feat: named and unnamed type assignment 3 of 3

[piux2]

This is the last part of the solution for issue #1141.
The 1 of 3 of the solution can be found in PR #1143.
The 2 of 3 of the solution can be found in PR #1246

It decomposes function calls that return multiple values in the preprocess.

Here is the problem to solve:

u1, n2 = x()

How do we ensure that the returned multiple values from a function call adhere to named and unnamed type assignment specifications? Additionally, we want to solve this problem during preprocessing instead of at runtime to minimize the impact on runtime performance.

The main ideas:

u1, n2 = x()  << decompose the statement to the following two lines 
// .tmp_1, .tmp_2 := x() 
// u1, n2 = .tmp_1, .tmp_2

then we can apply name and unname type conversion specs to the second line. 
u1, n2 = _tmp_1, _tmp_2

Here are the example code and the explanation

// decompose_filetest.gno
package main

  type nat []int

  func x() (nat, []int) {
    a := nat{1}
    b := []int{2}
    return a, b
  }

  func main() {
    var u1 []int
    var n2 nat

    u1, n2 = x() 
    // .tmp_1, .tmp_2 := x() 
    // u1, n2 = .tmp_1, .tmp_2

    println(u1)
    println(n2)

  }

  // Output:
  // slice[(1 int)]
  // (slice[(2 int)] main.nat)

Here is the simplified recursive tree of the transformation in the preprocess

Here are the major steps involved in this decomposition during preprocessing:

• Create hidden temporary name expressions .tmp1, .tmp2. In Go, a leading dot is not valid in variable names, ensuring that users cannot create names that clash with these hidden variables.

• Create two statements in the block: one for defining and one for assigning.

 .tmp1, .tmp2 := x() 
 u1, n2 = .tmp_1, .tmp_2

• Preprocess each newly created statements

• Replace the original statement with the two newly created statements.

Here are two additional changes to facilitate above.

• Update the FuncValue's body in updates := pn.PrepareNewValues(pv) since its source Body has been changed during preprocessing.

• Replace all for index := range Body with for i:=0; i < len(Body); i++ in transcribe.go since the body length might change due to the decomposition.

Contributors' checklist...

• Added new tests, or not needed, or not feasible
• Provided an example (e.g. screenshot) to aid review or the PR is self-explanatory
• Updated the official documentation or not needed
• No breaking changes were made
• Added references to related issues and PRs
• Provided any useful hints for running manual tests
• Added new benchmarks to generated graphs, if any. More info here.

[codecov]

Codecov Report

Attention: Patch coverage is 33.80282% with 47 lines in your changes missing coverage. Please review.

Project coverage is 54.99%. Comparing base (6432573) to head (e214285).
Report is 15 commits behind head on master.

Files	Patch %	Lines
gnovm/pkg/gnolang/preprocess.go	25.58%	29 Missing and 3 partials ⚠️
gnovm/pkg/gnolang/nodes.go	23.07%	10 Missing ⚠️
gnovm/pkg/gnolang/values.go	33.33%	3 Missing and 1 partial ⚠️
gnovm/pkg/gnolang/transcribe.go	88.88%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #2367      +/-   ##
==========================================
- Coverage   55.02%   54.99%   -0.03%     
==========================================
  Files         595      595              
  Lines       79645    79960     +315     
==========================================
+ Hits        43824    43974     +150     
- Misses      32503    32670     +167     
+ Partials     3318     3316       -2     

Flag	Coverage Δ
contribs/gnodev	26.00% <ø> (+0.34%)	⬆️
contribs/gnofaucet	15.31% <ø> (+0.85%)	⬆️
contribs/gnokeykc	0.00% <ø> (ø)
contribs/gnomd	0.00% <ø> (ø)
gno.land	64.24% <ø> (-0.16%)	⬇️
gnovm	60.25% <33.80%> (-0.06%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[deelawn]

This seems to work well. I made some suggestion to fix typos. Can you add some tests for a couple more cases?

• Multiple of these types assignments within the same block
• This type of assignment inside a global var block

[piux2]

There is conflict need to be resolved with #1426 first.
a,b = x(...) assignment check is not in the preprocess anymore.

[ltzmaxwell]

There is conflict need to be resolved with #1426 first. a,b = x(...) assignment check is not in the preprocess anymore.

Hi @piux2 , part of the logic for *AssignStmt has been moved to type_check.go:

gno/gnovm/pkg/gnolang/type_check.go

Line 710 in b5560e2

func (x *AssignStmt) AssertCompatible(store Store, last BlockNode) {

LMK if there'a any trouble about the conflicts.

[piux2]

There is conflict need to be resolved with #1426 first. a,b = x(...) assignment check is not in the preprocess anymore.

Hi @piux2 , part of the logic for *AssignStmt has been moved to type_check.go:

gno/gnovm/pkg/gnolang/type_check.go

Line 710 in b5560e2

func (x *AssignStmt) AssertCompatible(store Store, last BlockNode) {

LMK if there'a any trouble about the conflicts.

Thank you @ltzmaxwell the conflict is resulted, it is ready to merge

[thehowl]

overall lgtm, a few nits.

[jaekwon]

i'll take a look this week

[jaekwon]

just a few asks but otherwise good to go.