Validate migration files (go-gitea#18203)

JSON Schema validation for data used by Gitea during migrations

Discussion at https://forum.forgefriends.org/t/common-json-schema-for-repository-information/563

Co-authored-by: Loïc Dachary <loic@dachary.org>

.gitignore

@@ -36,6 +36,8 @@ _testmain.go
 coverage.all
 cpu.out
 
+/modules/migration/bindata.go
+/modules/migration/bindata.go.hash
 /modules/options/bindata.go
 /modules/options/bindata.go.hash
 /modules/public/bindata.go

cmd/restore_repo.go

@@ -43,6 +43,10 @@ var CmdRestoreRepository = cli.Command{
 			Usage: `Which items will be restored, one or more units should be separated as comma.
 wiki, issues, labels, releases, release_assets, milestones, pull_requests, comments are allowed. Empty means all units.`,
 		},
+		cli.BoolFlag{
+			Name:  "validation",
+			Usage: "Sanity check the content of the files before trying to load them",
+		},
 	},
 }
 
@@ -58,6 +62,7 @@ func runRestoreRepository(c *cli.Context) error {
 		c.String("owner_name"),
 		c.String("repo_name"),
 		c.StringSlice("units"),
+		c.Bool("validation"),
 	)
 	if statusCode == http.StatusOK {
 		return nil

go.mod

@@ -97,6 +97,7 @@ require (
 	github.com/quasoft/websspi v1.0.0
 	github.com/rs/xid v1.3.0 // indirect
 	github.com/russross/blackfriday/v2 v2.1.0 // indirect
+	github.com/santhosh-tekuri/jsonschema/v5 v5.0.0 // indirect
 	github.com/sergi/go-diff v1.2.0
 	github.com/shurcooL/httpfs v0.0.0-20190707220628-8d4bc4ba7749 // indirect
 	github.com/shurcooL/vfsgen v0.0.0-20200824052919-0d455de96546

go.sum

@@ -1039,6 +1039,8 @@ github.com/russross/blackfriday/v2 v2.1.0 h1:JIOH55/0cWyOuilr9/qlrm0BSXldqnqwMsf
 github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/ryanuber/columnize v0.0.0-20160712163229-9b3edd62028f/go.mod h1:sm1tb6uqfes/u+d4ooFouqFdy9/2g9QGwK3SQygK0Ts=
 github.com/samuel/go-zookeeper v0.0.0-20190923202752-2cc03de413da/go.mod h1:gi+0XIa01GRL2eRQVjQkKGqKF3SF9vZR/HnPullcV2E=
+github.com/santhosh-tekuri/jsonschema/v5 v5.0.0 h1:TToq11gyfNlrMFZiYujSekIsPd9AmsA2Bj/iv+s4JHE=
+github.com/santhosh-tekuri/jsonschema/v5 v5.0.0/go.mod h1:FKdcjfQW6rpZSnxxUvEA5H/cDPdvJ/SZJQLWWXWGrZ0=
 github.com/sean-/seed v0.0.0-20170313163322-e2103e2c3529/go.mod h1:DxrIzT+xaE7yg65j358z/aeFdxmN0P9QXhEzd20vsDc=
 github.com/sergi/go-diff v1.0.0/go.mod h1:0CfEIISq7TuYL3j771MWULgwwjU+GofnZX9QAmXWZgo=
 github.com/sergi/go-diff v1.1.0/go.mod h1:STckp+ISIX8hZLjrqAeVduY0gWCT9IjLuqbuNXdaHfM=

integrations/dump_restore_test.go

@@ -81,7 +81,7 @@ func TestDumpRestore(t *testing.T) {
 		//
 
 		newreponame := "restoredrepo"
-		err = migrations.RestoreRepository(ctx, d, repo.OwnerName, newreponame, []string{"labels", "milestones", "issues", "comments"})
+		err = migrations.RestoreRepository(ctx, d, repo.OwnerName, newreponame, []string{"labels", "milestones", "issues", "comments"}, false)
 		assert.NoError(t, err)
 
 		newrepo := unittest.AssertExistsAndLoadBean(t, &repo_model.Repository{Name: newreponame}).(*repo_model.Repository)

