Add map<string, AnyValue> to the `attribute` package by itssaharsh · Pull Request #7943 · open-telemetry/opentelemetry-go

itssaharsh · 2026-02-23T16:13:25Z

This PR focuses on adding the Support map<string, AnyValue> to attribute package
tracked in: #7935

pellared · 2026-02-23T16:19:58Z

Changing to draft as this does not look to be ready for review.
I suggest also looking at #7944 for reference.

itssaharsh · 2026-02-23T16:23:09Z

Okay this was an initial work overall and I will use the references now and change accordingly

codecov · 2026-02-23T18:09:03Z

Codecov Report

❌ Patch coverage is 93.02326% with 6 lines in your changes missing coverage. Please review.
✅ Project coverage is 81.7%. Comparing base (de5fb3a) to head (7c700fd).

Files with missing lines	Patch %	Lines
attribute/value.go	91.8%	4 Missing and 2 partials ⚠️

Additional details and impacted files

@@          Coverage Diff          @@
##            main   #7943   +/-   ##
=====================================
  Coverage   81.6%   81.7%           
=====================================
  Files        304     304           
  Lines      23424   23510   +86     
=====================================
+ Hits       19127   19211   +84     
- Misses      3909    3910    +1     
- Partials     388     389    +1

Files with missing lines	Coverage Δ
attribute/hash.go	`90.3% <100.0%> (+1.2%)`	⬆️
attribute/key.go	`100.0% <100.0%> (ø)`
attribute/kv.go	`100.0% <100.0%> (ø)`
attribute/type_string.go	`60.0% <ø> (ø)`
attribute/value.go	`92.3% <91.8%> (-0.3%)`	⬇️

... and 4 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

pellared · 2026-02-24T18:14:41Z

@itssaharsh, are you able to work on it this week? MAP is the only type that we do not have a PR "ready".

itssaharsh · 2026-02-25T06:00:48Z

@itssaharsh, are you able to work on it this week? MAP is the only type that we do not have a PR "ready".

I think it is ready for review. I viewed the other PRs and changed my code accordingly

pellared · 2026-02-25T10:10:24Z

I viewed the other PRs and changed my code accordingly

I kindly ask you do it once more with more attention.
It misses test coverage and the existing tests should be refactored.
Changlog entry is missing.
Misses benchmarks and the results in the PR description.
Also please update the PR title and other parts of the description.
This is what I found in 1 minute.

pellared · 2026-02-25T10:16:04Z

attribute/value.go

+// without type wrappers).
+func (v Value) mapEmitValue() any {
+	m := v.asMap()
+	raw := make(map[string]any, len(m))


This looks very inefficient as it would cause a lot of allocations. Can we try doing something similar to

opentelemetry-go/attribute/value.go

Lines 302 to 322 in dedc8f7

case SLICE:

slice := v.asSlice()

if len(slice) == 0 {

return "[]"

}

var b strings.Builder

b.WriteRune('[') //nolint:revive // No need to check error for strings.Builder.

for i, val := range slice {

if i > 0 {

b.WriteRune(',') //nolint:revive // No need to check error for strings.Builder.

}

if val.Type() == STRING {

b.WriteRune('"') //nolint:revive // No need to check error for strings.Builder.

b.WriteString(val.Emit()) //nolint:revive // No need to check error for strings.Builder.

b.WriteRune('"') //nolint:revive // No need to check error for strings.Builder.

} else {

b.WriteString(val.Emit()) //nolint:revive // No need to check error for strings.Builder.

}

}

b.WriteRune(']') //nolint:revive // No need to check error for strings.Builder.

return b.String()

?

Note that Emit is currently used by the Prometheus exporter (in getAttrs function) so it has to be efficient.

attribute/value.go

pellared · 2026-02-25T16:36:20Z

attribute/value.go

+				case math.IsInf(f, -1):
+					_, _ = b.WriteString(`"-Infinity"`)
+				default:
+					_, _ = b.WriteString(kv.Value.Emit())


What would be more efficient Emit or json.Marshal?
Or maybe we should even use https://pkg.go.dev/encoding/json#NewEncoder?

I think emit and marshal are around same benchmarks but encoder has less allocations.
https://github.com/bytedance/sonic found this it was shown that it is very fast but haven't used yet. Your thoughts?

