Supercharge traces #978

KtorZ · 2024-07-19T10:01:30Z

Preview

Default options

code	result
`trace @"my_label"`
`trace @"my_label": @"foo", @"bar"`
`trace @"my_label": [1, 2, 3], Some(#"00ff")`
`trace (14, 42)`
`trace fn(x) { x + 1 }`
`trace @"my_label", @"foo", @"bar"`

`--trace-level compact`

code	result
`trace @"my_label"`
`trace @"my_label": @"foo", @"bar"`
`trace @"my_label": [1, 2, 3], Some(#"00ff")`
`trace (14, 42)`
`trace fn(x) { x + 1 }`
`trace @"my_label", @"foo", @"bar"`

Changes

📍 Allow variadic arguments in trace
Although, doesn't do anything with them yet. The idea is to simplify
the use of trace to make it a lot more useful than it currently is.
📍 Type-check variadic traces & desugarize them.
📍 Add 'diagnostic' to the prelude, as well as companion functions.
This is not fully satisfactory as it pollutes a bit the prelude. Ideally, those functions should only be visible
and usable by the underlying trace code. But for now, we'll just go with it.
📍 Allow serialisable (Data-able) arguments to trace
Somehow, we have to patch some function in gen_uplc because of the
module name. I have to look further into this because it isn't normal.
📍 Remove unnecessary code_gen patch.
This is a little weird but, prelude functions are handled slightly
differently.
📍 re-introduce code-gen patch, but with a test.
Actually, this has been a bug for a long time it seems. Calling any
prelude functions using a qualified import would result in a codegen
crash. Whoopsie.

This is now fixed as shown by the regression test.
📍 Rework 'compact' mode for traces
- Trace-if-false are now completely discarded in compact mode.
- Only the label (i.e. first trace argument) is preserved.
- When compiling with tracing compact, the first label MUST unify to
  a string. This shouldn't be an issue generally speaking and would
  enforce that traces follow the pattern
```
label: arg_0[, arg_1, ..., arg_n]
```
Note that what isn't obvious with these changes is that we now support
what the "emit" keyword was trying to achieve; as we compile now with
user-defined traces only, and in compact mode to only keep event
labels in the final contract; while allowing larger payloads with
verbose tracing.
📍 Display expected patterns/tokens in parse error when applicable.
We've never been using those 'expected' tokens captured during
parsing, which is lame because they contain useful information!

This is much better than merely showing our infamous

"Try removing it!"
📍 Provide better parse errors in trace when using comma instead of colon.
📍 Fill-in CHANGELOG

Although, doesn't do anything with them yet. The idea is to simplify the use of trace to make it a lot more useful than it currently is.

This is not fully satisfactory as it pollutes a bit the prelude. Ideally, those functions should only be visible and usable by the underlying trace code. But for now, we'll just go with it.

Somehow, we have to patch some function in gen_uplc because of the module name. I have to look further into this because it isn't normal.

This is a little weird but, prelude functions are handled slightly differently.

Actually, this has been a bug for a long time it seems. Calling any prelude functions using a qualified import would result in a codegen crash. Whoopsie. This is now fixed as shown by the regression test.

- Trace-if-false are now completely discarded in compact mode. - Only the label (i.e. first trace argument) is preserved. - When compiling with tracing _compact_, the first label MUST unify to a string. This shouldn't be an issue generally speaking and would enforce that traces follow the pattern ``` label: arg_0[, arg_1, ..., arg_n] ``` Note that what isn't obvious with these changes is that we now support what the "emit" keyword was trying to achieve; as we compile now with user-defined traces only, and in compact mode to only keep event labels in the final contract; while allowing larger payloads with verbose tracing.

We've never been using those 'expected' tokens captured during parsing, which is lame because they contain useful information! This is much better than merely showing our infamous "Try removing it!"

