full data

README.md

@@ -23,6 +23,29 @@ Please refer to [github repo](https://github.com/tzshi/squall/) for source data.
 from datasets import load_dataset
 dataset = load_dataset("siyue/squall","0")
 ```
+Example:
+```python
+{
+	'nt': 'nt-10922', 
+	'tbl': '204_879', 
+	'columns': 
+		{
+			'raw_header': ['year', 'host / location', 'division i overall', 'division i undergraduate', 'division ii overall', 'division ii community college'], 
+			'tokenized_header': [['year'], ['host', '\\\\/', 'location'], ['division', 'i', 'overall'], ['division', 'i', 'undergraduate'], ['division', 'ii', 'overall'], ['division', 'ii', 'community', 'college']], 
+			'column_suffixes': [['number'], ['address'], [], [], [], []], 
+			'column_dtype': ['number', 'address', 'text', 'text', 'text', 'text'], 
+			'example': ['1997', 'penn', 'chicago', 'swarthmore', 'harvard', 'valencia cc']
+		}, 
+	'nl': ['when', 'was', 'the', 'last', 'time', 'the', 'event', 'was', 'held', 'in', 'minnesota', '?'], 
+	'nl_pos': ['WRB', 'VBD-AUX', 'DT', 'JJ', 'NN', 'DT', 'NN', 'VBD-AUX', 'VBN', 'IN', 'NNP', '.'], 
+	'nl_ner': ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'LOCATION', 'O'], 
+	'nl_incolumns': [False, False, False, False, False, False, False, False, False, False, False, False], 
+	'nl_incells': [False, False, False, False, False, False, False, False, False, False, True, False], 
+	'columns_innl': [False, False, False, False, False, False], 
+	'tgt': '2007', 
+	'sql': ['select', 'c1', 'from', 'w', 'where', 'c2', '=', "'minnesota'", 'order', 'by', 'c1_number', 'desc', 'limit', '1']
+}
+```
 
 ## Contact
 For any issues or questions, kindly email us at: Siyue Zhang (siyue001@e.ntu.edu.sg).

squall.py

@@ -98,8 +98,19 @@ class Squall(datasets.GeneratorBasedBuilder):
                     "nl_incells": datasets.features.Sequence(datasets.Value("bool_")),
                     "columns_innl": datasets.features.Sequence(datasets.Value("bool_")),
                     "tgt": datasets.Value("string"),
-                    "sql": datasets.features.Sequence(datasets.Value("string"))
-                    # "align" is not implemented
+                    "sql": {
+                        "sql_type": datasets.features.Sequence(datasets.Value("string")), 
+                        "value": datasets.features.Sequence(datasets.Value("string")), 
+                        "span_indices": datasets.features.Sequence(datasets.features.Sequence(datasets.Value("int32")))
+                    },
+                    "nl_ralign": {
+                        "aligned_sql_token_type":datasets.features.Sequence(datasets.Value("string")),
+                        "aligned_sql_token_info":datasets.features.Sequence(datasets.Value("string")),
+                    },
+                    "align":{
+                        "nl_indices": datasets.features.Sequence(datasets.Value("int32")),
+                        "sql_indices": datasets.features.Sequence(datasets.Value("int32"))
+                    }
                 }
             ),
             # No default supervised_keys (as we have to pass both question
@@ -225,6 +236,16 @@ class Squall(datasets.GeneratorBasedBuilder):
                 instance["numbers"] = numbers
                 instance["has_number"] = has_number
 
+            def transform(sample_key, keys):
+                cols = {}
+                n_col = len(sample[sample_key])
+                for k in range(len(keys)):
+                    tmp = []
+                    for j in range(n_col):
+                        tmp.append(sample[sample_key][j][k])
+                    cols[keys[k]] = tmp
+                return cols
+
             with open(dev_ids) as f:
                 dev_ids = json.load(f)
             if split_key == "train":
@@ -233,15 +254,20 @@ class Squall(datasets.GeneratorBasedBuilder):
                 set = [x for x in squall_full_data if x["tbl"] in dev_ids]
             idx = 0
             for sample in set:
-                cols = {}
+
+                # transform columns
                 keys = ["raw_header", "tokenized_header", "column_suffixes", "column_dtype", "example"]
-                n_col = len(sample["columns"])
-                for k in range(5):
-                    tmp = []
-                    for j in range(n_col):
-                        tmp.append(sample["columns"][j][k])
-                    cols[keys[k]] = tmp
-                sql = [x[1] for x in sample["sql"]]
+                cols = transform("columns", keys)
+                # transform sql
+                keys = ["sql_type", "value", "span_indices"]
+                sqls = transform("sql", keys)
+                # transform align
+                keys = ["nl_indices", "sql_indices"]
+                aligns = transform("align", keys)
+                # transform ralign
+                keys = ["aligned_sql_token_type", "aligned_sql_token_info"]
+                raligns = transform("nl_ralign", keys)
+
                 yield idx, {
                     "nt": sample["nt"],
                     "tbl": sample["tbl"],
@@ -249,13 +275,13 @@ class Squall(datasets.GeneratorBasedBuilder):
                     "nl": sample["nl"],
                     "nl_pos": sample["nl_pos"],
                     "nl_ner": sample["nl_ner"],
-                    # "nl_ralign": sample["nl_ralign"],
+                    "nl_ralign": raligns,
                     "nl_incolumns": sample["nl_incolumns"],
                     "nl_incells": sample["nl_incells"],
                     "columns_innl": sample["columns_innl"],
                     "tgt": sample["tgt"],
-                    "sql": sql,
-                    # "align": sample["align"]
+                    "sql": sqls,
+                    "align": aligns
                 }
                 idx += 1
         else:
@@ -263,14 +289,11 @@ class Squall(datasets.GeneratorBasedBuilder):
                 test_data = json.load(f)
             idx = 0
             for sample in test_data:
-                cols = {}
+                
+                # transform columns
                 keys = ["raw_header", "tokenized_header", "column_suffixes", "column_dtype", "example"]
-                n_col = len(sample["columns"])
-                for k in range(5):
-                    tmp = []
-                    for j in range(n_col):
-                        tmp.append(sample["columns"][j][k])
-                    cols[keys[k]] = tmp
+                cols = transform("columns", keys)
+
                 yield idx, {
                     "nt": sample["nt"],
                     "tbl": sample["tbl"],
@@ -278,12 +301,12 @@ class Squall(datasets.GeneratorBasedBuilder):
                     "nl": sample["nl"],
                     "nl_pos": sample["nl_pos"],
                     "nl_ner": sample["nl_ner"],
-                    # "nl_ralign": sample["nl_ralign"],
+                    "nl_ralign": [],
                     "nl_incolumns": sample["nl_incolumns"],
                     "nl_incells": sample["nl_incells"],
                     "columns_innl": sample["columns_innl"],
                     "tgt": '',
                     "sql": [],
-                    # "align": sample["align"]
+                    "align": []
                 }
                 idx += 1