I think emit and marshal are around same benchmarks but encoder has less allocations

Leaning towards encoder but we should have benchmark to validate it. I also plan to improve the Emit in my PR for SLICE. I just need to find some time for it

https://github.com/bytedance/sonic found this it was shown that it is very fast but haven't used yet. Your thoughts?

We do not want to add any new dependency for this.

Okay got it and I performed benchamrks locally and found that newencoder was around 10 percent slower and had around 7 percent more alocations
So should I use it and update the other json.Marshals also? and should I also post the benchmarks here

So should I use it and updta the other json.Marshals also?

Yes, given it makes it more efficient.

and should I also post the benchmarks here

It would be very good to have even for sake of retrospecting.

I added a extra function for the slices to be encoded to remove redudant code is that okay?

The tests are too long I used
go test -bench=. -benchmem -count=5 -benchtime=5s -timeout=30m ./attribute/benchmark_test.go so the outputs are too long. Apologies for that should I add all the benchmarks here?

No need to add these parameters -count=5 -benchtime=5s -timeout=30m

Co-authored-by: Robert Pająk <pellared@hotmail.com>

itssaharsh · 2026-02-26T16:22:35Z

I donot understand why the links(fail fast) is failing, I need some help with that one

itssaharsh · 2026-02-27T20:35:36Z

Marshal is faster and better allocation wise

pellared · 2026-02-27T20:38:18Z

@itssaharsh use benchstat for comparing benchmarks

I thought you just wanted to point out the "current state".

pellared · 2026-02-27T20:39:28Z

Marshal is faster and better allocation wise

I see the opposite.

itssaharsh · 2026-02-27T20:42:24Z

Marshal is faster and better allocation wise

I see the opposite.

I saw that Marshal take more time overall and indivisually also and allocs also letme use benchstat and upload them here sorry for the confusion

itssaharsh · 2026-02-27T21:06:26Z

I think that marshal is better somewhat so should I revert back to marshal? I think marshal also escapes &,<,> so should I specify these in case string also?

pellared · 2026-02-27T22:03:34Z

Please double check the output (with understanding 😉 )

¹ need >= 6 samples for confidence interval at level 0.95

Look that encoder for instance here is probably better

Map/Emit-16 24.00 ± ∞ ¹ 23.00 ± ∞ ¹ ~ (p=1.000 n=1) ³

For map type the encoder is reused multiple types.

