From f373f67048e2566af7fb5eaa68c3bec11607010a Mon Sep 17 00:00:00 2001
From: Adin Scannell <ascannell@google.com>
Date: Fri, 23 Apr 2021 17:31:18 -0700
Subject: Improve nogo action cache-ability.

Presently, the standard library facts are not serialized in a deterministic
order. This means that they have the possibility to change on each iteration,
requiring a large scale re-analysis of all downstream actions, which includes
all packages.

Improve cache-ability of nogo actions by improving the determinism of the both
facts and findings. Internally, default facts should be serialized as a sorted
list for this reason already.

PiperOrigin-RevId: 370188259
---
 tools/nogo/findings.go | 28 ++++++++++++++++++++++++++++
 tools/nogo/nogo.go     | 33 +++++++++++++++++++++++++++++----
 2 files changed, 57 insertions(+), 4 deletions(-)

(limited to 'tools')

diff --git a/tools/nogo/findings.go b/tools/nogo/findings.go
index 5bd850269..a00cfe813 100644
--- a/tools/nogo/findings.go
+++ b/tools/nogo/findings.go
@@ -19,6 +19,7 @@ import (
 	"fmt"
 	"go/token"
 	"io/ioutil"
+	"sort"
 )
 
 // Finding is a single finding.
@@ -44,6 +45,33 @@ func WriteFindingsToFile(findings []Finding, filename string) error {
 
 // WriteFindingsToBytes serializes findings as bytes.
 func WriteFindingsToBytes(findings []Finding) ([]byte, error) {
+	// N.B. Sort all the findings in order to maximize cacheability.
+	sort.Slice(findings, func(i, j int) bool {
+		switch {
+		case findings[i].Position.Filename < findings[j].Position.Filename:
+			return true
+		case findings[i].Position.Filename > findings[j].Position.Filename:
+			return false
+		case findings[i].Position.Line < findings[j].Position.Line:
+			return true
+		case findings[i].Position.Line > findings[j].Position.Line:
+			return false
+		case findings[i].Position.Column < findings[j].Position.Column:
+			return true
+		case findings[i].Position.Column > findings[j].Position.Column:
+			return false
+		case findings[i].Category < findings[j].Category:
+			return true
+		case findings[i].Category > findings[j].Category:
+			return false
+		case findings[i].Message < findings[j].Message:
+			return true
+		case findings[i].Message > findings[j].Message:
+			return false
+		default:
+			return false
+		}
+	})
 	return json.Marshal(findings)
 }
 
diff --git a/tools/nogo/nogo.go b/tools/nogo/nogo.go
index 779d4d6d8..c1b88e89f 100644
--- a/tools/nogo/nogo.go
+++ b/tools/nogo/nogo.go
@@ -34,6 +34,7 @@ import (
 	"path"
 	"path/filepath"
 	"reflect"
+	"sort"
 	"strings"
 
 	"golang.org/x/tools/go/analysis"
@@ -75,6 +76,12 @@ type loader func(string) ([]byte, error)
 // saver is a fact-saver function.
 type saver func([]byte) error
 
+// stdlibFact is used for serialiation.
+type stdlibFact struct {
+	Package string
+	Facts   []byte
+}
+
 // factLoader returns a function that loads facts.
 //
 // This resolves all standard library facts and imported package facts up
@@ -90,10 +97,17 @@ func (c *PackageConfig) factLoader() (loader, error) {
 		if err != nil {
 			return nil, fmt.Errorf("error loading stdlib facts from %q: %w", c.StdlibFacts, err)
 		}
-		var stdlibFacts map[string][]byte
-		if err := json.Unmarshal(data, &stdlibFacts); err != nil {
+		var (
+			stdlibFactsSorted []stdlibFact
+			stdlibFacts       = make(map[string][]byte)
+		)
+		// See below re: sorted serialization.
+		if err := json.Unmarshal(data, &stdlibFactsSorted); err != nil {
 			return nil, fmt.Errorf("error loading stdlib facts: %w", err)
 		}
+		for _, stdlibFact := range stdlibFactsSorted {
+			stdlibFacts[stdlibFact.Package] = stdlibFact.Facts
+		}
 		for pkg, data := range stdlibFacts {
 			allFacts[pkg] = data
 		}
@@ -327,8 +341,19 @@ func CheckStdlib(config *StdlibConfig, analyzers []*analysis.Analyzer) (allFindi
 		return nil, nil, fmt.Errorf("no stdlib facts found: misconfiguration?")
 	}
 
-	// Write out all findings.
-	factData, err := json.Marshal(stdlibFacts)
+	// Write out all findings. Note that the standard library facts
+	// must be serialized in a sorted order to ensure cacheability.
+	stdlibFactsSorted := make([]stdlibFact, 0, len(stdlibFacts))
+	for pkg, facts := range stdlibFacts {
+		stdlibFactsSorted = append(stdlibFactsSorted, stdlibFact{
+			Package: pkg,
+			Facts:   facts,
+		})
+	}
+	sort.Slice(stdlibFactsSorted, func(i, j int) bool {
+		return stdlibFactsSorted[i].Package < stdlibFactsSorted[j].Package
+	})
+	factData, err := json.Marshal(stdlibFactsSorted)
 	if err != nil {
 		return nil, nil, fmt.Errorf("error saving stdlib facts: %w", err)
 	}
-- 
cgit v1.2.3