|
|
|
""" |
|
The corpus is provided in a vertical format, where sentence boundaries are marked with |
|
a blank line. Every word form is written on a separate line, followed by five tab-separated |
|
attributes: lemma, tag, ID (word index in the sentence), head and deprel (analytical function, |
|
afun in the PDT formalism). The texts are shuffled in random chunks of maximum 100 words |
|
(respecting sentence boundaries). Each chunk is provided as a separate file, with the suggested |
|
division into train, dev and test sets written as file prefix. |
|
|
|
Korpus FicTree se skládá z osmi prozaických děl z žánru beletrie vydaných v České republice mezi |
|
lety 1991 a 2007. Šest z těchto literárních děl se (dle klasifikace textů podle žánrů používané |
|
v ČNK do r. 2015) považuje za „čistou“ beletrii, jedno dílo se řadí k memoárům, jedno dílo spadá |
|
do žánru „literatura pro děti a mládež“. Pět textů (80% tokenů) jsou původní české texty, dva texty |
|
jsou překlady z němčiny, jeden je překlad ze slovenštiny. |
|
""" |
|
|
|
import os |
|
import re |
|
|
|
from jsonlines import jsonlines |
|
|
|
|
|
documents = { |
|
"train": [], |
|
"dev": [], |
|
"test": [] |
|
} |
|
|
|
FOLDER = ".data/FICTREE_1.0/FicTree/" |
|
ws_before_punct = re.compile(r'\s+([.,!?])') |
|
|
|
for file in os.listdir(FOLDER): |
|
if file.endswith(".vert"): |
|
with open(os.path.join(FOLDER, file), "r") as f: |
|
if file.startswith("train"): |
|
documents["train"].append(f.read()) |
|
elif file.startswith("dev"): |
|
documents["dev"].append(f.read()) |
|
elif file.startswith("test"): |
|
documents["test"].append(f.read()) |
|
|
|
for split, docs in list(documents.items()): |
|
new_docs = [] |
|
for doc in docs: |
|
new_doc = [] |
|
for line in doc.splitlines(): |
|
token = line.split("\t")[0].strip() |
|
if token != "": |
|
new_doc.append(token) |
|
text = " ".join(new_doc) |
|
new_doc = re.sub(ws_before_punct, r'\1', text) |
|
new_docs.append(new_doc) |
|
documents[split] = new_docs |
|
|
|
|
|
OF = ".data/hf_dataset/cnc_fictree/test.jsonl" |
|
os.makedirs(os.path.dirname(OF), exist_ok=True) |
|
with jsonlines.open(OF, "w") as writer: |
|
for split, docs in list(documents.items()): |
|
for doc in docs: |
|
writer.write({"text": doc, "orig_split": split}) |