itssaharsh · 2026-02-28T10:24:14Z

                                            │ marshal.txt  │             encoder.txt              │
                                            │    sec/op    │    sec/op      vs base               │
Bool/Value-16                                 2.255n ±  8%    2.426n ±  6%   +7.58% (p=0.002 n=7)
Bool/KeyValue-16                              22.39n ±  1%    24.51n ±  3%   +9.47% (p=0.001 n=7)
Bool/AsBool-16                                2.221n ±  1%    2.438n ±  6%   +9.77% (p=0.001 n=7)
Bool/Emit-16                                  24.64n ±  1%    27.97n ±  5%  +13.51% (p=0.001 n=7)
BoolSlice/Value-16                            208.7n ±  1%    227.0n ±  3%   +8.77% (p=0.001 n=7)
BoolSlice/KeyValue-16                         210.4n ±  8%    226.2n ±  7%        ~ (p=0.072 n=7)
BoolSlice/AsBoolSlice-16                      84.56n ±  1%    86.47n ±  6%        ~ (p=0.710 n=7)
BoolSlice/Emit-16                             769.5n ±  1%    771.4n ±  4%        ~ (p=0.318 n=7)
Int/Value-16                                  2.222n ±  8%    2.436n ±  6%   +9.63% (p=0.004 n=7)
Int/KeyValue-16                               24.08n ±  0%    24.35n ±  7%        ~ (p=0.710 n=7)
Int/Emit-16                                   29.28n ±  1%    29.43n ±  7%        ~ (p=0.902 n=7)
IntSlice/Value-16                             251.8n ±  4%    260.8n ± 10%        ~ (p=0.259 n=7)
IntSlice/KeyValue-16                          269.2n ±  1%    273.4n ±  7%   +1.56% (p=0.038 n=7)
IntSlice/Emit-16                              594.4n ±  6%    716.9n ±  6%  +20.61% (p=0.001 n=7)
Int64/Value-16                                2.289n ±  2%    2.463n ±  7%   +7.60% (p=0.001 n=7)
Int64/KeyValue-16                             23.60n ±  2%    24.13n ±  6%        ~ (p=0.195 n=7)
Int64/AsInt64-16                              2.374n ±  4%    2.403n ±  5%        ~ (p=0.165 n=7)
Int64/Emit-16                                 28.81n ±  6%    29.74n ±  7%        ~ (p=0.259 n=7)
Int64Slice/Value-16                           243.5n ±  4%    245.3n ±  7%        ~ (p=0.097 n=7)
Int64Slice/KeyValue-16                        255.0n ±  5%    253.9n ±  3%        ~ (p=0.620 n=7)
Int64Slice/AsInt64Slice-16                    93.03n ±  2%    99.70n ±  5%   +7.17% (p=0.001 n=7)
Int64Slice/Emit-16                            575.5n ±  3%    716.8n ±  6%  +24.55% (p=0.001 n=7)
Float64/Value-16                              2.397n ±  4%    2.442n ±  6%        ~ (p=0.165 n=7)
Float64/KeyValue-16                           23.21n ±  0%    24.32n ±  2%   +4.78% (p=0.001 n=7)
Float64/AsFloat64-16                          2.368n ±  2%    2.401n ±  6%   +1.39% (p=0.034 n=7)
Float64/Emit-16                               251.2n ±  5%    259.0n ±  4%        ~ (p=0.778 n=7)
Float64Slice/Value-16                         247.9n ±  3%    254.1n ±  3%   +2.50% (p=0.004 n=7)
Float64Slice/KeyValue-16                      248.7n ±  1%    260.3n ±  5%   +4.66% (p=0.038 n=7)
Float64Slice/AsFloat64Slice-16                93.38n ±  1%    98.42n ±  6%   +5.40% (p=0.026 n=7)
Float64Slice/Emit-16                          933.4n ±  6%   1068.0n ±  2%  +14.42% (p=0.001 n=7)
String/Value-16                               3.080n ±  2%    3.238n ±  4%   +5.13% (p=0.010 n=7)
String/KeyValue-16                            22.36n ±  3%    22.74n ±  7%        ~ (p=0.097 n=7)
String/AsString-16                            2.861n ±  1%    2.749n ±  7%   -3.91% (p=0.024 n=7)
String/Emit-16                                24.53n ±  3%    25.96n ±  4%   +5.83% (p=0.001 n=7)
StringSlice/Value-16                          288.8n ±  2%    308.6n ±  3%   +6.86% (p=0.001 n=7)
StringSlice/KeyValue-16                       300.3n ±  4%    293.6n ±  8%        ~ (p=0.973 n=7)
StringSlice/AsStringSlice-16                  123.7n ±  1%    131.7n ±  5%   +6.47% (p=0.002 n=7)
StringSlice/Emit-16                           687.5n ±  3%    825.4n ±  6%  +20.06% (p=0.001 n=7)
Map/Value-16                                  1.352µ ±  2%    1.368µ ± 10%        ~ (p=0.105 n=7)
Map/KeyValue-16                               1.328µ ±  2%    1.268µ ± 10%        ~ (p=0.259 n=7)
Map/AsMap-16                                  834.5n ±  1%    769.9n ±  8%   -7.74% (p=0.004 n=7)
Map/Emit-16                                   3.620µ ±  2%    3.646µ ±  3%        ~ (p=0.057 n=7)
SetEquals/Empty-16                            19.03n ±  3%    19.57n ±  1%   +2.84% (p=0.001 n=7)
SetEquals/1_string_attribute-16               38.21n ±  3%    37.09n ±  4%   -2.93% (p=0.006 n=7)
SetEquals/10_string_attributes-16             191.1n ±  1%    180.9n ±  1%   -5.34% (p=0.001 n=7)
SetEquals/1_int_attribute-16                  38.51n ±  1%    37.23n ±  1%   -3.32% (p=0.001 n=7)
SetEquals/10_int_attributes-16                187.1n ± 11%    185.2n ±  3%        ~ (p=0.710 n=7)
EquivalentMapAccess/Empty-16                  14.49n ±  2%    15.38n ±  4%   +6.14% (p=0.001 n=7)
EquivalentMapAccess/1_string_attribute-16     14.58n ±  9%    15.53n ±  2%        ~ (p=0.118 n=7)
EquivalentMapAccess/10_string_attributes-16   15.97n ±  7%    15.41n ±  3%   -3.51% (p=0.002 n=7)
EquivalentMapAccess/1_int_attribute-16        16.92n ±  1%    15.05n ±  2%  -11.05% (p=0.001 n=7)
EquivalentMapAccess/10_int_attributes-16      16.98n ±  1%    14.84n ±  2%  -12.60% (p=0.001 n=7)
geomean                                       59.59n          61.58n         +3.34%