KtorZ · 2024-07-19T10:46:24Z

crates/aiken-lang/src/parser/error.rs

-    TypeIdent,
-    #[error("I found an unexpected identifier.")]
-    #[diagnostic(help("Try removing it!"))]
-    TermIdent,


KtorZ · 2024-07-19T10:46:29Z

crates/aiken-lang/src/parser/error.rs

-    #[diagnostic(help(
-        "If no label is provided then only variables\nmatching a field name are allowed."
-    ))]
-    RecordPunning,


KtorZ · 2024-07-19T10:46:51Z

crates/aiken-lang/src/gen_uplc.rs

+                                String::new()
+                            } else {
+                                module_name.clone()
+                            },


Actually a long-standing bug, see the corresponding gen_uplc test.

KtorZ · 2024-07-19T10:48:30Z

crates/aiken-lang/src/parser/token.rs

@@ -183,6 +183,6 @@ impl fmt::Display for Token {
            Token::Validator => "validator",
            Token::Via => "via",
        };
-        write!(f, "\"{s}\"")
+        write!(f, "{s}")


I am not 100% sure about the impact of this one. And also not sure why we were wrapping tokens in double-quotes when printing them. This was ugly in the error reporting so I removed them; but I am not certain that we don't show them in other places where the quotes might matter.

Anyway, we should add the quotes in-place where they're needed, and not from this function.

KtorZ · 2024-07-19T11:00:07Z

crates/aiken-project/src/tests/gen_uplc.rs

+            .delayed_if_then_else(constant_true, constant_false),
+        false,
+    )
+}


Note: ☝️ This would crash prior to the fix in this PR.

KtorZ added 2 commits July 18, 2024 10:01

Allow variadic arguments in trace

f9719af

Although, doesn't do anything with them yet. The idea is to simplify the use of trace to make it a lot more useful than it currently is.

Type-check variadic traces & desugarize them.

754ed07

KtorZ requested a review from a team as a code owner July 19, 2024 10:01

KtorZ added 8 commits July 19, 2024 12:28

Add 'diagnostic' to the prelude, as well as companion functions.

beb5ac4

This is not fully satisfactory as it pollutes a bit the prelude. Ideally, those functions should only be visible and usable by the underlying trace code. But for now, we'll just go with it.

Allow serialisable (Data-able) arguments to trace

f823681

Somehow, we have to patch some function in gen_uplc because of the module name. I have to look further into this because it isn't normal.

Remove unnecessary code_gen patch.

5afcc9b

This is a little weird but, prelude functions are handled slightly differently.

re-introduce code-gen patch, but with a test.

a9d782e

Actually, this has been a bug for a long time it seems. Calling any prelude functions using a qualified import would result in a codegen crash. Whoopsie. This is now fixed as shown by the regression test.

Display expected patterns/tokens in parse error when applicable.

2922c0a

We've never been using those 'expected' tokens captured during parsing, which is lame because they contain useful information! This is much better than merely showing our infamous "Try removing it!"

Provide better parse errors in trace when using comma instead of colon.

30ddfa2

Fill-in CHANGELOG

89890f3

KtorZ force-pushed the supercharge_traces branch from 3130429 to 89890f3 Compare July 19, 2024 10:44

KtorZ commented Jul 19, 2024

View reviewed changes

rvcas approved these changes Jul 19, 2024

View reviewed changes

KtorZ merged commit 967c264 into main Jul 23, 2024
12 checks passed

KtorZ deleted the supercharge_traces branch July 23, 2024 06:53

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Supercharge traces #978

Supercharge traces #978

KtorZ commented Jul 19, 2024 •

edited

Loading

KtorZ Jul 19, 2024

KtorZ Jul 19, 2024

KtorZ Jul 19, 2024

KtorZ Jul 19, 2024

KtorZ Jul 19, 2024

Supercharge traces #978

Supercharge traces #978

Conversation

KtorZ commented Jul 19, 2024 • edited Loading

Preview

Default options

--trace-level compact

Changes

KtorZ Jul 19, 2024

Choose a reason for hiding this comment

KtorZ Jul 19, 2024

Choose a reason for hiding this comment

KtorZ Jul 19, 2024

Choose a reason for hiding this comment

KtorZ Jul 19, 2024

Choose a reason for hiding this comment

KtorZ Jul 19, 2024

Choose a reason for hiding this comment

KtorZ commented Jul 19, 2024 •

edited

Loading

`--trace-level compact`