sqlglot.dialects.sqlite
1from __future__ import annotations 2 3from sqlglot import exp, generator, parser, tokens 4from sqlglot.dialects.dialect import ( 5 Dialect, 6 arrow_json_extract_scalar_sql, 7 arrow_json_extract_sql, 8 no_ilike_sql, 9 no_tablesample_sql, 10 no_trycast_sql, 11 rename_func, 12) 13from sqlglot.tokens import TokenType 14 15 16def _fetch_sql(self, expression): 17 return self.limit_sql(exp.Limit(expression=expression.args.get("count"))) 18 19 20# https://www.sqlite.org/lang_aggfunc.html#group_concat 21def _group_concat_sql(self, expression): 22 this = expression.this 23 distinct = expression.find(exp.Distinct) 24 if distinct: 25 this = distinct.expressions[0] 26 distinct = "DISTINCT " 27 28 if isinstance(expression.this, exp.Order): 29 self.unsupported("SQLite GROUP_CONCAT doesn't support ORDER BY.") 30 if expression.this.this and not distinct: 31 this = expression.this.this 32 33 separator = expression.args.get("separator") 34 return f"GROUP_CONCAT({distinct or ''}{self.format_args(this, separator)})" 35 36 37def _date_add_sql(self, expression): 38 modifier = expression.expression 39 modifier = expression.name if modifier.is_string else self.sql(modifier) 40 unit = expression.args.get("unit") 41 modifier = f"'{modifier} {unit.name}'" if unit else f"'{modifier}'" 42 return self.func("DATE", expression.this, modifier) 43 44 45class SQLite(Dialect): 46 class Tokenizer(tokens.Tokenizer): 47 IDENTIFIERS = ['"', ("[", "]"), "`"] 48 HEX_STRINGS = [("x'", "'"), ("X'", "'"), ("0x", ""), ("0X", "")] 49 50 KEYWORDS = { 51 **tokens.Tokenizer.KEYWORDS, 52 } 53 54 class Parser(parser.Parser): 55 FUNCTIONS = { 56 **parser.Parser.FUNCTIONS, # type: ignore 57 "EDITDIST3": exp.Levenshtein.from_arg_list, 58 } 59 60 class Generator(generator.Generator): 61 TYPE_MAPPING = { 62 **generator.Generator.TYPE_MAPPING, # type: ignore 63 exp.DataType.Type.BOOLEAN: "INTEGER", 64 exp.DataType.Type.TINYINT: "INTEGER", 65 exp.DataType.Type.SMALLINT: "INTEGER", 66 exp.DataType.Type.INT: "INTEGER", 67 exp.DataType.Type.BIGINT: "INTEGER", 68 exp.DataType.Type.FLOAT: "REAL", 69 exp.DataType.Type.DOUBLE: "REAL", 70 exp.DataType.Type.DECIMAL: "REAL", 71 exp.DataType.Type.CHAR: "TEXT", 72 exp.DataType.Type.NCHAR: "TEXT", 73 exp.DataType.Type.VARCHAR: "TEXT", 74 exp.DataType.Type.NVARCHAR: "TEXT", 75 exp.DataType.Type.BINARY: "BLOB", 76 exp.DataType.Type.VARBINARY: "BLOB", 77 } 78 79 TOKEN_MAPPING = { 80 TokenType.AUTO_INCREMENT: "AUTOINCREMENT", 81 } 82 83 TRANSFORMS = { 84 **generator.Generator.TRANSFORMS, # type: ignore 85 exp.DateAdd: _date_add_sql, 86 exp.ILike: no_ilike_sql, 87 exp.JSONExtract: arrow_json_extract_sql, 88 exp.JSONExtractScalar: arrow_json_extract_scalar_sql, 89 exp.JSONBExtract: arrow_json_extract_sql, 90 exp.JSONBExtractScalar: arrow_json_extract_scalar_sql, 91 exp.Levenshtein: rename_func("EDITDIST3"), 92 exp.TableSample: no_tablesample_sql, 93 exp.DateStrToDate: lambda self, e: self.sql(e, "this"), 94 exp.TimeStrToTime: lambda self, e: self.sql(e, "this"), 95 exp.TryCast: no_trycast_sql, 96 exp.GroupConcat: _group_concat_sql, 97 exp.Fetch: _fetch_sql, 98 } 99 100 def transaction_sql(self, expression): 101 this = expression.this 102 this = f" {this}" if this else "" 103 return f"BEGIN{this} TRANSACTION"