                                            │  marshal.txt   │              encoder.txt               │
                                            │      B/op      │     B/op      vs base                  │
Bool/Value-16                                   0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Bool/KeyValue-16                                0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Bool/AsBool-16                                  0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Bool/Emit-16                                    0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
BoolSlice/Value-16                              6.000 ± 0%       6.000 ± 0%         ~ (p=1.000 n=7) ¹
BoolSlice/KeyValue-16                           6.000 ± 0%       6.000 ± 0%         ~ (p=1.000 n=7) ¹
BoolSlice/AsBoolSlice-16                        3.000 ± 0%       3.000 ± 0%         ~ (p=1.000 n=7) ¹
BoolSlice/Emit-16                               54.00 ± 0%       54.00 ± 0%         ~ (p=1.000 n=7) ¹
Int/Value-16                                    0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int/KeyValue-16                                 0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int/Emit-16                                     0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
IntSlice/Value-16                               48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
IntSlice/KeyValue-16                            48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
IntSlice/Emit-16                                80.00 ± 0%      176.00 ± 0%  +120.00% (p=0.001 n=7)
Int64/Value-16                                  0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int64/KeyValue-16                               0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int64/AsInt64-16                                0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int64/Emit-16                                   0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Int64Slice/Value-16                             48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
Int64Slice/KeyValue-16                          48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
Int64Slice/AsInt64Slice-16                      24.00 ± 0%       24.00 ± 0%         ~ (p=1.000 n=7) ¹
Int64Slice/Emit-16                              80.00 ± 0%      176.00 ± 0%  +120.00% (p=0.001 n=7)
Float64/Value-16                                0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Float64/KeyValue-16                             0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Float64/AsFloat64-16                            0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
Float64/Emit-16                                 16.00 ± 0%       16.00 ± 0%         ~ (p=1.000 n=7) ¹
Float64Slice/Value-16                           48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
Float64Slice/KeyValue-16                        48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
Float64Slice/AsFloat64Slice-16                  24.00 ± 0%       24.00 ± 0%         ~ (p=1.000 n=7) ¹
Float64Slice/Emit-16                            80.00 ± 0%      176.00 ± 0%  +120.00% (p=0.001 n=7)
String/Value-16                                 0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
String/KeyValue-16                              0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
String/AsString-16                              0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
String/Emit-16                                  0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
StringSlice/Value-16                            96.00 ± 0%       96.00 ± 0%         ~ (p=1.000 n=7) ¹
StringSlice/KeyValue-16                         96.00 ± 0%       96.00 ± 0%         ~ (p=1.000 n=7) ¹
StringSlice/AsStringSlice-16                    48.00 ± 0%       48.00 ± 0%         ~ (p=1.000 n=7) ¹
StringSlice/Emit-16                             168.0 ± 0%       232.0 ± 0%   +38.10% (p=0.001 n=7)
Map/Value-16                                  1.016Ki ± 0%     1.016Ki ± 0%         ~ (p=1.000 n=7) ¹
Map/KeyValue-16                               1.016Ki ± 0%     1.016Ki ± 0%         ~ (p=1.000 n=7) ¹
Map/AsMap-16                                    944.0 ± 0%       944.0 ± 0%         ~ (p=1.000 n=7) ¹
Map/Emit-16                                     841.0 ± 0%       961.0 ± 0%   +14.27% (p=0.001 n=7)
SetEquals/Empty-16                              0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
SetEquals/1_string_attribute-16                 0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
SetEquals/10_string_attributes-16               0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
SetEquals/1_int_attribute-16                    0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
SetEquals/10_int_attributes-16                  0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
EquivalentMapAccess/Empty-16                    0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
EquivalentMapAccess/1_string_attribute-16       0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
EquivalentMapAccess/10_string_attributes-16     0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
EquivalentMapAccess/1_int_attribute-16          0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
EquivalentMapAccess/10_int_attributes-16        0.000 ± 0%       0.000 ± 0%         ~ (p=1.000 n=7) ¹
geomean                                                    ²                   +5.58%               ²
¹ all samples are equal
² summaries must be >0 to compute geomean

