3 SQL::Abstract::Manual::Specification
7 This discusses the specification for the AST provided by L<SQL::Abstract>. It is
8 meant to describe how the AST is structured, various components provided by
9 L<SQL::Abstract> for use with this AST, how to manipulate the AST, and various
10 uses for the AST once it is generated.
14 L<SQL::Abstract> has been in use for many years. Originally created to handle
15 the where-clause formation found in L<DBIx::Abstract>, it was generalized to
16 manage the creation of any SQL statement through the use of Perl structures.
17 Through the beating it received as the SQL generation syntax for L<DBIx::Class>,
18 various deficiencies were found and a generalized SQL AST was designed. This
19 document describes that AST.
23 The goals for this AST are as follows:
25 =head2 SQL-specific semantics
27 Instead of attempting to be an AST to handle any form of query, this will
28 instead be specialized to manage SQL queries (and queries that map to SQL
29 queries). This means that there will be support for SQL-specific features, such
32 =head2 Perl-specific semantics
34 This AST is meant to be used from within Perl5 only. So, it will take advantage
35 of as many Perl-specific features that make sense to use. No attempt whatosever
36 will be made to make this AST work within any other language, including Perl6.
38 =head2 Whole-lifecycle management
40 Whether a query is built out of whole cloth in one shot or cobbled together from
41 several snippets over the lifetime of a process, this AST will support any way
42 to construct the query. Queries can also be built from other queries, so an
43 UPDATE statement could be used as the basis for a SELECT statement, DELETE
44 statement, or even a DDL statement of some kind.
46 =head2 Dialect-agnostic usage
48 Even though SQL itself has several ANSI specifications (SQL-92 and SQL-99 among
49 them), this only serves as a basis for what a given RDBMS will expect. However,
50 every engine has its own specific extensions and specific ways of handling
51 common features. The AST will provide ways of expressing common functionality in
52 a common language. The emitters (objects that follow the Visitor pattern) will
53 be responsible for converting that common language into RDBMS-specific SQL.
57 The following are the restrictions upon the AST:
61 The AST will only support DML (Data Modelling Language). It will not (currently)
62 support DDL (Data Definition Language). Practically, this means that the only
63 statements supported will be:
77 Additional DML statements may be supported by specific Visitors (such as a
78 MySQL visitor supporting REPLACE INTO). q.v. the relevant sections of this
79 specification for details.
81 =head2 Dialect-agnostic construction
83 The AST will not attempt to be immediately readable to a human as SQL. In fact,
84 due to the dialect differences, particularly in terms of which use operators and
85 which use functions for a given action, the AST will provide simple units. It is
86 the responsibility of the Visitor to provide the appropriate SQL. Furthermore,
87 the AST will be very generic and only provide hints for a subset of SQL. If a
88 Visitor is sufficiently intelligent, pretty SQL may be emitted, but that is not
93 There are two major components to SQL::Abstract v2.
99 This is the Abstract Syntax Tree. It is a data structure that represents
100 everything necessary to construct the SQL statement in whatever dialect the
105 This object conforms to the Visitor pattern and is used to generate the SQL
106 represented by the AST. Each dialect will have a different Visitor object. In
107 addition, there will be visitors for at least one of the ANSI specifications.
111 The division of duties between the two components will focus on what the AST
112 can and cannot assume. For example, identifiers do not have 20 components in
113 any dialect, so the AST can validate that. However, determining what
114 constitutes a legal identifier can only be determined by the Visitor object
115 enforcing that dialect's rules.
119 The AST will be a HoHo..oH (hash of hash of ... of hashes). The keys to the
120 outermost hash will be the various clauses of a SQL statement, plus some
125 These are the additional metadata keys that the AST provides for.
129 This denotes what kind of query this AST should be interpreted as. Different
130 Visitors may accept additional values for type. For example, a MySQL Visitor
131 may choose to accept 'replace' for REPLACE INTO. If a type value is
132 unrecognized by the Visitor, the Visitor is expected to throw an error.
134 All Visitors are expected to handle the following values for type:
140 This is a SELECT statement.
144 This is an INSERT statement.
148 This is an UPDATE statement.
152 This is a DELETE statement.
158 This denotes the version of the AST. Different versions will indicate different
159 capabilities provided. Visitors will choose to respect the ast_version as needed
162 =head2 Structural units
164 All structural units will be hashes. These hashes will have, at minimum, the
171 This indicates the structural unit that this hash is representing. While this
172 specification provides for standard structural units, different Visitors may
173 choose to accept additional units as desired. If a Visitor encounters a unit it
174 doesn't know how to handle, it is expected to throw an exception.