46class SQLite(Dialect): 47 class Tokenizer(tokens.Tokenizer): 48 IDENTIFIERS = ['"', ("[", "]"), "`"] 49 HEX_STRINGS = [("x'", "'"), ("X'", "'"), ("0x", ""), ("0X", "")] 50 51 KEYWORDS = { 52 **tokens.Tokenizer.KEYWORDS, 53 } 54 55 class Parser(parser.Parser): 56 FUNCTIONS = { 57 **parser.Parser.FUNCTIONS, # type: ignore 58 "EDITDIST3": exp.Levenshtein.from_arg_list, 59 } 60 61 class Generator(generator.Generator): 62 TYPE_MAPPING = { 63 **generator.Generator.TYPE_MAPPING, # type: ignore 64 exp.DataType.Type.BOOLEAN: "INTEGER", 65 exp.DataType.Type.TINYINT: "INTEGER", 66 exp.DataType.Type.SMALLINT: "INTEGER", 67 exp.DataType.Type.INT: "INTEGER", 68 exp.DataType.Type.BIGINT: "INTEGER", 69 exp.DataType.Type.FLOAT: "REAL", 70 exp.DataType.Type.DOUBLE: "REAL", 71 exp.DataType.Type.DECIMAL: "REAL", 72 exp.DataType.Type.CHAR: "TEXT", 73 exp.DataType.Type.NCHAR: "TEXT", 74 exp.DataType.Type.VARCHAR: "TEXT", 75 exp.DataType.Type.NVARCHAR: "TEXT", 76 exp.DataType.Type.BINARY: "BLOB", 77 exp.DataType.Type.VARBINARY: "BLOB", 78 } 79 80 TOKEN_MAPPING = { 81 TokenType.AUTO_INCREMENT: "AUTOINCREMENT", 82 } 83 84 TRANSFORMS = { 85 **generator.Generator.TRANSFORMS, # type: ignore 86 exp.DateAdd: _date_add_sql, 87 exp.ILike: no_ilike_sql, 88 exp.JSONExtract: arrow_json_extract_sql, 89 exp.JSONExtractScalar: arrow_json_extract_scalar_sql, 90 exp.JSONBExtract: arrow_json_extract_sql, 91 exp.JSONBExtractScalar: arrow_json_extract_scalar_sql, 92 exp.Levenshtein: rename_func("EDITDIST3"), 93 exp.TableSample: no_tablesample_sql, 94 exp.DateStrToDate: lambda self, e: self.sql(e, "this"), 95 exp.TimeStrToTime: lambda self, e: self.sql(e, "this"), 96 exp.TryCast: no_trycast_sql, 97 exp.GroupConcat: _group_concat_sql, 98 exp.Fetch: _fetch_sql, 99 } 100 101 def transaction_sql(self, expression): 102 this = expression.this 103 this = f" {this}" if this else "" 104 return f"BEGIN{this} TRANSACTION"
Inherited Members
47 class Tokenizer(tokens.Tokenizer): 48 IDENTIFIERS = ['"', ("[", "]"), "`"] 49 HEX_STRINGS = [("x'", "'"), ("X'", "'"), ("0x", ""), ("0X", "")] 50 51 KEYWORDS = { 52 **tokens.Tokenizer.KEYWORDS, 53 }
Inherited Members
55 class Parser(parser.Parser): 56 FUNCTIONS = { 57 **parser.Parser.FUNCTIONS, # type: ignore 58 "EDITDIST3": exp.Levenshtein.from_arg_list, 59 }
Parser consumes a list of tokens produced by the sqlglot.tokens.Tokenizer
and produces
a parsed syntax tree.
Arguments:
- error_level: the desired error level. Default: ErrorLevel.RAISE
- error_message_context: determines the amount of context to capture from a query string when displaying the error message (in number of characters). Default: 50.
