add sentiment data

.gitignore

@@ -0,0 +1,3 @@
+data/*.txt
+.DS_Store
+data/*.tsv

HAIndicSentiment.py

@@ -0,0 +1,79 @@
+"""TODO(IndicSentiment): Add a description here."""
+
+
+import json
+
+import datasets
+
+
+_HOMEPAGE = ""
+
+_CITATION = """\
+
+"""
+
+_DESCRIPTION = """\
+
+"""
+
+_LANG = ["as", "bn", "bd", "gu", "hi", "kn", "ml", "mr", "or", "pa", "ta", "te", "ur"]
+_URL = "https://huggingface.co/datasets/ai4bharat/HAIndicSentiment/resolve/main/data/{split}/{language}.json"
+_VERSION = datasets.Version("1.0.0", "First version of IndicSentiment")
+
+
+class HAIndicSentiment(datasets.GeneratorBasedBuilder):
+    """TODO(IndicSentiment): Short description of my dataset."""
+    BUILDER_CONFIGS = [
+        datasets.BuilderConfig(
+            name=f"translation-{lang}",
+            description=f"translated sentiment data for {lang}",
+            version=_VERSION,
+        )
+        for lang in _LANG
+    ]
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION + self.config.description,
+            features=datasets.Features(
+                {
+                    "GENERIC CATEGORIES": datasets.Value("string"),
+                    "CATEGORY": datasets.Value("string"),
+                    "SUB-CATEGORY": datasets.Value("string"),
+                    "PRODUCT": datasets.Value("string"),
+                    "BRAND": datasets.Value("string"),
+                    "ASPECTS": datasets.Value("string"),
+                    "ASPECT COMBO": datasets.Value("string"),
+                    "ENGLISH REVIEW": datasets.Value("string"),
+                    "LABEL": datasets.Value("string"),
+                    "INDIC REVIEW": datasets.Value("string"),
+                }
+            ),
+            homepage=_HOMEPAGE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        *translation_prefix, language = self.config.name.split("-")
+        splits = {datasets.Split.VALIDATION: "validation", datasets.Split.TEST: "test"}
+
+        data_urls = {
+            split: _URL.format(language=language, split=splits[split]) for split in splits
+        }
+        dl_paths = dl_manager.download(data_urls)
+        return [
+            datasets.SplitGenerator(
+                name=split,
+                gen_kwargs={"filepath": dl_paths[split]},
+            )
+            for split in splits
+        ]
+
+    def _generate_examples(self, filepath):
+        """Yields examples."""
+
+        with open(filepath, encoding="utf-8") as f:
+            for idx, row in enumerate(f):
+                data = json.loads(row)
+                yield idx, data

data/compile.py

@@ -0,0 +1,25 @@
+import pandas as pd
+import numpy as np
+langs = ['Hindi', 'Punjabi', 'Assamese', 'Bengali', 'Gujarati', 'Kannada', 'Marathi', 'Tamil', 'Malayalam', 'Odia', 'Telugu', 'Bodo', 'Urdu']
+
+codes = ['hi', 'pa', 'as', 'bn', 'gu', 'kn', 'mr', 'ta', 'ml', 'or', 'te', 'bd', 'ur']
+
+metadata = pd.read_csv('metadata.tsv', sep='\t')
+idx = np.random.permutation(metadata.index)
+
+# for lang, code in zip(langs, codes):
+for lang in codes:
+    with open(f"{lang}.txt", "r") as f:
+        lines = f.readlines()
+    lines = [line.strip() for line in lines]
+
+    metadata = pd.read_csv('metadata.tsv', sep='\t')
+    metadata.columns = ['GENERIC CATEGORIES', 'CATEGORY', 'SUB-CATEGORY', 'PRODUCT', 'BRAND', 'ASPECTS', 'ASPECT COMBO', 'ENGLISH REVIEW', 'LABEL']
+    metadata['INDIC REVIEW'] = lines
+
+    metadata = metadata.reindex(idx)
+
+    metadata[:1000].to_json(f"test/{lang}.json", orient='records', lines=True)
+    metadata[1000:].to_json(f"dev/{lang}.json", orient='records', lines=True)
+
+    # metadata.to_json(f"dataset/{code}.json", orient='records', lines=True)