modules/migration/file_format.go

@@ -0,0 +1,112 @@
+// Copyright 2022 The Gitea Authors. All rights reserved.
+// Use of this source code is governed by a MIT-style
+// license that can be found in the LICENSE file.
+
+package migration
+
+import (
+	"fmt"
+	"os"
+	"strings"
+
+	"code.gitea.io/gitea/modules/json"
+	"code.gitea.io/gitea/modules/log"
+
+	"github.com/santhosh-tekuri/jsonschema/v5"
+	"gopkg.in/yaml.v2"
+)
+
+// Load project data from file, with optional validation
+func Load(filename string, data interface{}, validation bool) error {
+	isJSON := strings.HasSuffix(filename, ".json")
+
+	bs, err := os.ReadFile(filename)
+	if err != nil {
+		return err
+	}
+
+	if validation {
+		err := validate(bs, data, isJSON)
+		if err != nil {
+			return err
+		}
+	}
+	return unmarshal(bs, data, isJSON)
+}
+
+func unmarshal(bs []byte, data interface{}, isJSON bool) error {
+	if isJSON {
+		return json.Unmarshal(bs, data)
+	}
+	return yaml.Unmarshal(bs, data)
+}
+
+func getSchema(filename string) (*jsonschema.Schema, error) {
+	c := jsonschema.NewCompiler()
+	c.LoadURL = openSchema
+	return c.Compile(filename)
+}
+
+func validate(bs []byte, datatype interface{}, isJSON bool) error {
+	var v interface{}
+	err := unmarshal(bs, &v, isJSON)
+	if err != nil {
+		return err
+	}
+	if !isJSON {
+		v, err = toStringKeys(v)
+		if err != nil {
+			return err
+		}
+	}
+
+	var schemaFilename string
+	switch datatype := datatype.(type) {
+	case *[]*Issue:
+		schemaFilename = "issue.json"
+	case *[]*Milestone:
+		schemaFilename = "milestone.json"
+	default:
+		return fmt.Errorf("file_format:validate: %T has not a validation implemented", datatype)
+	}
+
+	sch, err := getSchema(schemaFilename)
+	if err != nil {
+		return err
+	}
+	err = sch.Validate(v)
+	if err != nil {
+		log.Error("migration validation with %s failed for\n%s", schemaFilename, string(bs))
+	}
+	return err
+}
+
+func toStringKeys(val interface{}) (interface{}, error) {
+	var err error
+	switch val := val.(type) {
+	case map[interface{}]interface{}:
+		m := make(map[string]interface{})
+		for k, v := range val {
+			k, ok := k.(string)
+			if !ok {
+				return nil, fmt.Errorf("found non-string key %T %s", k, k)
+			}
+			m[k], err = toStringKeys(v)
+			if err != nil {
+				return nil, err
+			}
+		}
+		return m, nil
+	case []interface{}:
+		l := make([]interface{}, len(val))
+		for i, v := range val {
+			l[i], err = toStringKeys(v)
+			if err != nil {
+				return nil, err
+			}
+		}
+		return l, nil
+	default:
+		return val, nil
+	}
+}

modules/migration/file_format_test.go

@@ -0,0 +1,39 @@
+// Copyright 2022 The Gitea Authors. All rights reserved.
+// Use of this source code is governed by a MIT-style
+// license that can be found in the LICENSE file.
+
+package migration
+
+import (
+	"strings"
+	"testing"
+
+	"github.com/santhosh-tekuri/jsonschema/v5"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestMigrationJSON_IssueOK(t *testing.T) {
+	issues := make([]*Issue, 0, 10)
+	err := Load("file_format_testdata/issue_a.json", &issues, true)
+	assert.NoError(t, err)
+	err = Load("file_format_testdata/issue_a.yml", &issues, true)
+	assert.NoError(t, err)
+}
+
+func TestMigrationJSON_IssueFail(t *testing.T) {
+	issues := make([]*Issue, 0, 10)
+	err := Load("file_format_testdata/issue_b.json", &issues, true)
+	if _, ok := err.(*jsonschema.ValidationError); ok {
+		errors := strings.Split(err.(*jsonschema.ValidationError).GoString(), "\n")
+		assert.Contains(t, errors[1], "missing properties")
+		assert.Contains(t, errors[1], "poster_id")
+	} else {
+		t.Fatalf("got: type %T with value %s, want: *jsonschema.ValidationError", err, err)
+	}
+}
+
+func TestMigrationJSON_MilestoneOK(t *testing.T) {
+	milestones := make([]*Milestone, 0, 10)
+	err := Load("file_format_testdata/milestones.json", &milestones, true)
+	assert.NoError(t, err)
+}

modules/migration/file_format_testdata/issue_a.json

@@ -0,0 +1,14 @@
+[
+  {
+      "number": 1,
+      "poster_id": 1,
+      "poster_name": "name_a",
+      "title": "title_a",
+      "content": "content_a",
+      "state": "closed",
+      "is_locked": false,
+      "created": "1985-04-12T23:20:50.52Z",
+      "updated": "1986-04-12T23:20:50.52Z",
+      "closed": "1987-04-12T23:20:50.52Z"
+  }
+]

