hnix/tests/ParserTests.hs

383 lines
14 KiB
Haskell
Raw Normal View History

2014-08-03 14:17:43 +02:00
{-# LANGUAGE TemplateHaskell #-}
{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE RankNTypes #-}
{-# OPTIONS_GHC -fno-warn-name-shadowing #-}
2014-08-03 14:17:43 +02:00
module ParserTests (tests) where
import Data.Fix
import Data.Semigroup
import Data.Text (Text, unpack)
import Nix.Atoms
import Nix.Expr
import Nix.Parser
import Test.Tasty
import Test.Tasty.HUnit
import Test.Tasty.TH
2014-08-03 14:17:43 +02:00
case_constant_int :: Assertion
case_constant_int = assertParseText "234" $ mkInt 234
case_constant_bool :: Assertion
case_constant_bool = do
assertParseText "true" $ mkBool True
assertParseText "false" $ mkBool False
2014-08-03 14:17:43 +02:00
case_constant_bool_respects_attributes :: Assertion
case_constant_bool_respects_attributes = do
assertParseText "true-foo" $ mkSym "true-foo"
assertParseText "false-bar" $ mkSym "false-bar"
2014-08-16 00:16:12 +02:00
case_constant_path :: Assertion
case_constant_path = do
assertParseText "./." $ mkPath False "./."
assertParseText "./+-_/cdef/09ad+-" $ mkPath False "./+-_/cdef/09ad+-"
assertParseText "/abc" $ mkPath False "/abc"
assertParseText "../abc" $ mkPath False "../abc"
assertParseText "<abc>" $ mkPath True "abc"
assertParseText "<../cdef>" $ mkPath True "../cdef"
assertParseText "a//b" $ mkOper2 NUpdate (mkSym "a") (mkSym "b")
assertParseText "rec+def/cdef" $ mkPath False "rec+def/cdef"
assertParseText "a/b//c/def//<g> < def/d" $ mkOper2 NLt
(mkOper2 NUpdate (mkPath False "a/b") $ mkOper2 NUpdate
(mkPath False "c/def") (mkPath True "g"))
(mkPath False "def/d")
assertParseText "a'b/c" $ Fix $ NBinary NApp (mkSym "a'b") (mkPath False "/c")
assertParseText "a/b" $ mkPath False "a/b"
assertParseText "4/2" $ mkPath False "4/2"
2014-08-16 00:16:12 +02:00
assertParseFail "."
assertParseFail ".."
assertParseFail "/"
assertParseFail "a/"
2014-08-30 10:10:28 +02:00
assertParseFail "a/def/"
case_constant_uri :: Assertion
case_constant_uri = do
assertParseText "a:a" $ mkUri "a:a"
assertParseText "http://foo.bar" $ mkUri "http://foo.bar"
assertParseText "a+de+.adA+-:%%%ads%5asdk&/" $ mkUri "a+de+.adA+-:%%%ads%5asdk&/"
assertParseText "rec+def:c" $ mkUri "rec+def:c"
assertParseText "f.foo:bar" $ mkUri "f.foo:bar"
2014-08-30 10:10:28 +02:00
assertParseFail "http://foo${\"bar\"}"
assertParseFail ":bcdef"
assertParseFail "a%20:asda"
assertParseFail ".:adasd"
assertParseFail "+:acdcd"
2014-08-16 00:16:12 +02:00
2014-08-03 14:17:43 +02:00
case_simple_set :: Assertion
case_simple_set = do
assertParseText "{ a = 23; b = 4; }" $ Fix $ NSet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar (mkSelector "a") $ mkInt 23
, NamedVar (mkSelector "b") $ mkInt 4
]
assertParseFail "{ a = 23 }"
case_set_inherit :: Assertion
case_set_inherit = do
assertParseText "{ e = 3; inherit a b; }" $ Fix $ NSet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar (mkSelector "e") $ mkInt 3
, Inherit Nothing $ flip StaticKey Nothing <$> ["a", "b"]
]
assertParseText "{ inherit; }" $ Fix $ NSet [ Inherit Nothing [] ]
case_set_scoped_inherit :: Assertion
case_set_scoped_inherit = assertParseText "{ inherit (a) b c; e = 4; inherit(a)b c; }" $ Fix $ NSet
[ Inherit (Just (mkSym "a")) $ flip StaticKey Nothing <$> ["b", "c"]
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
, NamedVar (mkSelector "e") $ mkInt 4
, Inherit (Just (mkSym "a")) $ flip StaticKey Nothing <$> ["b", "c"]
]
case_set_rec :: Assertion