178 Structural units in the AST are supported by loaded components. L<SQL::Abstract>
179 provides for the following structural units by default:
183 This is a (potentially) fully canonicalized identifier for a elemnt in the
184 query. This element could be a schema, table, or column. The Visitor will
185 determine validity within the context of that SQL dialect. The AST is only
186 responsible for validating that the elements are non-empty Strings.
188 The hash will be structured as follows:
191 type => 'Identifier',
197 If element3 exists, then element2 must exist. element1 must always exist. If a
198 given element exists, then it must be defined and of non-zero length.
200 Visitors are expected to, by default, quote all identifiers according to the SQL
201 dialect's quoting scheme.
203 Any of the elements may be '*', as in SELECT * or SELECT COUNT(*). Visitors must
204 be careful to I<not> quote asterisks.
208 A Value is a Perl scalar. Depending on the subtype, a Visitor may be able to
209 make certain decisions. The following are the minimally-valid subtypes:
215 A String is a quoted series of characters. The Visitor is expected to ensure
216 that embedded quotes are properly handled per the SQL dialect's quoting scheme.
220 A Number is an unquoted number in some numeric format.
224 Null is SQL's NULL and corresponds to Perl's C<undef>.
226 =item * BindParameter
228 This corresponds to a value that will be passed in. This value is normally
229 quoted in such a fashion so as to protect against SQL injection attacks. (q.v.
230 L<DBI/quote()> for an example.)
232 BindParameters are normally represented by a '?'.
236 The hash will be structured as follows:
240 subtype => [ 'String' | 'Number' | 'Null' | 'BindParameter' ]
244 The provided subtypes are the ones that all Visitors are expected to support.
245 Visitors may choose to support additional subtypes. Visitors are expected to
246 throw an exception upon encountering an unknown subtype.
250 An Operator would be, in SQL dialect terms, a unary operator, a binary operator,
251 a trinary operator, or a function. Since different dialects may have a given
252 functionality as an operator or a function (such as CONCAT in MySQl vs. || in
253 Oracle for string concatenation), they will be represented in the AST as generic
256 The hash will be structured as follows:
261 args => ExpressionList,
264 Operators have a cardinality, or expected number of arguments. Some operators,
265 such as MAX(), have a cardinality of 1. Others, such as IF(), have a cardinality
266 of N, meaning they can have any number of arguments greater than 0. Others, such
267 as NOW(), have a cardinality of 0. Several operators with the same meaning may
268 have a different cardinality in different SQL dialects as different engines may
269 allow different behaviors. As cardinality may differ between dialects, enforcing
270 cardinality is necessarily left to the Visitor.
272 Operators also have restrictions on the types of arguments they will accept. The
273 first argument may or may not restricted in the same fashion as the other
274 arguments. As with cardinality, this restriction will need to be managed by the
277 The operator name needs to take into account the possibility that the RDBMS may
278 allow UDFs (User-Defined Functions) that have the same name as an operator, such
279 as 'AND'. This will have to be managed by the Visitor.
283 A Subquery is another AST whose type metadata parameter is set to "SELECT".
285 Most places that a Subquery can be used would require a single value to be
286 returned (single column, single row), but that is not something that the AST can
287 easily enforce. The single-column restriction may possibly be enforced, but the
288 single-row restriction is much more difficult and, in most cases, probably
291 Subqueries, when expressed in SQL, must be bounded by parentheses.
295 An Expression can be any one of the following:
309 An Expression is a meta-syntactic unit. An "Expression" unit will never appear
310 within the AST. It acts as a junction.
312 =head3 ExpressionList
314 An ExpressionList is a list of Expressions, generally separated by commas
315 (though other separators may be appropriate at times or for different SQL
316 dialects). An null separator may also be used.
318 The hash for an ExpressionList is as follows:
321 type => 'ExpressionList',
323 elements => Array of Expressions,
326 An ExpressionList is always rendered in SQL with parentheses around it.
330 There is no specific operator or nodetype for nesting. Instead, nesting is
331 explicitly specified by node descent in the AST.
