Smalltalk-style syntax - Revision history

Kevin Reid: /* Verb mapping */ expand verb mapping table of examples per new thinking

2008-02-19T05:12:48Z

Verb mapping: expand verb mapping table of examples per new thinking

New page

[[User:Kevin Reid]]'s idle speculation on an [[alternate surface syntax]] for E.

Translated [[mint example]]:

makeMint does {
run: name = the any {
[sealer, unsealer] = makeBrandPair run: name.
mint does {
__printOn: out = the void {
out print: `<$name's mint>`.
}

makePurse: var balance an (int >= 0) = the any {
decr does {
run: amount an (0..balance) = the void {
set balance = balance - amount.
}
}.
purse does {
__printOn: out = the void {
out print: `<has $balance $name bucks>`.
}
getBalance = the int { balance }
sprout = the any { mint makePurse: 0 }
getDecr = the any { sealer seal: decr }

deposit: amount an int from: src = the void {
(unsealer unseal: src getDecr) run: amount.
set balance = balance + amount.
}
}.
^ purse
}
}.
^ mint
}
}

==Verb mapping==

{|
! Syntax
! Verb
! Arity
|-
| recip foo
| "foo"
| 0
|-
| recip foo: arg
| "foo"
| 1
|-
| recip foo: arg bar: arg
| "foo:bar"
| 2
|-
| recip foo: arg : arg
| "foo"
| 2
|-
| recip foo: arg bar: arg baz: arg
| "foo:bar:baz"
| 3
|-
| recip foo: arg bar: arg : arg
| "foo:bar"
| 3
|}

The traditional Smalltalk/ObjC form would include an additional colon at the end of the >0-arg verbs, but this would drop it for compatibility with normal E protocols.

There would also need to be a way to use arbitrary verbs as normal [[E surface syntax]] provides (the "[[verb-string]]" syntax).

[[Category:Syntax]]

Kevin Reid: verb mapping notes

2007-10-24T02:03:34Z

verb mapping notes

←Older revision		Revision as of 02:03, 24 October 2007
Line 36:		Line 36:
	}		}
	}		}
		+
		+	==Verb mapping==
		+
		+	{\|
		+	! Syntax
		+	! Verb
		+	! Arity
		+	\|-
		+	\| recip verb
		+	\| "verb"
		+	\| 0
		+	\|-
		+	\| recip verb: arg
		+	\| "verb"
		+	\| 1
		+	\|-
		+	\| recip verb: arg and: arg
		+	\| "verb:and"
		+	\| 2
		+	\|}
		+
		+	The traditional Smalltalk/ObjC form would include the colon in the 1-arg verb, but this would drop it for compatibility with normal E protocols.
		+
		+	There would also need to be a way to use arbitrary verbs as normal [[E surface syntax]] provides (the "[[verb-string]]" syntax).

	[[Category:Syntax]]		[[Category:Syntax]]

Kevin Reid: publishing old idea

2007-10-24T00:21:04Z

publishing old idea

←Older revision		Revision as of 00:21, 24 October 2007
Line 36:		Line 36:
	}		}
	}		}
-
-	~~==Verb mapping==~~
-
-	{\|
-	~~! Syntax~~
-	~~! Verb~~
-	~~! Arity~~
-	\|-
-	~~\| recip foo~~
-	~~\| "foo"~~
-	~~\| 0~~
-	\|-
-	~~\| recip foo: arg~~
-	~~\| "foo"~~
-	~~\| 1~~
-	\|-
-	~~\| recip foo: arg bar: arg~~
-	~~\| "foo:bar"~~
-	~~\| 2~~
-	\|-
-	~~\| recip foo: arg : arg~~
-	~~\| "foo"~~
-	~~\| 2~~
-	\|-
-	~~\| recip foo: arg bar: arg baz: arg~~
-	~~\| "foo:bar:baz"~~
-	~~\| 3~~
-	\|-
-	~~\| recip foo: arg bar: arg : arg~~
-	~~\| "foo:bar"~~
-	~~\| 3~~
-	\|}
-
-	~~The traditional Smalltalk/ObjC form would include an additional colon at the end of the >0-arg verbs, but this would drop it for compatibility with normal E protocols.~~
-
-	~~There would also need to be a way to use arbitrary verbs as normal [[E surface syntax]] provides (the "[[verb-string]]" syntax).~~

	[[Category:Syntax]]		[[Category:Syntax]]