add script

lasid.py

@@ -0,0 +1,326 @@
+# coding=utf-8
+# Copyright 2021 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
+# Copyright 2021 Phonetics and Speech Laboratory, Trinity College, Dublin
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# Lint as: python3
+
+import sys
+from pathlib import Path
+
+import datasets
+
+try:
+    import icu
+except ImportError:
+    sys.exit("ICU not found (hint: pip install pyicu)")
+
+_DESCRIPTION = """\
+Linguistic Atlas and Survey of Irish Dialects, volume 1
+"""
+
+_CITATION = """\
+@book{wagner1958linguistic,
+  title={Linguistic Atlas and Survey of Irish Dialects: Introduction, 300 maps.},
+  author={Wagner, H.},
+  number={v. 1},
+  year={1958},
+  publisher={Dublin Institute for Advanced Studies}
+}
+
+@phdthesis{mckendry1982computer,
+  title={Computer-aided contributions to the study of Irish dialects},
+  author={McKendry, Eugene},
+  year={1982},
+  school={Queen's University Belfast}
+}
+
+@article{mckendry1998linguistic,
+  title={The Linguistic Atlas and Survey of Irish Dialects (LASID) and the Computer},
+  author={McKendry, Eugene},
+  journal={Studia Celtica Upsaliensia},
+  volume={2},
+  pages={345--354},
+  year={1998}
+}
+"""
+
+_DATA_URL = "https://www3.smo.uhi.ac.uk/oduibhin/oideasra/lasid/lasid.zip"
+
+LASID_ICU = """
+\x07 → ᵏ ;
+\\\t → ᵉ ; # \x09
+\x0e → ᴵ ;
+\x11 → ʰ ;
+\x12 → ⁱ ;
+\x13 → ᵒ ;
+\x14 → ᵒ̤ ;
+\x15 → ʳ ;
+\x16 → ˢ ;
+\x17 → ᶴ ;
+\x18 → ᵗ ;
+\x19 → ᵘ ;
+\x1a → ᵘ̯ ;
+\x1c → ᵛ ;
+\x1d → ʷ ;
+\x1e → ᶾ ;
+\x1f → ᵊ ;
+\# → ᶠ ; # \x23
+\$ → ᵠ ; # \x24
+\% → ᵍ ; # \x25
+\& → ᵞ ; # \x26 ˠ for IPA
+\' → ’ ; # \x27
+\: → ː ; # \x3a
+\< → ⁱ̈ ; # \x3c
+\= → ⁱ̯ ; # \x3d
+\? → ʔ ; # \x3f
+\@ → ʲ ; # \x40
+E → ᴇ ; # \x45
+I → ɪ ; # \x49
+L → ʟ ;
+N → ɴ ;
+R → ʀ ;
+\^ → ᵐ ; # \x5e
+\_ → ǰ ; # crane, 021 # \x5f
+\` → ɛ̀̃ ; # limekiln, 078: \x60
+\| → ⁿ ; # lamb, 055: \x7c
+\~ → ᵑ ; # dreaming, 078; maybe ⁿ̠ ? # \x7e
+\x7f → ᴇ̃ ;
+\x80 → φ ; # ɸ
+\x81 → ü ;
+\x83 → ɛ \u0300 ;
+\x84 → è \u0323 ; # FIXME
+\\\x85 → è̃ ; # this is �, so it needs to be escaped
+\x86 → ũ̜ ; # lamb, 038
+\x87 → u̜ ; # finger-nails, 043
+\x88 → ʈ ; # looks like t̜ : toothache, 033
+\x89 → ᵃ ; # eggs, 066
+\x8a → è ;
+\x8b → ï ;
+\x8c → ɔ̜̃ ; # grandmother, 007
+\x8d → ɔ̜ ;
+\x8e → ɔ̆ ; # before i go, 078
+\x8f → õ̜ ; # as cute, 062
+\x91 → æ ;
+\x92 → o̜ ;
+\x93 → ɖ ;
+\x94 → ö ;
+\x95 → ɑ̜̃ ;
+\x96 → û ; # milking, 067
+\x97 → ɑ \u0323 ; # FIXME (maybe α̩  or ɑ̜ ?)
