dunhamsteve.bsky.social
Newt - a dependent typed programming language
todo list updates
+25
-22
+25
-22
TODO.md
+25
-22
TODO.md
···
3
3
4
4
- [ ] Use looping for TCO
5
5
- For single functions at least - I think this would be a performance win. I've learned that the slowness on `bun` goes away if I drop the TCO transform.
6
+
- Doing this manually for `lookupT23` got 3% speedup.
6
7
- [ ] Importing Prelude twice should be an error (currently it causes double hints and breaks auto)
7
-
- [ ] For errors in other files, point to the import
8
+
- [x] For errors in other files, point to the import
8
9
- [x] Unsolved metas should be errors (user metas are fine)
9
10
- [x] Better syntax for forward declared data (so we can distinguish from functions)
10
11
- [ ] maybe allow "Main" module name for any file
···
14
15
- [x] Case split for LSP mode
15
16
- [x] Require lowercase pattern variables
16
17
- I accidentally misspell a constructor and end up with a wildcard.
17
-
- [ ] Leverage LSP code for web playground
18
+
- [x] Leverage LSP code for web playground
18
19
- [ ] Improve handling of names:
19
20
- We need FC on names in a lot of places
20
21
- [x] FC for duplicate `data`, `record`, `class` name is wrong (points to `data`)
···
46
47
- Or the name on the VVar?
47
48
- Also we should render %pi, etc more readably (bonus points for _×_)
48
49
- Editor support:
49
-
- [ ] add missing cases should skip indented lines
50
-
- [ ] add missing cases should handle `_::_`
51
-
- [ ] add missing cases should share code between vscode and playground
50
+
- [x] add missing cases should skip indented lines
51
+
- [x] add missing cases should handle `_::_`
52
+
- [x] add missing cases should share code between vscode and playground
52
53
- [x] "Not in scope" should offer to import
53
-
- [ ] Case split
54
+
- [x] Case split
54
55
- [ ] Delay checking
55
56
- We have things like `foldr (\ x acc => case x : ...`, where the lambda doesn't have a good type, so we have to be explicit. If we could defer the checking of that expression until more things are solved, we might not need the annotation (e.g. checking the other arguments). Some `case` statements may have a similar situation.
56
57
- One idea is to throw the checks onto some sort of TODO list and run whatever works. (I think Idris may have a heuristic where it checks arguments backwards in some cases.)
57
58
- [x] Dependent records (I guess I missed that bit)
58
-
- [ ] Arguments on records
59
+
- [x] Arguments on records
59
60
- [ ] Add sugar for type aliases (maybe infer arguments)
60
61
- Lean has this, we maybe could run infer on the RHS and call it a day? We would need a simple LHS, though.
61
62
- [ ] see if we can get a better error if `for` is used instead of `for_` in do blocks
···
89
90
- [ ] Look into using holes for errors (https://types.pl/@AndrasKovacs/115401455046442009)
90
91
- This would let us hit more cases in a function when we hit an error.
91
92
- I've been wanting to try holes for parse errors too.
93
+
- Does softening up check errors break `auto`?
92
94
- [ ] Missing `∀ k` in type is error -> no declaration for, if we insert a hole, we can get the declaration.
93
95
- [ ] in-scope type at point in vscode
94
96
- So the idea here is that the references will be via FC, we remember the type at declaration and then point the usage back to the declaration (FC -> FC). We could dump all of this. (If we're still doing json.)
95
97
- This information _could_ support renaming, too (but there may be indentation issues).
96
98
- Do we want to (maybe later) keep the scope as a FC? We could do scope at point then.
97
-
- But ideally we'd switch to a server/repl, so we don't have to mess around with serializing stuff.
98
99
- [ ] LSP and/or more editor support
99
100
- [ ] refactor to query based? E.g. importing a module
100
-
- [ ] would be nice to have "add missing cases" and "case split"
101
+
- [x] would be nice to have "add missing cases" and "case split"
101
102
- [x] Probably need ranges for FC
102
-
- [ ] leave an interactive process running
103
+
- [x] leave an interactive process running
103
104
- [ ] restart mid file (we could save state per top level decl)
104
105
- [ ] rename in editor (need to accumulate all names and what they reference)
105
106
- [ ] who calls X? We can only do this scoped to the current context for now. Someday whole source dir. #lsp
106
107
- [ ] Pretty print
107
108
- Can we format code? Maybe pull nearby comments or attach them like FC to tokens?
108
109
- We would need to address stack and laziness issues in prettier printer (or make it merely pretty)
109
-
- [ ] Look into descriptions, etc.
110
-
- Can generating descriptions help with automatic "show" implementations
111
-
- We lost debug printing when switching to numeric tags
112
-
- Where did I see the idea of generating descriptions for inductive types?