case_set_rec = assertParseText "rec { a = 3; b = a; }" $ Fix $ NRecSet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar (mkSelector "a") $ mkInt 3
, NamedVar (mkSelector "b") $ mkSym "a"
]
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
case_set_complex_keynames :: Assertion
case_set_complex_keynames = do
assertParseText "{ \"\" = null; }" $ Fix $ NSet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar [DynamicKey (Plain "")] mkNull ]
assertParseText "{ a.b = 3; a.c = 4; }" $ Fix $ NSet
[ NamedVar [StaticKey "a" Nothing, StaticKey "b" Nothing] $ mkInt 3
, NamedVar [StaticKey "a" Nothing, StaticKey "c" Nothing] $ mkInt 4
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
]
assertParseText "{ ${let a = \"b\"; in a} = 4; }" $ Fix $ NSet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar [DynamicKey (Antiquoted letExpr)] $ mkInt 4 ]
assertParseText "{ \"a${let a = \"b\"; in a}c\".e = 4; }" $ Fix $ NSet
[ NamedVar [DynamicKey (Plain str), StaticKey "e" Nothing] $ mkInt 4 ]
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
where
letExpr = Fix $ NLet [ NamedVar (mkSelector "a") (mkStr "b") ] (mkSym "a")
str = DoubleQuoted [Plain "a", Antiquoted letExpr, Plain "c"]
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
case_set_inherit_direct :: Assertion
case_set_inherit_direct = assertParseText "{ inherit ({a = 3;}); }" $ Fix $ NSet
[ flip Inherit [] $ Just $ Fix $ NSet [NamedVar (mkSelector "a") $ mkInt 3]
2014-08-03 14:17:43 +02:00
]
2014-08-15 23:30:24 +02:00
case_inherit_selector :: Assertion
case_inherit_selector = do
assertParseText "{ inherit \"a\"; }" $ Fix $ NSet
[Inherit Nothing [DynamicKey (Plain "a")]]
2014-08-15 23:30:24 +02:00
assertParseFail "{ inherit a.x; }"
2014-08-03 14:17:43 +02:00
case_int_list :: Assertion
case_int_list = assertParseText "[1 2 3]" $ Fix $ NList
[ mkInt i | i <- [1,2,3] ]
2014-08-03 14:17:43 +02:00
case_int_null_list :: Assertion
case_int_null_list = assertParseText "[1 2 3 null 4]" $ Fix (NList (map (Fix . NConstant) [NInt 1, NInt 2, NInt 3, NNull, NInt 4]))
2014-08-03 14:17:43 +02:00
2014-08-15 23:30:24 +02:00
case_mixed_list :: Assertion
case_mixed_list = do
assertParseText "[{a = 3;}.a (if true then null else false) null false 4 [] c.d or null]" $ Fix $ NList
[ Fix (NSelect (Fix (NSet [NamedVar (mkSelector "a") (mkInt 3)])) (mkSelector "a") Nothing)
2014-08-15 23:30:24 +02:00
, Fix (NIf (mkBool True) mkNull (mkBool False))
, mkNull, mkBool False, mkInt 4, Fix (NList [])
, Fix (NSelect (mkSym "c") (mkSelector "d") (Just mkNull))
]
assertParseFail "[if true then null else null]"
assertParseFail "[a ? b]"
assertParseFail "[a : a]"
assertParseFail "[${\"test\")]"
2014-08-03 14:17:43 +02:00
case_simple_lambda :: Assertion
case_simple_lambda = assertParseText "a: a" $ Fix $ NAbs (Param "a") (mkSym "a")
2014-08-03 14:17:43 +02:00
case_lambda_or_uri :: Assertion
case_lambda_or_uri = do
assertParseText "a :b" $ Fix $ NAbs (Param "a") (mkSym "b")
assertParseText "a c:def" $ Fix $ NBinary NApp (mkSym "a") (mkUri "c:def")
assertParseText "c:def: c" $ Fix $ NBinary NApp (mkUri "c:def:") (mkSym "c")
assertParseText "a:{}" $ Fix $ NAbs (Param "a") $ Fix $ NSet []
assertParseText "a:[a]" $ Fix $ NAbs (Param "a") $ Fix $ NList [mkSym "a"]
assertParseFail "def:"
case_lambda_pattern :: Assertion
case_lambda_pattern = do
assertParseText "{b, c ? 1}: b" $
Fix $ NAbs (fixed args Nothing) (mkSym "b")
assertParseText "{ b ? x: x }: b" $
Fix $ NAbs (fixed args2 Nothing) (mkSym "b")
assertParseText "a@{b,c ? 1}: b" $
Fix $ NAbs (fixed args (Just "a")) (mkSym "b")
assertParseText "{b,c?1}@a: c" $
