diff options
Diffstat (limited to 'sqlglot/parser.py')
| -rw-r--r-- | sqlglot/parser.py | 78 |
1 files changed, 58 insertions, 20 deletions
diff --git a/sqlglot/parser.py b/sqlglot/parser.py index d6888c7..a9648ef 100644 --- a/sqlglot/parser.py +++ b/sqlglot/parser.py @@ -9,7 +9,7 @@ from sqlglot.errors import ErrorLevel, ParseError, concat_messages, merge_errors from sqlglot.helper import apply_index_offset, ensure_list, seq_get from sqlglot.time import format_time from sqlglot.tokens import Token, Tokenizer, TokenType -from sqlglot.trie import in_trie, new_trie +from sqlglot.trie import TrieResult, in_trie, new_trie if t.TYPE_CHECKING: from sqlglot._typing import E @@ -177,6 +177,7 @@ class Parser(metaclass=_Parser): TokenType.BIGSERIAL, TokenType.XML, TokenType.UNIQUEIDENTIFIER, + TokenType.USERDEFINED, TokenType.MONEY, TokenType.SMALLMONEY, TokenType.ROWVERSION, @@ -465,7 +466,7 @@ class Parser(metaclass=_Parser): } EXPRESSION_PARSERS = { - exp.Cluster: lambda self: self._parse_sort(exp.Cluster, "CLUSTER", "BY"), + exp.Cluster: lambda self: self._parse_sort(exp.Cluster, TokenType.CLUSTER_BY), exp.Column: lambda self: self._parse_column(), exp.Condition: lambda self: self._parse_conjunction(), exp.DataType: lambda self: self._parse_types(), @@ -484,7 +485,7 @@ class Parser(metaclass=_Parser): exp.Properties: lambda self: self._parse_properties(), exp.Qualify: lambda self: self._parse_qualify(), exp.Returning: lambda self: self._parse_returning(), - exp.Sort: lambda self: self._parse_sort(exp.Sort, "SORT", "BY"), + exp.Sort: lambda self: self._parse_sort(exp.Sort, TokenType.SORT_BY), exp.Table: lambda self: self._parse_table_parts(), exp.TableAlias: lambda self: self._parse_table_alias(), exp.Where: lambda self: self._parse_where(), @@ -540,8 +541,7 @@ class Parser(metaclass=_Parser): exp.Literal, this=token.text, is_string=False ), TokenType.STAR: lambda self, _: self.expression( - exp.Star, - **{"except": self._parse_except(), "replace": self._parse_replace()}, + exp.Star, **{"except": self._parse_except(), "replace": self._parse_replace()} ), TokenType.NULL: lambda self, _: self.expression(exp.Null), TokenType.TRUE: lambda self, _: self.expression(exp.Boolean, this=True), @@ -584,9 +584,10 @@ class Parser(metaclass=_Parser): "BLOCKCOMPRESSION": lambda self: self._parse_blockcompression(), "CHARACTER SET": lambda self: self._parse_character_set(), "CHECKSUM": lambda self: self._parse_checksum(), - "CLUSTER": lambda self: self._parse_cluster(), + "CLUSTER BY": lambda self: self._parse_cluster(), "COLLATE": lambda self: self._parse_property_assignment(exp.CollateProperty), "COMMENT": lambda self: self._parse_property_assignment(exp.SchemaCommentProperty), + "COPY": lambda self: self._parse_copy_property(), "DATABLOCKSIZE": lambda self, **kwargs: self._parse_datablocksize(**kwargs), "DEFINER": lambda self: self._parse_definer(), "DETERMINISTIC": lambda self: self.expression( @@ -780,6 +781,8 @@ class Parser(metaclass=_Parser): CLONE_KINDS = {"TIMESTAMP", "OFFSET", "STATEMENT"} + TABLE_INDEX_HINT_TOKENS = {TokenType.FORCE, TokenType.IGNORE, TokenType.USE} + WINDOW_ALIAS_TOKENS = ID_VAR_TOKENS - {TokenType.ROWS} WINDOW_BEFORE_PAREN_TOKENS = {TokenType.OVER} WINDOW_SIDES = {"FOLLOWING", "PRECEDING"} @@ -788,7 +791,8 @@ class Parser(metaclass=_Parser): STRICT_CAST = True - CONCAT_NULL_OUTPUTS_STRING = False # A NULL arg in CONCAT yields NULL by default + # A NULL arg in CONCAT yields NULL by default + CONCAT_NULL_OUTPUTS_STRING = False CONVERT_TYPE_FIRST = False @@ -1423,11 +1427,14 @@ class Parser(metaclass=_Parser): return self.expression(exp.ChecksumProperty, on=on, default=self._match(TokenType.DEFAULT)) def _parse_cluster(self) -> t.Optional[exp.Cluster]: - if not self._match_text_seq("BY"): + return self.expression(exp.Cluster, expressions=self._parse_csv(self._parse_ordered)) + + def _parse_copy_property(self) -> t.Optional[exp.CopyGrantsProperty]: + if not self._match_text_seq("GRANTS"): self._retreat(self._index - 1) return None - return self.expression(exp.Cluster, expressions=self._parse_csv(self._parse_ordered)) + return self.expression(exp.CopyGrantsProperty) def _parse_freespace(self) -> exp.FreespaceProperty: self._match(TokenType.EQ) @@ -1779,6 +1786,7 @@ class Parser(metaclass=_Parser): using=self._parse_csv(lambda: self._match(TokenType.USING) and self._parse_table()), where=self._parse_where(), returning=self._parse_returning(), + limit=self._parse_limit(), ) def _parse_update(self) -> exp.Update: @@ -1790,6 +1798,7 @@ class Parser(metaclass=_Parser): "from": self._parse_from(modifiers=True), "where": self._parse_where(), "returning": self._parse_returning(), + "limit": self._parse_limit(), }, ) @@ -2268,6 +2277,33 @@ class Parser(metaclass=_Parser): partition_by=self._parse_partition_by(), ) + def _parse_table_hints(self) -> t.Optional[t.List[exp.Expression]]: + hints: t.List[exp.Expression] = [] + if self._match_pair(TokenType.WITH, TokenType.L_PAREN): + # https://learn.microsoft.com/en-us/sql/t-sql/queries/hints-transact-sql-table?view=sql-server-ver16 + hints.append( + self.expression( + exp.WithTableHint, + expressions=self._parse_csv( + lambda: self._parse_function() or self._parse_var(any_token=True) + ), + ) + ) + self._match_r_paren() + else: + # https://dev.mysql.com/doc/refman/8.0/en/index-hints.html + while self._match_set(self.TABLE_INDEX_HINT_TOKENS): + hint = exp.IndexTableHint(this=self._prev.text.upper()) + + self._match_texts({"INDEX", "KEY"}) + if self._match(TokenType.FOR): + hint.set("target", self._advance_any() and self._prev.text.upper()) + + hint.set("expressions", self._parse_wrapped_id_vars()) + hints.append(hint) + + return hints or None + def _parse_table_part(self, schema: bool = False) -> t.Optional[exp.Expression]: return ( (not schema and self._parse_function(optional_parens=False)) @@ -2335,12 +2371,7 @@ class Parser(metaclass=_Parser): if not this.args.get("pivots"): this.set("pivots", self._parse_pivots()) - if self._match_pair(TokenType.WITH, TokenType.L_PAREN): - this.set( - "hints", - self._parse_csv(lambda: self._parse_function() or self._parse_var(any_token=True)), - ) - self._match_r_paren() + this.set("hints", self._parse_table_hints()) if not self.ALIAS_POST_TABLESAMPLE: table_sample = self._parse_table_sample() @@ -2610,8 +2641,8 @@ class Parser(metaclass=_Parser): exp.Order, this=this, expressions=self._parse_csv(self._parse_ordered) ) - def _parse_sort(self, exp_class: t.Type[E], *texts: str) -> t.Optional[E]: - if not self._match_text_seq(*texts): + def _parse_sort(self, exp_class: t.Type[E], token: TokenType) -> t.Optional[E]: + if not self._match(token): return None return self.expression(exp_class, expressions=self._parse_csv(self._parse_ordered)) @@ -3655,7 +3686,11 @@ class Parser(metaclass=_Parser): def _parse_concat(self) -> t.Optional[exp.Expression]: args = self._parse_csv(self._parse_conjunction) if self.CONCAT_NULL_OUTPUTS_STRING: - args = [exp.func("COALESCE", arg, exp.Literal.string("")) for arg in args] + args = [ + exp.func("COALESCE", exp.cast(arg, "text"), exp.Literal.string("")) + for arg in args + if arg + ] # Some dialects (e.g. Trino) don't allow a single-argument CONCAT call, so when # we find such a call we replace it with its argument. @@ -4553,13 +4588,16 @@ class Parser(metaclass=_Parser): curr = self._curr.text.upper() key = curr.split(" ") this.append(curr) + self._advance() result, trie = in_trie(trie, key) - if result == 0: + if result == TrieResult.FAILED: break - if result == 2: + + if result == TrieResult.EXISTS: subparser = parsers[" ".join(this)] return subparser + self._retreat(index) return None |