modules/migration/file_format_testdata/issue_a.yml

@@ -0,0 +1,10 @@
+- number: 1
+  poster_id: 1
+  poster_name: name_a
+  title: title_a
+  content: content_a
+  state: closed
+  is_locked: false
+  created: 2021-05-27T15:24:13+02:00
+  updated: 2021-11-11T10:52:45+01:00
+  closed: 2021-11-11T10:52:45+01:00

modules/migration/file_format_testdata/issue_b.json

@@ -0,0 +1,5 @@
+[
+  {
+      "number": 1
+  }
+]

modules/migration/file_format_testdata/milestones.json

@@ -0,0 +1,20 @@
+[
+  {
+      "title": "title_a",
+      "description": "description_a",
+      "deadline": "1988-04-12T23:20:50.52Z",
+      "created": "1985-04-12T23:20:50.52Z",
+      "updated": "1986-04-12T23:20:50.52Z",
+      "closed": "1987-04-12T23:20:50.52Z",
+      "state": "closed"
+  },
+  {
+      "title": "title_b",
+      "description": "description_b",
+      "deadline": "1998-04-12T23:20:50.52Z",
+      "created": "1995-04-12T23:20:50.52Z",
+      "updated": "1996-04-12T23:20:50.52Z",
+      "closed": null,
+      "state": "open"
+  }
+]

modules/migration/issue.go

@@ -28,21 +28,21 @@ func (c BasicIssueContext) ForeignID() int64 {
 
 // Issue is a standard issue information
 type Issue struct {
-	Number      int64
-	PosterID    int64  `yaml:"poster_id"`
-	PosterName  string `yaml:"poster_name"`
-	PosterEmail string `yaml:"poster_email"`
-	Title       string
-	Content     string
-	Ref         string
-	Milestone   string
-	State       string // closed, open
-	IsLocked    bool   `yaml:"is_locked"`
-	Created     time.Time
-	Updated     time.Time
-	Closed      *time.Time
-	Labels      []*Label
-	Reactions   []*Reaction
-	Assignees   []string
+	Number      int64        `json:"number"`
+	PosterID    int64        `yaml:"poster_id" json:"poster_id"`
+	PosterName  string       `yaml:"poster_name" json:"poster_name"`
+	PosterEmail string       `yaml:"poster_email" json:"poster_email"`
+	Title       string       `json:"title"`
+	Content     string       `json:"content"`
+	Ref         string       `json:"ref"`
+	Milestone   string       `json:"milestone"`
+	State       string       `json:"state"` // closed, open
+	IsLocked    bool         `yaml:"is_locked" json:"is_locked"`
+	Created     time.Time    `json:"created"`
+	Updated     time.Time    `json:"updated"`
+	Closed      *time.Time   `json:"closed"`
+	Labels      []*Label     `json:"labels"`
+	Reactions   []*Reaction  `json:"reactions"`
+	Assignees   []string     `json:"assignees"`
 	Context     IssueContext `yaml:"-"`
 }

modules/migration/label.go

@@ -7,7 +7,7 @@ package migration
 
 // Label defines a standard label information
 type Label struct {
-	Name        string
-	Color       string
-	Description string
+	Name        string `json:"name"`
+	Color       string `json:"color"`
+	Description string `json:"description"`
 }

modules/migration/milestone.go

@@ -9,11 +9,11 @@ import "time"
 
 // Milestone defines a standard milestone
 type Milestone struct {
-	Title       string
-	Description string
-	Deadline    *time.Time
-	Created     time.Time
-	Updated     *time.Time
-	Closed      *time.Time
-	State       string // open, closed
+	Title       string     `json:"title"`
+	Description string     `json:"description"`
+	Deadline    *time.Time `json:"deadline"`
+	Created     time.Time  `json:"created"`
+	Updated     *time.Time `json:"updated"`
+	Closed      *time.Time `json:"closed"`
+	State       string     `json:"state"` // open, closed
 }

modules/migration/reaction.go

@@ -6,7 +6,7 @@ package migration
 
 // Reaction represents a reaction to an issue/pr/comment.
 type Reaction struct {
-	UserID   int64  `yaml:"user_id"`
-	UserName string `yaml:"user_name"`
-	Content  string
+	UserID   int64  `yaml:"user_id" json:"user_id"`
+	UserName string `yaml:"user_name" json:"user_name"`
+	Content  string `json:"content"`
 }