Fix $ NAbs (fixed args (Just "a")) (mkSym "c")
assertParseText "{b,c?1,...}@a: c" $
Fix $ NAbs (variadic vargs (Just "a")) (mkSym "c")
assertParseText "{...}: 1" $
Fix $ NAbs (variadic mempty Nothing) (mkInt 1)
assertParseFail "a@b: a"
assertParseFail "{a}@{b}: a"
where
2018-04-02 05:53:19 +02:00
fixed args = ParamSet args False
variadic args = ParamSet args True
args = [("b", Nothing), ("c", Just $ mkInt 1)]
vargs = [("b", Nothing), ("c", Just $ mkInt 1)]
args2 = [("b", Just lam)]
lam = Fix $ NAbs (Param "x") (mkSym "x")
2014-08-03 14:17:43 +02:00
case_lambda_app_int :: Assertion
case_lambda_app_int = assertParseText "(a: a) 3" $ Fix (NBinary NApp lam int) where
int = mkInt 3
lam = Fix (NAbs (Param "a") asym)
asym = mkSym "a"
2014-08-03 14:17:43 +02:00
case_simple_let :: Assertion
case_simple_let = do
assertParseText "let a = 4; in a" $ Fix (NLet binds $ mkSym "a")
assertParseFail "let a = 4 in a"
where
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
binds = [NamedVar (mkSelector "a") $ mkInt 4]
2014-08-03 14:17:43 +02:00
case_let_body :: Assertion
case_let_body = assertParseText "let { body = 1; }" letBody
where
letBody = Fix $ NSelect aset (mkSelector "body") Nothing
aset = Fix $ NRecSet [NamedVar (mkSelector "body") (mkInt 1)]
case_nested_let :: Assertion
case_nested_let = do
assertParseText "let a = 4; in let b = 5; in a" $ Fix $ NLet
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ NamedVar (mkSelector "a") $ mkInt 4 ]
(Fix $ NLet [NamedVar (mkSelector "b") $ mkInt 5] $ mkSym "a")
assertParseFail "let a = 4; let b = 3; in b"
case_let_scoped_inherit :: Assertion
case_let_scoped_inherit = do
assertParseText "let a = null; inherit (b) c; in c" $ Fix $ NLet
[ NamedVar (mkSelector "a") mkNull
, Inherit (Just $ mkSym "b") [StaticKey "c" Nothing] ]
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
(mkSym "c")
assertParseFail "let inherit (b) c in c"
2014-08-30 10:10:28 +02:00
case_if :: Assertion
case_if = do
assertParseText "if true then true else false" $ Fix $ NIf (mkBool True) (mkBool True) (mkBool False)
2014-08-30 10:10:28 +02:00
assertParseFail "if true then false"
assertParseFail "else"
assertParseFail "if true then false else"
assertParseFail "if true then false else false else"
assertParseFail "1 + 2 then"
case_identifier_special_chars :: Assertion
case_identifier_special_chars = do
assertParseText "_a" $ mkSym "_a"
assertParseText "a_b" $ mkSym "a_b"
assertParseText "a'b" $ mkSym "a'b"
assertParseText "a''b" $ mkSym "a''b"
assertParseText "a-b" $ mkSym "a-b"
assertParseText "a--b" $ mkSym "a--b"
assertParseText "a12a" $ mkSym "a12a"
assertParseFail ".a"
assertParseFail "'a"
case_identifier_keyword_prefix :: Assertion
case_identifier_keyword_prefix = do
assertParseText "true-name" $ mkSym "true-name"
assertParseText "trueName" $ mkSym "trueName"
assertParseText "null-name" $ mkSym "null-name"
assertParseText "nullName" $ mkSym "nullName"
assertParseText "[ null-name ]" $ mkList [ mkSym "null-name" ]
makeTextParseTest :: Text -> Assertion
makeTextParseTest str = assertParseText ("\"" <> str <> "\"") $ mkStr str
case_simple_string :: Assertion
case_simple_string = mapM_ makeTextParseTest ["abcdef", "a", "A", " a a ", ""]
case_string_dollar :: Assertion
case_string_dollar = mapM_ makeTextParseTest ["a$b", "a$$b", "$cdef", "gh$i"]
case_string_escape :: Assertion
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
case_string_escape = do
assertParseText "\"\\$\\n\\t\\r\\\\\"" $ mkStr "$\n\t\r\\"
assertParseText "\" \\\" \\' \"" $ mkStr " \" ' "
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