                                            │ marshal.txt  │             encoder.txt             │
                                            │  allocs/op   │ allocs/op   vs base                 │
Bool/Value-16                                 0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Bool/KeyValue-16                              0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Bool/AsBool-16                                0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Bool/Emit-16                                  0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
BoolSlice/Value-16                            2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
BoolSlice/KeyValue-16                         2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
BoolSlice/AsBoolSlice-16                      1.000 ± 0%     1.000 ± 0%        ~ (p=1.000 n=7) ¹
BoolSlice/Emit-16                             6.000 ± 0%     6.000 ± 0%        ~ (p=1.000 n=7) ¹
Int/Value-16                                  0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int/KeyValue-16                               0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int/Emit-16                                   0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
IntSlice/Value-16                             2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
IntSlice/KeyValue-16                          2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
IntSlice/Emit-16                              4.000 ± 0%     5.000 ± 0%  +25.00% (p=0.001 n=7)
Int64/Value-16                                0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64/KeyValue-16                             0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64/AsInt64-16                              0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64/Emit-16                                 0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64Slice/Value-16                           2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64Slice/KeyValue-16                        2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64Slice/AsInt64Slice-16                    1.000 ± 0%     1.000 ± 0%        ~ (p=1.000 n=7) ¹
Int64Slice/Emit-16                            4.000 ± 0%     5.000 ± 0%  +25.00% (p=0.001 n=7)
Float64/Value-16                              0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64/KeyValue-16                           0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64/AsFloat64-16                          0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64/Emit-16                               2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64Slice/Value-16                         2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64Slice/KeyValue-16                      2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64Slice/AsFloat64Slice-16                1.000 ± 0%     1.000 ± 0%        ~ (p=1.000 n=7) ¹
Float64Slice/Emit-16                          4.000 ± 0%     5.000 ± 0%  +25.00% (p=0.001 n=7)
String/Value-16                               0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
String/KeyValue-16                            0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
String/AsString-16                            0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
String/Emit-16                                0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
StringSlice/Value-16                          2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
StringSlice/KeyValue-16                       2.000 ± 0%     2.000 ± 0%        ~ (p=1.000 n=7) ¹
StringSlice/AsStringSlice-16                  1.000 ± 0%     1.000 ± 0%        ~ (p=1.000 n=7) ¹
StringSlice/Emit-16                           4.000 ± 0%     5.000 ± 0%  +25.00% (p=0.001 n=7)
Map/Value-16                                  4.000 ± 0%     4.000 ± 0%        ~ (p=1.000 n=7) ¹
Map/KeyValue-16                               4.000 ± 0%     4.000 ± 0%        ~ (p=1.000 n=7) ¹
Map/AsMap-16                                  3.000 ± 0%     3.000 ± 0%        ~ (p=1.000 n=7) ¹
Map/Emit-16                                   24.00 ± 0%     23.00 ± 0%   -4.17% (p=0.001 n=7)
SetEquals/Empty-16                            0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
SetEquals/1_string_attribute-16               0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
SetEquals/10_string_attributes-16             0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
SetEquals/1_int_attribute-16                  0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
SetEquals/10_int_attributes-16                0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
EquivalentMapAccess/Empty-16                  0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
EquivalentMapAccess/1_string_attribute-16     0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
EquivalentMapAccess/10_string_attributes-16   0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
EquivalentMapAccess/1_int_attribute-16        0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
EquivalentMapAccess/10_int_attributes-16      0.000 ± 0%     0.000 ± 0%        ~ (p=1.000 n=7) ¹
geomean                                                  ²                +1.65%               ²
¹ all samples are equal
² summaries must be >0 to compute geomean

itssaharsh · 2026-02-28T10:35:15Z

I read the benchmarks and I think that Encoder is slightly better in case of map and in other cases marshal is better or they are even. So let's use encoder for map and marshal for everything else.