113
110
- [ ] Add info to Ref/VRef (is dcon, arity, etc)
114
111
- To save lookups during compilation and it might make eval faster
115
112
- [x] number tags for data constructors
···
123
120
- [ ] Maybe add qualified names to surface syntax and allow / detect conflicts on reference
124
121
- [ ] Add `export` keywords
125
122
- [x] vscode - run newt when switching editors
126
-
- [ ] case split
123
+
- [x] case split
127
124
- We could fake this up:
128
125
- given a name and a point in the editor
129
126
- walk through the function looking for the binder
···
131
128
- enumerate valid constructors (and their arity)
132
129
- Repeat the line with each, applied to args
133
130
- For `<-` or `let` we'd want to fudge some `|` lines
131
+
- [ ] `let` has issues for multiline split / add missing
132
+
- [ ] `derive` has phantom splits in it
134
133
- [x] Support "Add missing cases"
135
134
- This has been hakced together
135
+
- LSP version now.
136
136
- We could possibly fake up missing cases, too. Since they're listed and have an FC pointing at the first one
137
137
- [x] inline struct getters during code generation (We'd like `x.h1.h2`)
138
138
- [ ] Better FC for parse errors (both EOF and the ones that show up just after the error)
···
149
149
- [x] Maybe tag some functions as inline
150
150
- [x] Eq Nat is not tail recursive because of the call to `==`
151
151
- [ ] Eq Nat does things the hard way, can we turn it into `==`?
152
+
- We could have a compile time substitution for the function
153
+
- Maybe this could reify how we're hard coding functions to js operators
152
154
- [x] use hint table for auto solving. (I think walking the `toList` is a big chunk of performance in `Elab.newt`.)
153
155
- [x] implement string enum (or number, but I'm using strings for tags at the moment)
154
156
- [x] use monaco input method instead of lean's
155
157
- [x] `Def` is shadowed between Types and Syntax (TCon vs DCon), detect this
156
-
- [ ] constructor magic for Bool?
158
+
- [x] constructor magic for Bool?
157
159
- We'd have to make assumptions about order.
158
160
- we could special case some translations
159
161
- extra code.
···
161
163
- Two issues
162
164
- I'm rewriting stuff in the context, leaving it in a bad state (forward references). I think I can avoid this.
163
165
- The variables at the end of pattern matching have types with references in the wrong order. I think we can reorder them on dependencies.
164
-
- should w
166
+
- In some cases Idris stuffs Erased into types.
165
167
- Improve `auto`
166
168
- [ ] Improve cases where the auto isn't solved because of a type error
167
169
- [ ] Handle `Foo Blah`, `Foo a => Bar a` vs `Bar Blah`
168
170
- [ ] Add some optimizations
169
-
- [ ] see if we can make the typeclass stuff a little leaner, e.g. inline a projection of a static record.
171
+
- [ ] see if we can make the typeclass stuff a little leaner, e.g. inline a projection of a static record. (Maybe done now with the inlining, but verify.)
170
172
- It would be nice if IO looked like imperative JS, but that might be a bit of a stretch.
171
-
172
-
173
173
- [ ] warn on unused imports?
174
174
- Probably have to mark on name lookup, maybe wait until we have query-based
175
175
- [x] redo code to determine base path
···
185
185
- [x] get port to run
186
186
- [x] something goes terribly wrong with traverse_ and for_ (related to erasure, I think)
187
187
- [x] ~~don't use `take` - it's not stack safe~~ The newt version is stack safe
188
-
- [ ] show user hole info even in case of error
188
+
- [x] show user hole info even in case of error
189
189
- [x] tokenizer that can be ported to newt
190
190
- [ ] Add default path for library, so we don't need symlinks everywhere and can write tests for the library
191
191
- We need this to work for tests / dev and for installed newt.
192
+
- The new `FileSource` mechanism might help here.
192
193
- [x] string interpolation?
193
194
- The tricky part here is the `}` - I need to run the normal tokenizer in a way that treats `}` specially.
194
195
- Idris handles `putStrLn "done \{ show $ add {x=1} 2}"` - it recurses for `{` `}` pairs. Do we want that complexity?
···
202
203
- [x] record update sugar
203
204
- [x] Change `Ord` to be more like Idris - LT / EQ / GT (and entail equality)
204
205
- [ ] Consider making `<` independent of `Ord`, so we get the `<` oper in the javascript.
206
+
- Maybe a "default implementation" deal.
205
207
- [x] Keep a `compare` function on `SortedMap` (like lean)
206
208
- `emptyMap` helper defaults to `compare` from `Ord a`
207
209
- [x] keymap for monaco
···
258
260
- Probably fixed by trying to solve auto immediately
259
261
- [x] Can't skip an auto. We need `{{_}}` to be auto or have a `%search` syntax.
260
262
- [x] add filenames to FC
261
-
- [ ] Add full ranges to FC
263
+
- [x] Add full ranges to FC
264
+
- [ ] Done, but we need to fix a bunch of FC in the parser
262
265
- [x] maybe use backtick for javascript so we don't highlight strings as JS
263
266
- [x] dead code elimination
264
267
- [x] imported files leak info messages everywhere