case_string_antiquote :: Assertion
case_string_antiquote = do
assertParseText "\"abc${ if true then \"def\" else \"abc\" } g\"" $
Fix $ NStr $ DoubleQuoted
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
[ Plain "abc"
, Antiquoted $ Fix $ NIf (mkBool True) (mkStr "def") (mkStr "abc")
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
, Plain " g"
]
assertParseText "\"\\${a}\"" $ mkStr "${a}"
antiquotes + improve pretty printer + restructure This commit improves the pretty printing and adds support for antiquotes. It also fixes an issue with the parser that caused `[if true then false else true]` to parse successfully, even though that is not a valid nix expression. The pretty printer now produces a lot more readable output and also supports operator precedences. The changes to the AST are: * strings are no longer atomic, because they may contain other expressions in the form of antiquotes. For strings, the new type NString is introduced and the constructor NStr is added to NExprF * the unused NVar constructor of NExprF is removed * operators are now represented uniformly so that the pretty printer can lookup information about operators (in particular, associativity and precedence) * the NArgs constructor is removed. The first argument of the NAbs constructor now directly represents the lambda arguments. * the select and the hasattr operator are moved into NExpr because they are special (they only accept a selector as second argument, and select also supports 'or') The list of operators is now in Types.hs and Parser.hs re-uses that list to build the parser. This is required because the pretty printer and parser both need access to operator precedences. Parser and evaluator also support dynamic attributes and attributes with dots now. As an example, `let b.a = 3; b.c = { e = {}; }; b.c.e.${"f"} = 4; in b` is parsed and evaluated correctly. As a side effect, NSym values now don't evaluate to themselves anymore, but instead to the value retrieved by looking up the variable in the current environment. Support for evaluating `inherit` bindings was removed because it was broken before (`{ inherit a; }` would evaluate to a set where the attribute `a` had the value `NSym a`, not the value of `a`). The manual Show instances for the AST were replaced by derived ones, because the manual instances often resulted in output were it was difficult to determine the missing parentheses.
2014-08-15 22:11:54 +02:00
assertParseFail "\"a"
assertParseFail "${true}"
assertParseFail "\"${true\""
2014-08-16 01:31:41 +02:00
case_select :: Assertion
case_select = do
assertParseText "a . e .di. f" $ Fix $ NSelect (mkSym "a")
[ StaticKey "e" Nothing, StaticKey "di" Nothing, StaticKey "f" Nothing ]
2014-08-16 01:31:41 +02:00
Nothing
assertParseText "a.e . d or null" $ Fix $ NSelect (mkSym "a")
[ StaticKey "e" Nothing, StaticKey "d" Nothing ]
2014-08-16 01:31:41 +02:00
(Just mkNull)
assertParseText "{}.\"\"or null" $ Fix $ NSelect (Fix (NSet []))
2014-08-16 01:31:41 +02:00
[ DynamicKey (Plain "") ] (Just mkNull)
2018-04-10 20:22:54 +02:00
assertParseText "{ a = [1]; }.a or [2] ++ [3]" $ Fix $ NBinary NConcat
(Fix (NSelect
(Fix (NSet [NamedVar [StaticKey "a" Nothing]
(Fix (NList [Fix (NConstant (NInt 1))]))]))
[StaticKey "a" Nothing]
(Just (Fix (NList [Fix (NConstant (NInt 2))])))))
(Fix (NList [Fix (NConstant (NInt 3))]))
2014-08-16 01:31:41 +02:00
2014-08-16 01:17:01 +02:00
case_select_path :: Assertion
case_select_path = do
assertParseText "f ./." $ Fix $ NBinary NApp (mkSym "f") (mkPath False "./.")
