def readZip(f, **kwargs):
from mien.parsers.nmpml import elements as dialect
f=ZipFile(f, 'r')
xml=f.read('xml')
xml=StringIO(xml)
doc=xm.readTree(xml)
xml.close()
doc=xm.assignClasses(doc, dialect)
try:
dat=f.read('data')
except:
print "No data archive in zip file"
return doc
from mien.parsers.datahash import readMD
dat=StringIO(dat)
dat=readMD(dat, return_raw_hash=True)
des=doc.getElements('Data')
for de in des:
try:
d, h=dat[de.upath()]
except:
print "can't find data for element %s" % (de.upath(),)
d, h=(zeros(0), {})
de.datinit(d, h)
f.close()
return doc
def read(fname, toObject=False):
'''returns a tree structure for the document (made of dicts)'''
if type(fname) in [str, unicode]:
fname = urllib.urlopen(fname)
sp = QuickParser()
sp.feed(fname.read())
sp.close()
# checkdata(sp.elements[0])
# print len(sp.elements)
if not toObject:
return sp.elements[0]
return assignClasses(sp.elements[0], {})
def deserialize(f, **kwargs):
#st=time.time()
if not type(f) in [str, unicode]:
f=f.read()
l=struct.unpack('<I', f[:4])[0]
doc=zlib.decompress(f[4:l+4])
doc=StringIO(doc)
doc=xm.readTree(doc)
from mien.parsers.nmpml import elements as dialect
doc=xm.assignClasses(doc, dialect)
f=f[l+4:]
try:
if f:
from mien.parsers.datahash import readMD
try:
f2=StringIO(f)
f=readMD(f2, return_raw_hash=True)
except:
f=zlib.decompress(f)
f2=StringIO(f)
f=readMD(f2, return_raw_hash=True)
del(f2)
des=doc.getElements('Data')
for de in des:
try:
d, h=f[de.upath()]
except:
print "can't find data for element %s" % (de.upath(),)
d, h=(zeros(0), {})
de.datinit(d, h)
except:
print "cant load data"
#print time.time()-st;st=time.time()
return doc
def read(fname, **kwargs):
'''reads from file fname. This function will attempt to automatically
identify the format of the file, and will ask the user for confirmation
if it can't make a unique ID. To override this behavior use the keyword
argument "format". This function will return None if it fails, or an
NmpmlObject if it succeeds.
fname may be the name of a local file, or it may be a url.
Key word arguments:
"format" - set this to a key of "filetypes" to force the format of the input
file to be treated as a particular format.
"gui" - Set this to a mien.wx.base.BaseGui instance to use that GUI's methods
for user interaction during the load. Set it to True to cause this
function to make its own GUI for interaction (otherwise, you will get
text-mode interaction)
"select" - ifTrue, ask the user to select subsets of a document to load
(by default, the whole document is automatically loaded)
"convertxml" - if True, cast generic xml to a simplified version of the
nmpml dialect. This will cause some advanced function of the interfaces
to work. On the other hand, this will may cause the xml to be modified.
Tags without an attribute "Name" will be given one, and the value of the
"Name" tag will be rendered sibling-unique. This may mean that if the
resulting xml is saved back to a file, other parsers will not read it
correctly.
By default, a compatibility class is used for generic XML that enables
most of the features of the Nmpml interfaces without modifying the xml.S
Note that this flag will only convert "generic" xml, so if there is a
user-defined xml dialect that is non-empty, but not nmpml-compliant,
this flag will not make it functional.
'''
url=fname
parts=parseurl(url)
fname=parts[2]
kwargs['unparsed_url']=url
kwargs['parsed_url']=parts
format=kwargs.get('format')
if not format:
format=get_file_format(fname, kwargs.get('gui'))
kwargs['format']=format
if format=="unknown xml":
ft={'xml dialect':'automatic', 'read':True}
elif not filetypes.has_key(format):
fl=match_extension(format)
if not fl:
print "Unknown format %s" % format
return
kwargs['format']=fl[0]
ft=filetypes[fl[0]]
else:
ft=filetypes[format]
if not ft["read"]:
print "format %s is write only" % format
return
fileobj, cleanup=openurlread(url)
if not ft.has_key('xml dialect'):
doc=ft["read"](fileobj, **kwargs)
cleanup()
else:
doc=xm.readTree(fileobj)
cleanup()
if ft['xml dialect']=='automatic':
format = get_xml_dialect(doc)
kwargs['format']=format
ft=filetypes[format]
dialect=ft['xml dialect']
if dialect=={}:
if kwargs.get("convertxml"):
dialect={'default class':filetypes['nmpml']['xml dialect']['default class']}
else:
from mien.nmpml.basic_tools import NmpmlCompat
dialect={'default class':NmpmlCompat}
doc=xm.assignClasses(doc, dialect)
if kwargs.get('select'):
if doc.fileinformation.get('select_done'):
del(doc.fileinformation['select_done'])
else:
doc=select_elements(doc, **kwargs)
doc.fileinformation["filename"]=url
doc.fileinformation["type"]=format
if not ft.get('autoload'):
doc.onLoad()
return doc