335 These are all the legal and acceptable clauses within the AST that would
336 correpsond to clauses in a SQL statement. Not all clauses are legal within a
337 given RDBMS engine's SQL dialect and some clauses may be required in one and
338 optional in another. Detecting and enforcing those engine-specific restrictions
339 is the responsibility of the Visitor object.
341 The clauses are defined with a yacc-like syntax. The various parts are:
347 This means "defined" and is used to create a new term to be used below.
351 This means optional and indicates that the items within it are optional.
355 This means optional and repeating as many times as desired.
359 This means alternation. It is a binary operator and indicates that either the
360 left or right hand sides may be used, but not both.
364 This is a grouping construct. It means that all elements within this construct
365 are treated together for the purposes of optional, repeating, alternation, etc.
369 The expected clauses are (name and structure):
373 This corresponds to the SELECT clause of a SELECT statement.
375 A select clause unit is an array of one or more SelectComponent units.
377 The hash for a SelectComponent unit is composed as follows:
380 type => 'SelectComponent',
385 The 'as' component is optional. Visitors may choose to make it required in
390 This is a list of tables that this clause is affecting. It corresponds to the
391 FROM clause in a SELECT statement and the INSERT INTO/UPDATE/DELETE clauses in
392 those respective statements. Depending on the type metadata entry, the
393 appropriate clause name will be used.
395 The tables clause has several RDBMS-specific variations. The AST will support
396 all of them and it is up to the Visitor object constructing the actual SQL to
397 validate and/or use what is provided as appropriate.
399 A TableJoin is a junction of the following elements:
403 =item * TableIdentifier
409 The hash for a TableIdentifier will be composed as follows:
413 type => 'TableIdentifier',
418 The value should be either an Identifier or a SubQuery.
420 The hash for an Operator within a tables clause will be composed as follows:
425 op => '< LEFT|RIGHT|FULL [ OUTER ] > | INNER | CROSS',
429 A USING clause is syntactic sugar for an ON clause and, as such, is not provided
430 for by the AST. A join of a comma is identical to a CROSS JOIN and, as such, is
431 not provided for by the AST. The on clause is optional.
435 This corresponds to the WHERE clause in a SELECT, UPDATE, or DELETE statement.
437 A where clause is composed of an Expression.
441 This corresponds to the SET clause in an INSERT or UPDATE statement.
443 A set clause unit is an array of one or more SetComponent units.
445 The hash for SetComponent unit is composed as follows:
448 type => 'SetComponent',
455 This corresponds to the optional list of columns in an INSERT statement.
457 A columns clause unit is an array of one or more Identifier units.
461 This corresponds to the VALUES clause in an INSERT statement.
463 A values clause unit is an array of one or more Expression units.
465 If there is a columns clause, the number of entries in the values clause must be
466 equal to the number of entries in the columns clause.
470 This corresponds to the ORDER BY clause in a SELECT statement.
472 A orderby clause unit is an array of one or more OrderbyComponent units.
474 The hash for a OrderbyComponent unit is composed as follows:
477 type => 'OrderbyComponent',
478 value => < Identifier | Number >
479 dir => '< ASC | DESC >',
482 The dir element, if omitted, will be defaulted to ASC by the AST. The number
483 corresponds to a column in the select clause.
487 This corresponds to the GROUP BY clause in a SELECT statement.
489 A groupby clause unit is an array of one or more GroupbyComponent units.
491 The hash for a GroupbyComponent unit is composed as follows:
494 type => 'GroupbyComponent',
495 value => < Identifier | Number >
498 The number corresponds to a column in the select clause.
502 This corresponds to the clause that is used in some RDBMS engines to limit the
503 number of rows returned by a query. In MySQL, this would be the LIMIT clause.
505 The hash for a rows clause is composed as follows:
512 The start attribute, if ommitted, will default to 0. The count attribute is
517 This corresponds to the clause that is used in some RDBMS engines to indicate
518 what locks are to be taken by this SELECT statement.
520 The hash for a for clause is composed as follows:
523 value => '< UPDATE | DELETE >',
528 This corresponds to the clause that is used in some RDBMS engines to provide for
529 an adjacency-list query.
531 The hash for a for clause is composed as follows:
534 start_with => ExpressionList,
536 option => '< PRIOR | NOCYCLE >'
537 cond => ExpressionList,
539 order_siblings => orderby-clause,
542 Both the start_with and order_siblings clauses are optional.
554 robkinyon: Rob Kinyon C<< <rkinyon@cpan.org> >>
558 You may distribute this code under the same terms as Perl itself.