assertParseText "f.b ../a" $ Fix $ NBinary NApp select (mkPath False "../a")
assertParseText "{}./def" $ Fix $ NBinary NApp (Fix (NSet [])) (mkPath False "./def")
assertParseText "{}.\"\"./def" $ Fix $ NBinary NApp
(Fix $ NSelect (Fix (NSet [])) [DynamicKey (Plain "")] Nothing)
2014-08-30 10:10:28 +02:00
(mkPath False "./def")
2014-08-16 01:17:01 +02:00
where select = Fix $ NSelect (mkSym "f") (mkSelector "b") Nothing
2014-08-16 00:16:12 +02:00
case_fun_app :: Assertion
case_fun_app = do
assertParseText "f a b" $ Fix $ NBinary NApp (Fix $ NBinary NApp (mkSym "f") (mkSym "a")) (mkSym "b")
assertParseText "f a.x or null" $ Fix $ NBinary NApp (mkSym "f") $ Fix $
NSelect (mkSym "a") (mkSelector "x") (Just mkNull)
assertParseFail "f if true then null else null"
case_indented_string :: Assertion
case_indented_string = do
assertParseText "''a''" $ mkIndentedStr 0 "a"
assertParseText "''\n foo\n bar''" $ mkIndentedStr 2 "foo\nbar"
assertParseText "'' ''" $ mkIndentedStr 0 ""
assertParseText "'''''''" $ mkIndentedStr 0 "''"
assertParseText "'' ${null}\n a${null}''" $ Fix $ NStr $ Indented 3
[ Antiquoted mkNull
, Plain "\na"
, Antiquoted mkNull
]
assertParseFail "'''''"
assertParseFail "'' '"
case_indented_string_escape :: Assertion
case_indented_string_escape = assertParseText
"'' ''\\n ''\\t ''\\\\ ''${ \\ \\n ' ''' ''" $
mkIndentedStr 1 "\n \t \\ ${ \\ \\n ' '' "
case_operator_fun_app :: Assertion
case_operator_fun_app = do
assertParseText "a ++ b" $ mkOper2 NConcat (mkSym "a") (mkSym "b")
assertParseText "a ++ f b" $ mkOper2 NConcat (mkSym "a") $ Fix $ NBinary NApp
(mkSym "f") (mkSym "b")
case_operators :: Assertion
case_operators = do
assertParseText "1 + 2 - 3" $ mkOper2 NMinus
(mkOper2 NPlus (mkInt 1) (mkInt 2)) (mkInt 3)
assertParseFail "1 + if true then 1 else 2"
assertParseText "1 + (if true then 2 else 3)" $ mkOper2 NPlus (mkInt 1) $ Fix $ NIf
(mkBool True) (mkInt 2) (mkInt 3)
assertParseText "{ a = 3; } // rec { b = 4; }" $ mkOper2 NUpdate
(Fix $ NSet [NamedVar (mkSelector "a") (mkInt 3)])
(Fix $ NRecSet [NamedVar (mkSelector "b") (mkInt 4)])
assertParseText "--a" $ mkOper NNeg $ mkOper NNeg $ mkSym "a"
assertParseText "a - b - c" $ mkOper2 NMinus
(mkOper2 NMinus (mkSym "a") (mkSym "b")) $
mkSym "c"
assertParseText "foo<bar" $ mkOper2 NLt (mkSym "foo") (mkSym "bar")
2014-08-30 10:10:28 +02:00
assertParseFail "+ 3"
assertParseFail "foo +"
case_comments :: Assertion
case_comments = do
Success expected <- parseNixFile "data/let.nix"
assertParseFile "let-comments-multiline.nix" expected
assertParseFile "let-comments.nix" expected
2014-08-03 14:17:43 +02:00
tests :: TestTree
tests = $testGroupGenerator
--------------------------------------------------------------------------------
assertParseText :: Text -> NExpr -> Assertion
assertParseText str expected = case parseNixText str of
Success actual ->
assertEqual ("When parsing " ++ unpack str)
(stripPositionInfo expected) (stripPositionInfo actual)
Failure err ->
assertFailure $ "Unexpected error parsing `" ++ unpack str ++ "':\n" ++ show err
assertParseFile :: FilePath -> NExpr -> Assertion
assertParseFile file expected = do
res <- parseNixFile $ "data/" ++ file
case res of
Success actual -> assertEqual ("Parsing data file " ++ file)
(stripPositionInfo expected) (stripPositionInfo actual)
Failure err ->
assertFailure $ "Unexpected error parsing data file `"
++ file ++ "':\n" ++ show err
assertParseFail :: Text -> Assertion
assertParseFail str = case parseNixText str of
Failure _ -> return ()
Success r ->
assertFailure $ "Unexpected success parsing `"
++ unpack str ++ ":\nParsed value: " ++ show r