 'home', 'work', 'foaf'] #unofficial
            }

relationshipModifiers  = { 'home':1, 'work':1, 'main':1 } # These make work-adr etc
# Others are a class of phone/email/etc
 
fieldProperties = {   # @@@ These are rather long localnames IMHO
    'n': [ 'family-name', 'given-name', 'additional-name', 'prefix', 'suffix' ],

    'adr': [ 'post-office-box', 'extended-address', 'street-address',
            'locality', 'region', 'postal-code', 'country-name' ],
    'org': [ 'organization-name', 'organization-unit']
            }
           
             
def zapOut(str, allowed):
    """Only allow the characters given. Strings of consecutive
    unallowed characters are replaced with a single underscore character"""
    str2 = ""
    for i in range(len(str)):
        if str[i] in allowed:
            str2 = str2 + str[i]
        else:
            if str2[-1:] != "_": str2 = str2 + "_"
    return str2

def munge(str):
    return zapOut(str, "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789")

def splitBy(stri, delim, unescape=1):
    "Split by unescaped delimiters"
    result = []
    begin = 0
    escaped = 0
    escape = '\\'
    while begin < len(stri):
        i, unesc = begin, ""
        while i < len(stri):
            ch = stri[i]
            if escaped:
                escaped = 0
                if ch in "nN": unesc += '\n'
                else: unesc += ch
            else:
                if ch == delim: break
                if ch == escape: escaped = 1
                else: unesc += ch
            i = i + 1
        if unescape: result.append(unesc)
        else: result.append(stri[begin:i])
        begin = i+1
    return result

def wr(s):
    sys.stdout.write(s.encode('utf-8'))

def extract(path):
    global nochange
    global verbose
    total = 0
    
    wr( """# n3  http://www.w3.org/DesignIssues/Notation3.
# From vCard data in %s
# Extracted by $Id: vcard2n3.py,v 1.8 2014/09/09 13:12:23 timbl Exp $ 
@prefix : <#>.
@prefix loc: <#loc_>.
@prefix s: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix log: <http://www.w3.org/2000/10/swap/log#>.
@prefix v:  <http://www.w3.org/2006/vcard/ns#>.
@prefix vc:  <http://www.w3.org/2006/vcard/class#>.
@prefix abl:  <http://www.w3.org/2006/vcard/abl#>.
@prefix user: <#>.
""" % path)

    input = open(path, "r")
    b = input.read()
    input.close()
    
    wr("# Length: " + `len(b)`+ "starts ")
    wr(" ".join(["%2x"%ord(ch) for ch in b[:8]]))
    wr("\n")
    if ord(b[0])==0 and ord(b[1]) == ord('B'):  #  UTF16 with MSB byte order unmarked    
        d = "\xfe\xff"   # Add byte order mark
        buf = (d+b).decode('utf-16')
        wr( " #Warning:  UTF-16 was not byte order marked.\n")
    else:
        buf = b.decode('utf-8')

    group_line = re.compile(r'^([a-zA-Z0-9]+)\.(.*)')
    field_value = re.compile(r'^([A-Za-z0-9_-]*):(.*)')
    group = None    # 
    
    cardData = ""
    groupData = {}
    groupPred = {}
    

    def readBareLine(buf):
        "Return None for EOF or a line, even including a final widowed line"
        global bufpo # can't pass by ref
        begin = bufpo
        if begin == len(buf): return None
        bufpo = buf.find('\n', begin)
        if bufpo < 0:
            bufpo = len(buf)
            line = buf[begin:]
        else:
            line = buf[begin:bufpo]
            bufpo += 1  # After \n

        while line [-1:] == "\r": line = line[:-1] # Strip triling CRs
        return line
        
    def startGroup(g):
#       print "# Start group  <%s>" % `g`
        groupData[g] = []   # Pairs of p and o
        groupPred[g] = "loc:%s" % munge(g)  # Unless overwritten 


    def endGroup(g):
        #print "# End group <%s> data:%s" % (`g`, groupData[g])
        pos = groupData[g]
        kludges, data = [], []
        for i in range(len(pos)):
            p, o = pos[i]
            if p in kludgeTags: kludges.append((p,o))
            else: data.append((p,o))
        if len(data) == 1:  # The AddressBook model, one data item + kludges
            dp, do = data[0]
            for p,o in kludges:
                if p == 'v:x-ablabel':
                    dp = o  # Override predicate
                if p == 'v:x-abadr':
                    assert do[-1:] == ']';
                    do = do[:-1]+' '+ p +' '+ o +';]' # annnotate object
            return "%s %s;\n" % (dp, do)

        if len(kludges) != 0:
            raise ValueError("Unknown Group pattern:"+`pos`)

        res = ""
        for p,o in  pos:
            res += " %s %s;" %(p,o)
        return "%s [ # %s\n%s];\n" % (groupPred[g], g, res)

    def lineFold(str1, str2):
        x = str1.rfind('\n')
        if x < 0: x = 0
        if len(str1) - x + len(str2) > lineLength:
            return str1+ "\n\t" + str2
        return str1 + str2

    def orderedFields(value, map):
            cardData = ""
            beg = 0
            for i in range(len(map)):
                end = beg
                while 1:
                    end = value.find(";", end)
                    if end>0 and value[end-1] == "\\":
                        end += 1
                        continue
                    break
                if end < 0:
                    end = len(value)
                st = " ".join(splitBy(value[beg:end], ','))
                if st: cardData = lineFold(cardData, ' v:%s %s;' % \
                                    (map[i], stringToN3(st, singleLine=1)))
                beg=end+1
                if beg > len(value):
                    break
            return cardData