+\x98 → v̠ ;
+\x99 → t̠ ; # toothache, 021
+\x9a → r̠ ;
+\x9b → ø ;
+\x9c → ɴ̠ ; # sick, 034
+\x9d → ŋ̠ ; # grazing, 002
+\x9e → n̠ ;
+\x9f → l̠ ; # plumage, 068
+\xa4 → k̠ ; # plumage, 068
+\xa5 → g̠ ;
+\xa6 → d̠ ; # wedge, 021
+\xa7 → ŭ ;
+\xa8 → ö̆ ;
+\xa9 → ŏ ;
+\xaa → ĭ ;
+\xab → ɛ̆ ;
+\xac → ĕ ;
+\xad → o̤ ;
+\xae → λ ;
+\xaf → ɑ ; # α in the software
+\xb0 → ɔ ;
+\xb1 → ɑ̆ \u0323 ; # FIXME
+\xb2 → ə ;
+\xb4 → ᵈ ; # tail, 007
+\xb6 → ɑ̆ ; # ᾰ in the software
+\xb7 → ă ;
+\xb8 → λ \u0323 ; # FIXME
+\xb9 → ɛ ;
+\xba → ʃ \u030c ; # calling, 067
+\xbb → š ;
+\xbc → ř ;
+\xbd → ɑ̃ ;
+\xbe → ẽ ; # tied, 88N
+\xc1 → ′ ; # superscript prime
+\xc5 → ᴍ̠ ; # fart, 071
+\xc6 → ã ; # calf, 046
+\xc7 → t \u0323 ; # probably t̞
+\xc8 → λ̯ ; # mane, 067
+\xc9 → o̯ ; # hare, 088
+\xca → Ɫ ; # loaf, 001
+\xcb → ɫ ; # loaf, 003
+\xcc → m̥ ; # awake, 001
+\xcd → ʀ̥ ; # thieving, 003
+\xce → ˈ ;
+\xcf → ˌ ; # cattle, 040
+\xd0 → ð ; # boar, 88N
+\xd1 → s \u0323 ; # FIXME # slime 008
+\xd2 → r \u0323 ; # FIXME # bulls 067
+\xd3 → ɪ̆ ; # suit of clothes 039
+\xd4 → ᴇ̀ ;
+\xd5 → p \u0323 ; # FIXME # castrating 053
+\xd7 → ɪ̃ ; # slime, 007
+\xd8 → ɪ̈ ; # calf 027
+\xdb → o \u0323 ; # FIXME # cow 028
+\xdc → ŋ \u0323 ; # FIXME # tied 078
+\xdd → ö̤ ;
+\xde → k \u0323 ; # FIXME
+\xdf → i \u0323 ; # FIXME # sick 069
+\xe1 → g \u0323 ; # FIXME
+\xe2 → e \u0323 ; # FIXME
+\xe3 → d \u0323 ; # FIXME # agut 052
+\xe4 → õ ; # I shall tie 062
+\xe5 → b \u0323 ; # FIXME # castrating 071
+\xe6 → ɑ̃ \u0323 ; #FIXME # barking 049
+\xe7 → ɑ \u0323 ; # FIXME # slime 008
+\xe8 → ỹ ;
+\xea → λ̃ ;
+\xeb → ü̃ ; # churn-dash, 011
+\xec → ũ ;
+\xed → ɔ̃ ; # cow 074
+\xee → õ̤ ; # barking 055
+\xef → ′ ;
+\xf0 → ″ ;
+\xf1 → ö̤̃ ; # dreaming, 078
+\xf2 → ö̃ ; # sheep shears 074
+\xf3 → ï̃ ; # churn-dash, 034
+\xf4 → ĩ ; # sick 001
+\xf5 → ɣ̃ ; # tied 075
+\xf6 → ɛ̃ ; # tied 067
+\xf7 → n̥ ; # awake, 059
+\xf8 → r̥ ; # slime 002
+\xf9 → ʃ ;
+\xfb → · ; # slime 058
+\xfa → ɣ ;
+\xfc → χ ; # limekiln, 080
+\xfd → ʒ ; # sheep shears 054
+\xfe → ŋ ;
+"""
+
+LASID_TITLES_ICU = """
+\xb5 → Á ;
+\xd6 → Í ;
+\x90 → É ;
+\xe0 → Ó ;
+\xe9 → Ú ;
+"""
+
+def transliterator_from_rules(name, rules):
+    fromrules = icu.Transliterator.createFromRules(name, rules)
+    icu.Transliterator.registerInstance(fromrules)
+    return icu.Transliterator.createInstance(name)
+
+LASID = transliterator_from_rules('lasid_icu', LASID_ICU)
+TITLES = transliterator_from_rules('lasid_titles', LASID_TITLES_ICU)
+
+def translit_phon(text):
+  # could have been any 8-bit encoding
+  return LASID.transliterate(text.decode('ISO-8859-1').rstrip())
+
+def translit_irish(text):
+  return TITLES.transliterate(text.decode('ISO-8859-1').rstrip())
+
+