- index_offset: Index offset for arrays eg ARRAY[0] vs ARRAY[1] as the head of a list. Default: 0
- alias_post_tablesample: If the table alias comes after tablesample. Default: False
- max_errors: Maximum number of error messages to include in a raised ParseError. This is only relevant if error_level is ErrorLevel.RAISE. Default: 3
- null_ordering: Indicates the default null ordering method to use if not explicitly set. Options are "nulls_are_small", "nulls_are_large", "nulls_are_last". Default: "nulls_are_small"
Inherited Members
61 class Generator(generator.Generator): 62 TYPE_MAPPING = { 63 **generator.Generator.TYPE_MAPPING, # type: ignore 64 exp.DataType.Type.BOOLEAN: "INTEGER", 65 exp.DataType.Type.TINYINT: "INTEGER", 66 exp.DataType.Type.SMALLINT: "INTEGER", 67 exp.DataType.Type.INT: "INTEGER", 68 exp.DataType.Type.BIGINT: "INTEGER", 69 exp.DataType.Type.FLOAT: "REAL", 70 exp.DataType.Type.DOUBLE: "REAL", 71 exp.DataType.Type.DECIMAL: "REAL", 72 exp.DataType.Type.CHAR: "TEXT", 73 exp.DataType.Type.NCHAR: "TEXT", 74 exp.DataType.Type.VARCHAR: "TEXT", 75 exp.DataType.Type.NVARCHAR: "TEXT", 76 exp.DataType.Type.BINARY: "BLOB", 77 exp.DataType.Type.VARBINARY: "BLOB", 78 } 79 80 TOKEN_MAPPING = { 81 TokenType.AUTO_INCREMENT: "AUTOINCREMENT", 82 } 83 84 TRANSFORMS = { 85 **generator.Generator.TRANSFORMS, # type: ignore 86 exp.DateAdd: _date_add_sql, 87 exp.ILike: no_ilike_sql, 88 exp.JSONExtract: arrow_json_extract_sql, 89 exp.JSONExtractScalar: arrow_json_extract_scalar_sql, 90 exp.JSONBExtract: arrow_json_extract_sql, 91 exp.JSONBExtractScalar: arrow_json_extract_scalar_sql, 92 exp.Levenshtein: rename_func("EDITDIST3"), 93 exp.TableSample: no_tablesample_sql, 94 exp.DateStrToDate: lambda self, e: self.sql(e, "this"), 95 exp.TimeStrToTime: lambda self, e: self.sql(e, "this"), 96 exp.TryCast: no_trycast_sql, 97 exp.GroupConcat: _group_concat_sql, 98 exp.Fetch: _fetch_sql, 99 } 100 101 def transaction_sql(self, expression): 102 this = expression.this 103 this = f" {this}" if this else "" 104 return f"BEGIN{this} TRANSACTION"
Generator interprets the given syntax tree and produces a SQL string as an output.
Arguments:
- time_mapping (dict): the dictionary of custom time mappings in which the key represents a python time format and the output the target time format
- time_trie (trie): a trie of the time_mapping keys
- pretty (bool): if set to True the returned string will be formatted. Default: False.
- quote_start (str): specifies which starting character to use to delimit quotes. Default: '.
- quote_end (str): specifies which ending character to use to delimit quotes. Default: '.
- identifier_start (str): specifies which starting character to use to delimit identifiers. Default: ".
- identifier_end (str): specifies which ending character to use to delimit identifiers. Default: ".
- identify (bool): if set to True all identifiers will be delimited by the corresponding character.
- normalize (bool): if set to True all identifiers will lower cased
- string_escape (str): specifies a string escape character. Default: '.
- identifier_escape (str): specifies an identifier escape character. Default: ".
- pad (int): determines padding in a formatted string. Default: 2.
- indent (int): determines the size of indentation in a formatted string. Default: 4.
- unnest_column_only (bool): if true unnest table aliases are considered only as column aliases
- normalize_functions (str): normalize function names, "upper", "lower", or None Default: "upper"
- alias_post_tablesample (bool): if the table alias comes after tablesample Default: False
- unsupported_level (ErrorLevel): determines the generator's behavior when it encounters unsupported expressions. Default ErrorLevel.WARN.