    def predicateObject(n, props, value):
        "Return a pair of the predicate and object as N3 strings"
        modifiers = ""
        datatype = None
        classes = []
        for prop, val in props:
            if prop == 'type':
                vals = val.lower()
                for val in splitBy(vals, ','):
                    if val == 'internet' and n == 'email':
                        pass
                    elif val == 'pref':   # Preferred @@ - how represent?
                        pass
                    elif val in typeFields.get(n, []):
                        if relationshipModifiers.get(val, 0):
                            if modifiers: print "# @@ multiple modifiers in: "+line
                            modifiers = val + '-' + modifiers
                        else: classes.append('vc:'+val[0].upper()+val[1:])
                    else:
                        raise ValueError("Unhandled type %s in: %s" %(val, line))
            elif prop == 'value':  # This means datatype
                datatype = val
                if val == 'date':
                    pass # Date-times from AB certainly look like w3c not iCal dates
                elif val == 'uri':
                    pass
                else:
                    raise ValueError ('Unimplemented data type:'+val)
            elif prop == 'base64' or (prop == 'encoding' and val.lower() == 'b'):
                value = value.replace(' ','')
                res = ""
                while value:
                    res += value[:lineLength] + "\n"
                    value = value[lineLength:]
                return 'v:'+n, '[ v:base64 """%s"""]\n' % (res)  # Special case
                                
            else: raise ValueError('Unknown property %s with value %s' & (prop, val))

        classSpec = ""
        if classes: classSpec = 'a '+(', '.join(classes))

        map = fieldProperties.get(n,None)
        pred = 'v:%s%s' % (modifiers, n)
        if map:
            if classSpec: classSpec = '\n\t'+classSpec
            if n == 'n': # Special case 
                assert classSpec == ""
                return  '', orderedFields(value, map)  # Naked fields - see notes
            return pred, '[' + orderedFields(value, map) + classSpec + ']'
        if n == 'version':
            assert value == "3.0", "value found: "+`value`
            return "", ""
        if n == 'x.ablabel':
            return "", "" # used elsewhere

        if n == 'categories':   # Really should relate these to classes, but this roundtrips
            obj = ", ".join(['"'+x+'"' for x in splitBy(value, ',')])
            return pred,  obj 


        unesc = splitBy(value, ';')
        if len(unesc) != 1: raise ValueError("Unescaped semicolon in value: "+ value)
        unesc = unesc[0]

        obj = None
        if n == 'tel':
            if value[0] != '+':
                print "# @@ Warning: not international form tel: "+value
            obj = '<tel:%s>' % (value.replace(' ','-'))
        elif n == 'url':
            obj = '<%s>' % (value)
        elif n == 'email':
            obj = '<mailto:%s>' % (value)

        if obj:  # Any case so far is a form of URI
            if classSpec: wr('%s %s.\n'  %(obj, classSpec))
            return pred, obj

        elif n in singleTextField :  # Single text
            if classSpec:
                raiseValueError("Unexpected class on %s: %s"%(n,`classSpec`))
            return pred, stringToN3(unesc, singleLine=0) # @@@ N3 escaping

        raise ValueError('Unknown tag:'+n)
    
    global bufpo
    bufpo = 0
    nextLine = readBareLine(buf)
    while 1:

        line = nextLine
        while 1:
            nextLine = readBareLine(buf)
            if not nextLine or nextLine[0] != ' ': break
            line += nextLine[1:]

        if line is None : break # EOF
#       wr( "# line: " +line[:100])
        m = group_line.match(line)
        if m:
            g = m.group(1)
            line = m.group(2)
            if group != g:
                if group is not None: cardData += endGroup(group)
                if g is not None: startGroup(g)
            group = g
            
            n, props, value = icslex.parseLine(line)
#           for prop, val in props:
#               if prop == 'type':
#                   val = val.lower()
#                   if group .startswith("item"): # AB hack
#                       groupPred[group] = "loc:"+val
                        
            if n == 'x-ablabel':
                pred = 'v:'+n
                if value[:4] == "_$!<" and value[-4:] == ">!$_": # [sic]
                    obj = "abl:"+munge(value[4:-4]).lower()
                else:  # User generated
                    obj = "user:"+munge(value).lower()
            else:
                pred, obj = predicateObject(n, props, value)
            groupData[group].append((pred, obj))

        else:
            if group is not None:  # End group
                cardData += endGroup(group);
                group = None
            n, props, value = icslex.parseLine(line)    
            if n == 'begin':
                cardData = ""
                cardID = "[]"
            elif n == "uid":
                cardID = "<uid:%s>" % value
            elif n == 'end':
                wr("%s %s." % (cardID, cardData))
            else:
                if n == 'n': # ugh special case
                    map = fieldProperties.get(n,None)
                    cardData +=  orderedFields(value, map) +'\n' # Naked fields - see notes
                else:
                    p, o = predicateObject(n, props, value)
                    if p: cardData+= "    %s %s;\n" %(p, o)
            
    wr("\n\n#ends\n")            
    input.close()


def do(path):
    if verbose: sys.stderr.write("Doing " + path + "\n")
    extract(path) 

###################################

def _test():
    import doctest
    doctest.testmod()


######################################## Main program

recursive = 0
nochange = 1
verbose = 0
files = []

for arg in sys.argv[1:]:
    if arg[0:1] == "-":
        if arg == "-v": verbose = 1  
        if arg == '-t':
            _test()
            sys.exit(0)
        else:
            print """Bad option argument.
            -v  verbose
            -t  self-test

"""
            sys.exit(-1)
    else:
        files.append(arg)

if files == []: files = [ "." ] # Default to this directory

for path in files:
    do(path)