Added HQL Keywords

Author: mathilde.oustlant

sqlparse/keywords.py

@@ -16,6 +16,7 @@ def is_keyword(value):
     return (KEYWORDS_COMMON.get(val)
             or KEYWORDS_ORACLE.get(val)
             or KEYWORDS_PLPGSQL.get(val)
+            or KEYWORDS_HQL.get(val)
             or KEYWORDS.get(val, tokens.Name)), value
 
 
@@ -63,8 +64,8 @@ def is_keyword(value):
         (r'[A-ZÀ-Ü]\w*(?=\()', tokens.Name),  # side effect: change kw to func
         (r'-?0x[\dA-F]+', tokens.Number.Hexadecimal),
         (r'-?\d*(\.\d+)?E-?\d+', tokens.Number.Float),
-        (r'-?(\d+(\.\d*)|\.\d+)', tokens.Number.Float),
-        (r'-?\d+(?![_A-ZÀ-Ü])', tokens.Number.Integer),
+        (r'(?![_A-ZÀ-Ü])-?(\d+(\.\d*)|\.\d+)(?![_A-ZÀ-Ü])', tokens.Number.Float),
+        (r'(?![_A-ZÀ-Ü])-?\d+(?![_A-ZÀ-Ü])', tokens.Number.Integer),
         (r"'(''|\\\\|\\'|[^'])*'", tokens.String.Single),
         # not a real string literal in ANSI SQL:
         (r'"(""|\\\\|\\"|[^"])*"', tokens.String.Symbol),
@@ -82,6 +83,7 @@ def is_keyword(value):
         (r'DOUBLE\s+PRECISION\b', tokens.Name.Builtin),
         (r'GROUP\s+BY\b', tokens.Keyword),
         (r'ORDER\s+BY\b', tokens.Keyword),
+        (r'(LATERAL\s+VIEW\s+)(EXPLODE|INLINE|PARSE_URL_TUPLE|POSEXPLODE|STACK)\b', tokens.Keyword),
 