+class LasidDataset(datasets.GeneratorBasedBuilder):
+    """Scraper dataset for LASID."""
+
+    VERSION = datasets.Version("1.1.0")
+
+    BUILDER_CONFIGS = [
+        datasets.BuilderConfig(name="lasid"),
+    ]
+
+    def _info(self):
+        features = datasets.Features(
+            {
+                "english": datasets.Value("string"),
+                "irish": datasets.Value("string"),
+                "map_id": datasets.Value("string"),
+                "place_ids": datasets.Sequence(datasets.Value("string")),
+                "transcripts": datasets.Sequence(datasets.Value("string")),
+            }
+        )
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            supervised_keys=None,
+            citation=_CITATION
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        dl_path = dl_manager.download_and_extract(_DATA_URL)
+        infile = f"{dl_path}/mapdata.dat"
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "split": "train",
+                    "data_file": infile
+                },
+            ),
+        ]
+
+    def _generate_examples(
+        self, split, data_file
+    ):
+        """ Yields examples as (key, example) tuples. """
+        data = process_lasid(data_file)
+        _id = 1
+        for map in data.keys():
+            item = data[map]
+            place_ids = list(item["data"])
+            transcripts = [item["data"][a] for a in place_ids]
+
+            yield _id, {
+                "english": item.get("en", ""),
+                "irish": item.get("ga", ""),
+                "map_id": item.get("id", ""),
+                "place_ids": place_ids,
+                "transcripts": transcripts
+            }
+            _id += 1
+
+
+def process_lasid(filename):
+    data = {}
+    cur = {}
+    en = ''
+    ga = ''
+    id = ''
+    with open(filename, "rb") as file:
+        for line in file.readlines():
+            if b'{M' in line:
+                if en and id:
+                    tmp = {}
+                    tmp['en'] = en
+                    tmp['id'] = id
+                    tmp['ga'] = ga
+                    tmp['data'] = cur
+                    data[id] = tmp
+                text = line.decode('ISO-8859-1').rstrip()
+                id = text[3:7].strip()
+                en = text[7:-1].strip()
+                cur = {}
+            elif b'{F' in line:
+                raw = translit_irish(line)
+                ga = raw[3:-1].strip()
+            elif line.decode('ISO-8859-1')[0:1].isnumeric():
+                pid = line.decode('ISO-8859-1')[0:3]
+                ptext = translit_phon(line[3:-1])
+                if ptext[-1] == '*':
+                    ptext = ptext[0:-1]
+                cur[pid] = ptext.strip()
+    return data