Support table aliases without AS (7/8)

nickolay · nickolay · commit 76ec175d2067 · 2019-01-31T03:57:17.000+03:00
...as in `FROM foo bar WHERE bar.x &gt; 1`.

To avoid ambiguity as to whether a token is an alias or a keyword, we
maintain a blacklist of keywords, that can follow a "table factor", to
prevent parsing them as an alias. This "context-specific reserved
keyword" approach lets us accept more SQL that's valid in some dialects,
than a list of globally reserved keywords. Also some dialects (e.g.
Oracle) apparently don't reserve some keywords (like JOIN), while
presumably they won't accept them as an alias (`FROM foo JOIN` meaning
`FROM foo AS JOIN`).
diff --git a/src/dialect/keywords.rs b/src/dialect/keywords.rs
@@ -1,14 +1,16 @@
-///! This module defines a list of constants for every keyword that
+///! This module defines
+/// 1) a list of constants for every keyword that
 /// can appear in SQLWord::keyword:
 ///    pub const KEYWORD = "KEYWORD"
-/// and an `ALL_KEYWORDS` array with every keyword in it.
+/// 2) an `ALL_KEYWORDS` array with every keyword in it
+///     This is not a list of *reserved* keywords: some of these can be
+///     parsed as identifiers if the parser decides so. This means that
+///     new keywords can be added here without affecting the parse result.
 ///
-/// This is not a list of *reserved* keywords: some of these can be
-/// parsed as identifiers if the parser decides so. This means that
-/// new keywords can be added here without affecting the parse result.
-///
-/// As a matter of fact, most of these keywords are not used at all
-/// and could be removed.
+///     As a matter of fact, most of these keywords are not used at all
+///     and could be removed.
+/// 3) a `RESERVED_FOR_TABLE_ALIAS` array with keywords reserved in a
+/// "table alias" context.
 
 macro_rules! keyword {
     ($($ident:ident),*) => {
@@ -707,3 +709,12 @@ pub const ALL_KEYWORDS: &'static [&'static str] = &[
     ZONE,
     END_EXEC,
 ];
+
+/// These keywords can't be used as a table alias, so that `FROM table_name alias`
+/// can be parsed unambiguously without looking ahead.
+pub const RESERVED_FOR_TABLE_ALIAS: &'static [&'static str] = &[
+    WHERE, GROUP, ON, // keyword is 'reserved' in most dialects
+    JOIN, INNER, CROSS, FULL, LEFT, RIGHT, // not reserved in Oracle
+    NATURAL, USING, // not reserved in Oracle & MSSQL
+    // UNION, EXCEPT, INTERSECT, ORDER // TODO add these with tests.
+];
diff --git a/src/sqlparser.rs b/src/sqlparser.rs
@@ -14,6 +14,7 @@
 
 //! SQL Parser
 
+use super::dialect::keywords;
 use super::dialect::Dialect;
 use super::sqlast::*;
 use super::sqltokenizer::*;
@@ -950,13 +951,18 @@ impl Parser {
     /// `SELECT ... FROM t1 foo, t2 bar`, `SELECT ... FROM (...) AS bar`
     pub fn parse_optional_alias(
         &mut self,
+        reserved_kwds: &[&str],
     ) -> Result<Option<SQLIdent>, ParserError> {
         let after_as = self.parse_keyword("AS");
         let maybe_alias = self.next_token();
         match maybe_alias {
             // Accept any identifier after `AS` (though many dialects have restrictions on
-            // keywords that may appear here).
-            Some(Token::SQLWord(ref w)) if after_as  =>
+            // keywords that may appear here). If there's no `AS`: don't parse keywords,
+            // which may start a construct allowed in this position, to be parsed as aliases.
+            // (For example, in `FROM t1 JOIN` the `JOIN` will always be parsed as a keyword,
+            // not an alias.)
+            Some(Token::SQLWord(ref w))
+                if after_as || !reserved_kwds.contains(&w.keyword.as_str()) =>
             {
                 // have to clone here until #![feature(bind_by_move_pattern_guards)] is enabled by default
                 Ok(Some(w.value.clone()))
@@ -1157,7 +1163,7 @@ impl Parser {
         } else {
             self.parse_compound_identifier(&Token::Period)?
         };
-        let alias = self.parse_optional_alias()?;
+        let alias = self.parse_optional_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?;
         Ok(ASTNode::TableFactor {
             relation: Box::new(relation),
             alias,
diff --git a/tests/sqlparser_generic.rs b/tests/sqlparser_generic.rs
@@ -601,6 +601,10 @@ fn parse_joins_on() {
             JoinOperator::Inner
         )]
     );
+    parses_to(
+        "SELECT * FROM t1 JOIN t2 foo ON c1 = c2",
+        "SELECT * FROM t1 JOIN t2 AS foo ON c1 = c2",
+    );
     // Test parsing of different join operators
     assert_eq!(
         joins_from(verified("SELECT * FROM t1 JOIN t2 ON c1 = c2")),
@@ -644,6 +648,10 @@ fn parse_joins_using() {
             JoinOperator::Inner
         )]
     );
+    parses_to(
+        "SELECT * FROM t1 JOIN t2 foo USING(c1)",
+        "SELECT * FROM t1 JOIN t2 AS foo USING(c1)",
+    );
     // Test parsing of different join operators
     assert_eq!(
         joins_from(verified("SELECT * FROM t1 JOIN t2 USING(c1)")),
