19 commits

Author	SHA1	Message	Date
Folkert	4da01c720e	monomorphize the ext_var of records/tag unions	2020-03-12 21:32:58 +01:00
Folkert	3dbaac210a	add snapshot functions to Subs	2020-03-12 02:38:38 +01:00
Folkert	a037173cdb	complete the Content hashing	2020-03-11 23:00:44 +01:00
Folkert	5c9cf0ef37	add type hash function	2020-03-11 22:01:56 +01:00
Folkert	fb8e8570bc	Merge remote-tracking branch 'origin/trunk' into num	2020-03-11 14:30:50 +01:00
Richard Feldman	3ecbe0325c	Plain number literals (e.g. 5) are now Num *	2020-03-10 22:27:13 -04:00
Folkert	62f34c41cc	hardcode Bool alias ... previously it would use the tag union, which would result in slightly worse error messages down the line	2020-03-10 15:21:39 +01:00
Folkert	7fea957979	hardcode Bool	2020-03-10 14:50:35 +01:00
Richard Feldman	72ecb231f1	clip clop clip clop clippy	2020-03-07 23:11:42 -05:00
Richard Feldman	db4ef45708	Upgrade to List.set_in_place when appropriate	2020-03-07 12:43:24 -05:00
Folkert	bca43a2564	improve is_unique	2020-03-07 12:43:24 -05:00
Richard Feldman	a092f59b0f	First pass at is_unique	2020-03-07 12:43:24 -05:00
Folkert	94f2b6b6f6	switch back to Apply where possible	2020-03-07 00:50:38 +01:00
Folkert	eb7d9d98e6	hardcode variables for empty records and tag unions	2020-03-07 00:25:16 +01:00
Folkert	12a819e762	extract tags from the ext variable ... turns out this is always needed. I also removed the explicit check for an empty tag union: there could still be tags in the ext even if the union itself seems empty. seems weird to optimize for the empty case because it'll be very rare in practice	2020-03-06 13:45:59 +01:00
Richard Feldman	3b26a982f0	Move pretty_print_types into types crate	2020-03-06 02:01:05 -05:00
Richard Feldman	908e485fca	Move constrain and its deps into their own crates	2020-03-06 01:43:39 -05:00
Richard Feldman	8ef2ae3de7	clippy clip clip	2020-03-05 23:03:29 -05:00
Richard Feldman	3b6ed43126	Extract can/ into its own crate, plus its deps	2020-03-05 23:01:32 -05:00