         (r'[0-9_A-ZÀ-Ü][_$#\w]*', is_keyword),
 
@@ -634,26 +636,39 @@ def is_keyword(value):
     'DATE': tokens.Name.Builtin,
     'DEC': tokens.Name.Builtin,
     'DECIMAL': tokens.Name.Builtin,
+    'FILE_TYPE': tokens.Name.Builtin,
     'FLOAT': tokens.Name.Builtin,
     'INT': tokens.Name.Builtin,
     'INT8': tokens.Name.Builtin,
     'INTEGER': tokens.Name.Builtin,
     'INTERVAL': tokens.Name.Builtin,
     'LONG': tokens.Name.Builtin,
+    'NATURALN': tokens.Name.Builtin,
+    'NVARCHAR': tokens.Name.Builtin,
     'NUMBER': tokens.Name.Builtin,
     'NUMERIC': tokens.Name.Builtin,
+    'PLS_INTEGER': tokens.Name.Builtin,
+    'POSITIVE': tokens.Name.Builtin,
+    'POSITIVEN': tokens.Name.Builtin,
     'REAL': tokens.Name.Builtin,
     'ROWID': tokens.Name.Builtin,
     'ROWLABEL': tokens.Name.Builtin,
     'ROWNUM': tokens.Name.Builtin,
     'SERIAL': tokens.Name.Builtin,
     'SERIAL8': tokens.Name.Builtin,
     'SIGNED': tokens.Name.Builtin,
+    'SIGNTYPE': tokens.Name.Builtin,
+    'SIMPLE_DOUBLE': tokens.Name.Builtin,
+    'SIMPLE_FLOAT': tokens.Name.Builtin,
+    'SIMPLE_INTEGER': tokens.Name.Builtin,
     'SMALLINT': tokens.Name.Builtin,
+    'SYS_REFCURSOR': tokens.Name.Builtin,
     'SYSDATE': tokens.Name,
     'TEXT': tokens.Name.Builtin,
     'TINYINT': tokens.Name.Builtin,
     'UNSIGNED': tokens.Name.Builtin,
+    'UROWID': tokens.Name.Builtin,
+    'UTL_FILE': tokens.Name.Builtin,
     'VARCHAR': tokens.Name.Builtin,
     'VARCHAR2': tokens.Name.Builtin,
     'VARYING': tokens.Name.Builtin,
@@ -854,3 +869,83 @@ def is_keyword(value):
     'IN': tokens.Keyword,
     'LOOP': tokens.Keyword,
 }
+
+# Hive Syntax
+KEYWORDS_HQL = {
+    'EXPLODE': tokens.Keyword,
+    'DIRECTORY': tokens.Keyword,
+    'DISTRIBUTE': tokens.Keyword,
+    'INCLUDE': tokens.Keyword,
+    'LOCATE': tokens.Keyword,
+    'OVERWRITE': tokens.Keyword,
+    'POSEXPLODE': tokens.Keyword,
+
+    'ARRAY_CONTAINS': tokens.Keyword,
+    'CMP': tokens.Keyword,
+    'COLLECT_LIST': tokens.Keyword,
+    'CONCAT': tokens.Keyword,
+    'CONDITION': tokens.Keyword,
+    'DATE_ADD': tokens.Keyword,
+    'DATE_SUB': tokens.Keyword,
+    'DECODE': tokens.Keyword,
+    'DBMS_OUTPUT': tokens.Keyword,
+    'ELEMENTS': tokens.Keyword,
+    'EXCHANGE': tokens.Keyword,
+    'EXTENDED': tokens.Keyword,
+    'FLOOR': tokens.Keyword,
+    'FOLLOWING': tokens.Keyword,
+    'FROM_UNIXTIME': tokens.Keyword,
+    'FTP': tokens.Keyword,
+    'HOUR': tokens.Keyword,
+    'INLINE': tokens.Keyword,
+    'INSTR': tokens.Keyword,
+    'LEN': tokens.Keyword,
+    'MAXELEMENT': tokens.Keyword,
+    'MAXINDEX': tokens.Keyword,
+    'MAX_PART_DATE': tokens.Keyword,
+    'MAX_PART_INT': tokens.Keyword,
+    'MAX_PART_STRING': tokens.Keyword,
+    'MINELEMENT': tokens.Keyword,
+    'MININDEX': tokens.Keyword,
+    'MIN_PART_DATE': tokens.Keyword,
+    'MIN_PART_INT': tokens.Keyword,
+    'MIN_PART_STRING': tokens.Keyword,
+    'NOW': tokens.Keyword,
+    'NVL': tokens.Keyword,
+    'NVL2': tokens.Keyword,
+    'PARSE_URL_TUPLE': tokens.Keyword,
+    'PART_LOC': tokens.Keyword,
+    'PART_COUNT': tokens.Keyword,
+    'PART_COUNT_BY': tokens.Keyword,
+    'PRINT': tokens.Keyword,
+    'PUT_LINE': tokens.Keyword,
+    'RANGE': tokens.Keyword,
+    'REDUCE': tokens.Keyword,
+    'REGEXP_REPLACE': tokens.Keyword,
+    'RESIGNAL': tokens.Keyword,
+    'RTRIM': tokens.Keyword,
+    'SIGN': tokens.Keyword,
+    'SIGNAL': tokens.Keyword,
+    'SIN': tokens.Keyword,
+    'SPLIT': tokens.Keyword,
+    'SQRT': tokens.Keyword,
+    'STACK': tokens.Keyword,
+    'STR': tokens.Keyword,
+    'SUBSTR': tokens.Keyword,
+    'SUMMARY': tokens.Keyword,
+    'TBLPROPERTIES': tokens.Keyword,
+    'TIMESTAMP_ISO': tokens.Keyword,
+    'TO_CHAR': tokens.Keyword,
+    'TO_DATE': tokens.Keyword,
+    'TO_TIMESTAMP': tokens.Keyword,
+    'TRUNC': tokens.Keyword,
+    'UNBOUNDED': tokens.Keyword,
+    'UNIQUEJOIN': tokens.Keyword,
+    'UNIX_TIMESTAMP': tokens.Keyword,
+    'UTC_TIMESTAMP': tokens.Keyword,
+    'VIEWS': tokens.Keyword,
+
+    'EXIT': tokens.Keyword,
+    'BREAK': tokens.Keyword,
+    'LEAVE': tokens.Keyword,
+}