Clean up string-related variants in Token and Value

Remove sqlparser_generic.rs/sqlparser_postgres.rs duplication

https://jakewheat.github.io/sql-overview/sql-2011-foundation-grammar.html#character-string-type

https://jakewheat.github.io/sql-overview/sql-2011-foundation-grammar.html#exact-numeric-type

Some unsupported features are noted as TODOs.

Widely used in MS SQL and specified in ANSI.

The dialect information is from https://en.wikibooks.org/wiki/SQL_Dialects_Reference/Data_struct...

By not swallowing the Err from parse_data_type().

Also switch to `match` to enable parsing tabl...

Don't need the duplicate `columns.push()` + we advance the tokenizer,
so no need to peek first.

...by reusing `parse_column_names` instead of extracting identifiers
out of the `parse_expr_list...

This happens all the time when I forget to check that the keyword I wanted
to use is actually li...

(the tests affected by "unboxing" in the previous commits.)

Instead change ASTNode::SQLSubquery to be Box<SQLSelect>

This used to be needed when it was a variant in the ASTNode enum itself.

It was probably copied from somewhere else when most types were variants
in ASTNode, and needed ...

ASTNode can now be renamed SQLExpression, as it represents a node in
the "expression" part of th...

Also move more things to use SQLIdent instead of String in the hope of
making it a newtype event...

Store (and parse) `table_name: SQLObjectName` instead of
`relation: Option<Box<ASTNode>>`, which...

...instead make `parse_compound_identifier()` return the underlying
Vec<> directly, and rename i...

...to match the name of the recently introduced `SQLObjectName` struct
and to avoid any reservat...

(To store "A name of a table, view, custom type, etc., possibly
multi-part, i.e. db.schema.obj"....

This makes the parser more strict when handling SELECTs nested
somewhere in the main statement:
...

I checked the docs of a few of the most popular RDBMSes, and it seems
there's consensus that the...

Parser::parse_sql() can now parse a semicolon-separated list of
statements, returning them in a ...

(The primary motivation was that it makes the tests more resilient to
the upcoming changes to th...

Before this change an expression like `(a+b)-(c+d)` was parsed correctly
(as a Minus node with t...

Before this commit there was a single `parse_expr(u8)` method, which
was called both

1) from wi...

Continuing from https://github.com/andygrove/sqlparser-rs/pull/33#issuecomment-453060427

This s...

This will allow re-using it for SQLStatement in a later commit.

(Also split the new struct into...

The SQLAssignment *struct* is used directly in ASTNode::SQLUpdate (will
change to SQLStatement::...

...as in `FROM foo bar WHERE bar.x > 1`.

To avoid ambiguity as to whether a token is an alias o...

A "table factor" (name borrowed from the ANSI SQL grammar) is a table
name or a derived table (s...

This part changes behavior:
- Fail when no identifier is found.
- Avoid rewinding if EOF was hit...

Now populating SQLWord.keyword based on the list of globally supported
keywords.

...instead of parse_keyword / consume_token - to reduce nesting of `if`s.

(Follow-up to PR #35)

Fold Token::{Keyword, Identifier, DoubleQuotedString} into one
Token::SQLWord, which has the nec...

i.e. ASC/DESC/unspecified - so that we don't lose information about
source code.

Also don't tak...

because they share implementation.

This reduces amount of boilerplate and avoids cloning the `expr` param.

These have identical copies in sqlparser_generic.rs

...as this syntax is not specific to the PostgreSQL dialect.

Also use verified() to assert that...

...and parser support for the corresponding token, as "..." in SQL[*] is
not a literal string li...

Its existence alongside SingleQuotedString simply doesn't make sense:
`'a string'` is a string l...

Indeed, given that there is Token::SingleQuotedString and
Token::Identifier, there's no other "s...

An identifier is not a literal value, and parse_value is not called on
such a token anyway.

Mainly by replacing `assert_eq!(sql, ast.to_string())` with a call to
the recently introduced `v...

Clean up consume_token() and parse/expect_keyword()

Add #[must_use] to warn against unchecked results of parse_keyword/s in
the future.

Before this missing keywords THEN/WHEN/AS would be parsed as if they
were in the text as the cod...

The function is invoked after a DoubleColon was already matched.

implement NOT

Miscellaneous fixes

...it gets handled just as well by the infix parser.
(Add a test while we're at it.)

1) Simplified the bit in parse_datatype()
2) Made sure it was covered by the test (the "public.y...

There's no Token::Period in such situation, so fractional part (from sec) was silently truncated...

Remove `pub` from the "internal" ones. Remove ones that duplicate each
other completely.