itssaharsh added 2 commits February 23, 2026 13:50

map implement

22a35d5

specs match

15de6d7

itssaharsh requested review from MrAlias, XSAM, dashpole, dmathieu, flc1125 and pellared as code owners February 23, 2026 16:13

lint

be03f02

pellared marked this pull request as draft February 23, 2026 16:20

itssaharsh added 2 commits February 23, 2026 16:31

lint

446da2a

lint

1d29018

pellared reviewed Feb 25, 2026

View reviewed changes

itssaharsh force-pushed the map branch from 231ed16 to 1d29018 Compare February 25, 2026 14:04

itssaharsh added 2 commits February 25, 2026 14:06

fixes

5120aa9

lint

50d8058

pellared reviewed Feb 25, 2026

View reviewed changes

attribute/value.go Outdated Show resolved Hide resolved

attribute/value.go Outdated Show resolved Hide resolved

itssaharsh added 2 commits February 25, 2026 15:08

edge cases handling

2b5a605

edge cases handling

43eabe1

pellared reviewed Feb 25, 2026

View reviewed changes

attribute/value.go Outdated Show resolved Hide resolved

json

e9a1046

pellared reviewed Feb 25, 2026

View reviewed changes

itssaharsh and others added 2 commits February 25, 2026 22:19

Update attribute/value.go

2b60037

Co-authored-by: Robert Pająk <pellared@hotmail.com>

json newencoder

ed4b301

itssaharsh added 2 commits February 26, 2026 15:17

lint

45926f7

fix

4cf698f

itssaharsh force-pushed the map branch from 7b20a04 to 4cf698f Compare February 26, 2026 15:43

itssaharsh and others added 3 commits February 26, 2026 21:14

Merge branch 'main' into map

4efede6

lint

1bd33f2

lint

9945db7

itssaharsh and others added 3 commits March 2, 2026 11:20

encoder for map

3a1b94e

Merge branch 'main' into map

112bc8c

Merge branch 'main' into map

7c700fd

itssaharsh requested a review from pellared March 6, 2026 12:09

	case SLICE:
	slice := v.asSlice()
	if len(slice) == 0 {
	return "[]"
	}
	var b strings.Builder
	b.WriteRune('[') //nolint:revive // No need to check error for strings.Builder.
	for i, val := range slice {
	if i > 0 {
	b.WriteRune(',') //nolint:revive // No need to check error for strings.Builder.
	}
	if val.Type() == STRING {
	b.WriteRune('"') //nolint:revive // No need to check error for strings.Builder.
	b.WriteString(val.Emit()) //nolint:revive // No need to check error for strings.Builder.
	b.WriteRune('"') //nolint:revive // No need to check error for strings.Builder.
	} else {
	b.WriteString(val.Emit()) //nolint:revive // No need to check error for strings.Builder.
	}
	}
	b.WriteRune(']') //nolint:revive // No need to check error for strings.Builder.
	return b.String()

Conversation

itssaharsh commented Feb 23, 2026

Uh oh!

pellared commented Feb 23, 2026

Uh oh!

itssaharsh commented Feb 23, 2026

Uh oh!

codecov bot commented Feb 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

pellared commented Feb 24, 2026

Uh oh!

itssaharsh commented Feb 25, 2026

Uh oh!

pellared commented Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pellared Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

pellared Feb 25, 2026

Choose a reason for hiding this comment

Uh oh!

itssaharsh Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pellared Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

itssaharsh Feb 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pellared Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

itssaharsh Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

itssaharsh Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

pellared Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

itssaharsh commented Feb 26, 2026

Uh oh!

itssaharsh commented Feb 27, 2026

Uh oh!

pellared commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pellared commented Feb 27, 2026

Uh oh!

itssaharsh commented Feb 27, 2026

Uh oh!

itssaharsh commented Feb 27, 2026

Uh oh!

pellared commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

itssaharsh commented Feb 28, 2026

Uh oh!

itssaharsh commented Feb 28, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

codecov bot commented Feb 23, 2026 •

edited

Loading

pellared commented Feb 25, 2026 •

edited

Loading

pellared Feb 25, 2026 •

edited

Loading

itssaharsh Feb 25, 2026 •

edited

Loading

pellared Feb 25, 2026 •

edited

Loading

itssaharsh Feb 25, 2026 •

edited

Loading

pellared commented Feb 27, 2026 •

edited

Loading

pellared commented Feb 27, 2026 •

edited

Loading