- null_ordering (str): Indicates the default null ordering method to use if not explicitly set. Options are "nulls_are_small", "nulls_are_large", "nulls_are_last". Default: "nulls_are_small"
- max_unsupported (int): Maximum number of unsupported messages to include in a raised UnsupportedError. This is only relevant if unsupported_level is ErrorLevel.RAISE. Default: 3
- leading_comma (bool): if the the comma is leading or trailing in select statements Default: False
- max_text_width: The max number of characters in a segment before creating new lines in pretty mode. The default is on the smaller end because the length only represents a segment and not the true line length. Default: 80
- comments: Whether or not to preserve comments in the output SQL code. Default: True
Inherited Members
- sqlglot.generator.Generator
- Generator
- generate
- unsupported
- sep
- seg
- pad_comment
- maybe_comment
- wrap
- no_identify
- normalize_func
- indent
- sql
- uncache_sql
- cache_sql
- characterset_sql
- column_sql
- columndef_sql
- columnconstraint_sql
- autoincrementcolumnconstraint_sql
- compresscolumnconstraint_sql
- generatedasidentitycolumnconstraint_sql
- notnullcolumnconstraint_sql
- primarykeycolumnconstraint_sql
- uniquecolumnconstraint_sql
- create_sql
- describe_sql
- prepend_ctes
- with_sql
- cte_sql
- tablealias_sql
- bitstring_sql
- hexstring_sql
- datatype_sql
- directory_sql
- delete_sql
- drop_sql
- except_sql
- except_op
- fetch_sql
- filter_sql
- hint_sql
- index_sql
- identifier_sql
- national_sql
- partition_sql
- properties_sql
- root_properties
- properties
- with_properties
- locate_properties
- property_sql
- likeproperty_sql
- fallbackproperty_sql
- journalproperty_sql
- freespaceproperty_sql
- afterjournalproperty_sql
- checksumproperty_sql
- mergeblockratioproperty_sql
- datablocksizeproperty_sql
- blockcompressionproperty_sql
- isolatedloadingproperty_sql
- lockingproperty_sql
- withdataproperty_sql
- insert_sql
- intersect_sql
- intersect_op
- introducer_sql
- pseudotype_sql
- rowformatdelimitedproperty_sql
- table_sql
- tablesample_sql
- pivot_sql
- tuple_sql
- update_sql
- values_sql
- var_sql
- into_sql
- from_sql
- group_sql
- having_sql
- join_sql
- lambda_sql
- lateral_sql
- limit_sql
- offset_sql
- lock_sql
- literal_sql
- loaddata_sql
- null_sql
- boolean_sql
- order_sql
- cluster_sql
- distribute_sql
- sort_sql
- ordered_sql
- matchrecognize_sql
- query_modifiers
- select_sql
- schema_sql
- star_sql
- structkwarg_sql
- parameter_sql
- sessionparameter_sql
- placeholder_sql
- subquery_sql
- qualify_sql
- union_sql
- union_op
- unnest_sql
- where_sql
- window_sql
- partition_by_sql
- window_spec_sql
- withingroup_sql
- between_sql
- bracket_sql
- all_sql
- any_sql
- exists_sql
- case_sql
- constraint_sql
- extract_sql
- trim_sql
- concat_sql
- check_sql
- foreignkey_sql
- primarykey_sql
- unique_sql
- if_sql
- in_sql
- in_unnest_op
- interval_sql
- return_sql
- reference_sql
- anonymous_sql
- paren_sql
- neg_sql
- not_sql
- alias_sql
- aliases_sql
- attimezone_sql
- add_sql
- and_sql
- connector_sql
- bitwiseand_sql
- bitwiseleftshift_sql
- bitwisenot_sql
- bitwiseor_sql
- bitwiserightshift_sql
- bitwisexor_sql
- cast_sql
- currentdate_sql
- collate_sql
- command_sql
- commit_sql
- rollback_sql
- altercolumn_sql
- renametable_sql
- altertable_sql
- droppartition_sql
- addconstraint_sql
- distinct_sql
- ignorenulls_sql
- respectnulls_sql
- intdiv_sql
- dpipe_sql
- div_sql
- distance_sql
- dot_sql
- eq_sql
- escape_sql
- glob_sql
- gt_sql
- gte_sql
- ilike_sql
- is_sql
- like_sql
- similarto_sql
- lt_sql
- lte_sql
- mod_sql
- mul_sql
- neq_sql
- nullsafeeq_sql
- nullsafeneq_sql
- or_sql
- slice_sql
- sub_sql
- trycast_sql
- use_sql
- binary
- function_fallback_sql
- func
- format_args
- text_width
- format_time
- expressions
- op_expressions
- naked_property
- set_operation
- tag_sql
- token_sql
- userdefinedfunction_sql
- joinhint_sql
- kwarg_sql
- when_sql
- merge_sql