keeping large csv files as tar and modified script to support them

snow-mountain.py

@@ -62,8 +62,8 @@ for lang in _LANGUAGES:
         "train_full": f"data/experiments/{lang}/train_full.csv",
         "val_full": f"data/experiments/{lang}/val_full.csv",
         "test_common": f"data/experiments/{lang}/test_common.csv",
-        "all_verses": f"data/cleaned/{lang}/all_verses.csv",
-        "short_verses": f"data/cleaned/{lang}/short_verses.csv",
+        "all_verses": f"data/cleaned/{lang}/all_verses.tar.gz",
+        "short_verses": f"data/cleaned/{lang}/short_verses.tar.gz",
     }
     _FILES[lang] = file_dic
 
@@ -75,7 +75,8 @@ OT_BOOKS = ['GEN', 'EXO', 'LEV', 'NUM', 'DEU', 'JOS', 'JDG', 'RUT', '1SA', '2SA'
 
 BOOKS_DIC = {'hindi':OT_BOOKS, 'bhadrawahi':NT_BOOKS, 'bilaspuri':NT_BOOKS, 'dogri':NT_BOOKS, 'gaddi':
             NT_BOOKS, 'haryanvi':NT_BOOKS, 'kangri':NT_BOOKS, 'kulvi':NT_BOOKS, 'kulvi_outer_seraji':NT_BOOKS
-            , 'mandeali':NT_BOOKS, 'pahari_mahasui':NT_BOOKS}
+            , 'mandeali':NT_BOOKS, 'pahari_mahasui':NT_BOOKS, 'malayalam':OT_BOOKS+NT_BOOKS, 'tamil': 
+            OT_BOOKS+NT_BOOKS, 'kannada': OT_BOOKS+NT_BOOKS, 'telugu': OT_BOOKS+NT_BOOKS}
 
 class Test(datasets.GeneratorBasedBuilder):
 
@@ -110,6 +111,7 @@ class Test(datasets.GeneratorBasedBuilder):
         
         downloaded_files = dl_manager.download(_FILES[self.config.name]) 
 
+        '''Downloads full audio here'''
         audio_data = {}
         for book in BOOKS_DIC[self.config.name]:
             archive_url = f"data/cleaned/{self.config.name}/{book}.tar.gz"
@@ -118,8 +120,7 @@ class Test(datasets.GeneratorBasedBuilder):
                 audio_ = path.split('/')[-1]
                 if audio_ not in audio_data:
                     content = file.read()
-                    audio_data[audio_] = content
-
+                    audio_data[audio_] = content     
 
         data_size = ['500', '1000', '2500', 'short', 'full']
 
@@ -131,6 +132,7 @@ class Test(datasets.GeneratorBasedBuilder):
                     gen_kwargs={
                         "filepath": downloaded_files[f"train_{size}"],
                         "audio_data": audio_data,
+                        "dl_manager":dl_manager,
                     },
                 )
             )
@@ -140,6 +142,7 @@ class Test(datasets.GeneratorBasedBuilder):
                     gen_kwargs={
                         "filepath": downloaded_files[f"val_{size}"],
                         "audio_data": audio_data,
+                        "dl_manager":dl_manager,
                     },
                 )
             )
@@ -149,6 +152,7 @@ class Test(datasets.GeneratorBasedBuilder):
                 gen_kwargs={
                     "filepath": downloaded_files["test_common"],
                     "audio_data": audio_data,
+                    "dl_manager":dl_manager,
                 },
             )
         )
@@ -158,6 +162,7 @@ class Test(datasets.GeneratorBasedBuilder):
                 gen_kwargs={
                     "filepath": downloaded_files["all_verses"],
                     "audio_data": audio_data,
+                    "dl_manager":dl_manager,
                 },
             )
         )
@@ -167,27 +172,49 @@ class Test(datasets.GeneratorBasedBuilder):
                 gen_kwargs={
                     "filepath": downloaded_files["short_verses"],
                     "audio_data": audio_data,
+                    "dl_manager":dl_manager,
                 },
             )
         )
         return splits
 
         
-    def _generate_examples(self, filepath, audio_data):
+    def _generate_examples(self, filepath, audio_data, dl_manager):
+
+        '''Function for parsing large csv archives (all_verses, short_verses)'''
+        def parse_archive(archive):
+            temp_df = pd.DataFrame() 
+            for path, file in dl_manager.iter_archive(archive):
+                if path.endswith('_verses.csv'):
+                    verses_filepath = file 
+                    verses_lines = file.readlines()
+                    verses_lines = [line.decode("utf-8") for line in verses_lines]
+                    column_names = verses_lines[0].strip().split(",")
+                    rows = [row.split(',') for row in verses_lines[1:]]
+                    rows = [[i[0], i[1], ','.join(i[2:])]for i in rows]
+                    temp_df = pd.DataFrame(rows, columns =column_names) 
+                    break
+            return temp_df
+
+        if filepath.endswith('all_verses.tar.gz'):
+            data_df = parse_archive(filepath)
+        elif filepath.endswith('short_verses.tar.gz'):    
+            data_df = parse_archive(filepath)
+        else:
+            with open(filepath) as f:
+                data_df = pd.read_csv(f,sep=',')  
+        
         key = 0
-        #print(list(audio_data.keys()))
-        with open(filepath) as f:
-            data_df = pd.read_csv(f,sep=',')
-            for index,row in data_df.iterrows():
-                audio = row['path'].split('/')[-1]
-                content = ''
-                if audio in list(audio_data.keys()):
-                    content = audio_data[audio]
-                else:    
-                    print(f"*********** Couldn't find audio: {audio} **************")
-                yield key, {
-                        "sentence": row["sentence"],
-                        "path": row["path"],
-                        "audio":{"path": row["path"], "bytes": content}
-                    }
-                key+=1
+        for index,row in data_df.iterrows():
+            audio = row['path'].split('/')[-1]
+            content = ''
+            if audio in list(audio_data.keys()):
+                content = audio_data[audio]
+            else:    
+                print(f"*********** Couldn't find audio: {audio} **************")
+            yield key, {
+                    "sentence": row["sentence"],
+                    "path": row["path"],
+                    "audio":{"path": row["path"], "bytes": content}
+                }
+            key+=1