def importann(pathtofile):
"""
Import ann and .txt files from a folder.
:param pathtofile: (string) the path to the folder containing both the
.ann and .txt files.
:return: a tuple containing a dictionary of annotations and a string,
representing the text of the document
"""
annotations = readannfile(pathtofile)
path, extension = os.path.splitext(pathtofile)
sentences = []
text = []
char_index = 0
for sent_index, line in enumerate(open(path + ".txt", encoding='utf-8')):
sentences.append(Sentence(sent_index, line, char_index))
char_index += len(line)
_join(annotations.values(), sentences)
with open(path + ".txt", encoding='utf-8') as input_file:
text = input_file.read()
return sentences, text
def importann(pathtofile):
"""
Import ann and .txt files from a folder.
:param pathtofile: (string) the path to the folder containing both the
.ann and .txt files.
:return: a tuple containing a dictionary of annotations and a string,
representing the text of the document
"""
annotations = readannfile(pathtofile)
path, extension = os.path.splitext(pathtofile)
sentences = []
char_index = 0
##this is a horrible hack because i want to preserve newlines
with open(path + ".txt", encoding='utf8') as f:
for sent_index, line in enumerate(f.read().split('\n\n')):
sentences.append(Sentence(sent_index, line, char_index))
char_index += len(line) + len('\n\n')
# for sent_index, line in enumerate(open(path + ".txt", encoding='utf-8')):
# sentences.append(Sentence(sent_index, line, char_index))
# char_index += len(line)
_join(annotations.values(), sentences)
return sentences
def importann(pathtofile):
"""
Imports ann and .txt files from a folder.
:param pathtofile: (string) the path to the folder containing both the .ann and .txt files.
:return: a tuple containing a dictionary of annotations and a string, representing the text of the document
"""
annotations = readannfile(pathtofile)
context = _readcontext(".".join(pathtofile.split(".")[:-1] + ["txt"]))
sentences = []
index = 0
for sindex, line in enumerate(context.splitlines()):
sentences.append(Sentence(sindex, line, index))
index += len(line)+1
_couple(annotations.values(), sentences)
return annotations, sentences
def importann(pathtofile):
"""
Import ann and .txt files from a folder.
:param pathtofile: (string) the path to the folder containing both the
.ann and .txt files.
:return: a tuple containing a dictionary of annotations and a string,
representing the text of the document
"""
annotations = readannfile(pathtofile)
path, extension = os.path.splitext(pathtofile)
sentences = []
char_index = 0
# for sent_index, line in enumerate(open((path + ".txt"), errors='ignore')):
for sent_index, line in enumerate(
open((path + ".txt"), errors='ignore', encoding="utf-8")):
sentences.append(Sentence(sent_index, line, char_index))
char_index += len(line) + 1
_join(annotations.values(), sentences)
return sentences
def importxml(filename):
"""
Imports an XML file formatted with the format created by this program. Used for persistency and to operate on
RepoModels in memory.
:param filename: (string) the path to the file to be imported.
:return: A tuple containing a dictionary of annotations and a list of dictionaries representing the context.
"""
anndict = OrderedDict()
sentobjects = []
with codecs.open(filename, 'r', encoding='utf-8') as f:
data = f.read()
doc = etree.fromstring(data)
sentences, annotations = doc.getchildren()
for s in sentences.getchildren():
repr = " ".join([w.text for w in s.getchildren()])
sentobjects.append(
Sentence(key=s.get('id').split(".")[1],
line=repr,
start=int(s.get("start"))))
for annotation in annotations.getchildren():
id = unicode(annotation.get('id')[3:])
repr = unicode(annotation.get('repr'))
spans = [[int(y) for y in x.split("|")]
for x in annotation.get('spans').split(",")]
ann = Annotation(id, repr, spans)
for span in ann.spans:
for s in sentobjects:
start, end = span
ann.words.extend(s.getwordsinspan(start, end))
anndict[id] = ann
for annotation in annotations.getchildren():
id = unicode(annotation.get('id')[3:])
ann = anndict[id]
for x, y in {
unicode(x): unicode(y)
for x, y in annotation.attrib.items()
if x not in ["id", "repr", "spans", "words"]
}.items():
if x.startswith("link."):
ann.links[x[5:]].extend(
[anndict[key[3:]] for key in y.split()])
else:
ann.labels[x].append(y)
return anndict, sentobjects