covermerger: add web-git source provider

It allows to speed up the single file calculations because we don't need git clone.

3 files changed, 68 insertions, 3 deletions

diff --git a/pkg/covermerger/covermerger.go b/pkg/covermerger/covermerger.go
index 602a0a263..f9faa9387 100644
--- a/pkg/covermerger/covermerger.go
+++ b/pkg/covermerger/covermerger.go
@@ -122,7 +122,7 @@ type Config struct {
 	Workdir          string
 	skipRepoClone    bool
 	Base             RepoBranchCommit
-	FileVersProvider fileVersProvider
+	FileVersProvider FileVersProvider
 }
 
 func isSchema(fields, schema []string) bool {
diff --git a/pkg/covermerger/monorepo.go b/pkg/covermerger/provider_monorepo.go
index 5ed67e311..21d6792e2 100644
--- a/pkg/covermerger/monorepo.go
+++ b/pkg/covermerger/provider_monorepo.go
@@ -12,7 +12,7 @@ import (
 	"github.com/google/syzkaller/sys/targets"
 )
 
-type fileVersProvider interface {
+type FileVersProvider interface {
 	GetFileVersions(c *Config, targetFilePath string, rbcs []RepoBranchCommit,
 	) (fileVersions, error)
 }
@@ -77,7 +77,7 @@ func (mr *monoRepo) addRepoBranch(rbc RepoBranchCommit) error {
 	return nil
 }
 
-func MakeMonoRepo(workdir string) fileVersProvider {
+func MakeMonoRepo(workdir string) FileVersProvider {
 	rbcPath := workdir + "/repos/linux_kernels"
 	mr := &monoRepo{
 		branches: map[RepoBranchCommit]struct{}{},
diff --git a/pkg/covermerger/provider_web.go b/pkg/covermerger/provider_web.go
new file mode 100644
index 000000000..72933f280
--- /dev/null
+++ b/pkg/covermerger/provider_web.go
@@ -0,0 +1,65 @@
+// Copyright 2024 syzkaller project authors. All rights reserved.
+// Use of this source code is governed by Apache 2 LICENSE that can be found in the LICENSE file.
+
+package covermerger
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+)
+
+type webGit struct {
+}
+
+func (mr *webGit) GetFileVersions(c *Config, targetFilePath string, rbcs []RepoBranchCommit,
+) (fileVersions, error) {
+	res := make(fileVersions)
+	for _, rbc := range rbcs {
+		fileBytes, err := loadFile(targetFilePath, rbc.Repo, rbc.Commit)
+		// It is ok if some file doesn't exist. It means we have repo FS diff.
+		if err == errFileNotFound {
+			continue
+		}
+		if err != nil {
+			return nil, fmt.Errorf("failed to loadFile: %w", err)
+		}
+		res[rbc] = string(fileBytes)
+	}
+	return res, nil
+}
+
+var errFileNotFound = errors.New("file not found")
+
+func loadFile(filePath, repo, commit string) ([]byte, error) {
+	uri := fmt.Sprintf("%s/plain/%s?id=%s", repo, filePath, commit)
+	u, err := url.Parse(uri)
+	if err != nil {
+		return nil, fmt.Errorf("failed to parse %v: %w", uri, err)
+	}
+	u.Scheme = "https"
+	uri = u.String()
+	res, err := http.Get(uri)
+	if err != nil {
+		return nil, fmt.Errorf("failed to http.Get: %w", err)
+	}
+	defer res.Body.Close()
+
+	if res.StatusCode == 404 {
+		return nil, errFileNotFound
+	}
+	if res.StatusCode != 200 {
+		return nil, fmt.Errorf("error: status %d getting %s", res.StatusCode, uri)
+	}
+	body, err := io.ReadAll(res.Body)
+	if err != nil {
+		return nil, fmt.Errorf("failed to io.ReadAll from body: %w", err)
+	}
+	return body, nil
+}
+
+func MakeWebGit() FileVersProvider {
+	return &webGit{}
+}