Datasets:
Commit
•
828708b
1
Parent(s):
652d6f4
Fix data URL (#3)
Browse files- Update data URL (7e706635743f5d0b09d6350db9154183eae23d18)
- Fix DuplicatedKeysError (5b2b5ee5532c8ccc4d4dd368984d39fe72152d7d)
- hkcancor.py +6 -3
hkcancor.py
CHANGED
@@ -61,7 +61,9 @@ _HOMEPAGE = "http://compling.hss.ntu.edu.sg/hkcancor/"
|
|
61 |
|
62 |
_LICENSE = "CC BY 4.0"
|
63 |
|
64 |
-
|
|
|
|
|
65 |
|
66 |
|
67 |
class Hkcancor(datasets.GeneratorBasedBuilder):
|
@@ -243,7 +245,7 @@ class Hkcancor(datasets.GeneratorBasedBuilder):
|
|
243 |
|
244 |
def _generate_examples(self, data_dir, split):
|
245 |
"""Yields examples."""
|
246 |
-
|
247 |
downloaded_files = [os.path.join(data_dir, fn) for fn in sorted(os.listdir(data_dir))]
|
248 |
for filepath in downloaded_files:
|
249 |
# Each file in the corpus contains one conversation
|
@@ -299,7 +301,7 @@ class Hkcancor(datasets.GeneratorBasedBuilder):
|
|
299 |
# to create an identifier.
|
300 |
id_from_transcriptions = "".join(transcriptions[:5])[:5].upper()
|
301 |
id_ = f"{tape_number}-{date_recorded}-{id_from_transcriptions}"
|
302 |
-
yield
|
303 |
"conversation_id": id_,
|
304 |
"speaker": current_speaker,
|
305 |
"turn_number": turn_number,
|
@@ -308,3 +310,4 @@ class Hkcancor(datasets.GeneratorBasedBuilder):
|
|
308 |
"pos_tags_prf": pos_prf,
|
309 |
"pos_tags_ud": pos_ud,
|
310 |
}
|
|
|
|
61 |
|
62 |
_LICENSE = "CC BY 4.0"
|
63 |
|
64 |
+
# Original URL gives connection error
|
65 |
+
# _URL = "http://compling.hss.ntu.edu.sg/hkcancor/data/hkcancor-utf8.zip"
|
66 |
+
_URL = "https://github.com/fcbond/hkcancor/raw/master/data/hkcancor-utf8.zip"
|
67 |
|
68 |
|
69 |
class Hkcancor(datasets.GeneratorBasedBuilder):
|
|
|
245 |
|
246 |
def _generate_examples(self, data_dir, split):
|
247 |
"""Yields examples."""
|
248 |
+
key = 0
|
249 |
downloaded_files = [os.path.join(data_dir, fn) for fn in sorted(os.listdir(data_dir))]
|
250 |
for filepath in downloaded_files:
|
251 |
# Each file in the corpus contains one conversation
|
|
|
301 |
# to create an identifier.
|
302 |
id_from_transcriptions = "".join(transcriptions[:5])[:5].upper()
|
303 |
id_ = f"{tape_number}-{date_recorded}-{id_from_transcriptions}"
|
304 |
+
yield key, {
|
305 |
"conversation_id": id_,
|
306 |
"speaker": current_speaker,
|
307 |
"turn_number": turn_number,
|
|
|
310 |
"pos_tags_prf": pos_prf,
|
311 |
"pos_tags_ud": pos_ud,
|
312 |
}
|
313 |
+
key += 1
|