adding support for public and private git providers

README.md

@@ -12,6 +12,42 @@ The Devfile Parser library is a Golang module that:
 2. writes to the devfile.yaml with the updated data.
 3. generates Kubernetes objects for the various devfile resources.
 4. defines util functions for the devfile.
+5. downloads resources from a parent devfile if specified in the devfile.yaml
+
+## Private repository support
+
+Tokens are required to be set in the following cases:
+1. parsing a devfile from a private repository
+2. parsing a devfile containing a parent devfile from a private repository [1]
+3. parsing a devfile from a private repository containing a parent devfile from a public repository [2]
+
+Set the token for the repository:
+```go
+parser.ParserArgs{
+	...
+	// URL must point to a devfile.yaml
+	URL: <url-to-devfile-on-supported-git-provider-repo>/devfile.yaml
+	Token: <repo-personal-access-token>
+	...
+}
+```
+Note: The url must also be set with a supported git provider repo url.
+
+Minimum token scope required:
+1. GitHub: Read access to code
+2. GitLab: Read repository
+3. Bitbucket: Read repository
+
+Note: To select token scopes for GitHub, a fine-grained token is required.
+
+For more information about personal access tokens:
+1. [GitHub docs](https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/creating-a-personal-access-token)
+2. [GitLab docs](https://docs.gitlab.com/ee/user/profile/personal_access_tokens.html#create-a-personal-access-token)
+3. [Bitbucket docs](https://support.atlassian.com/bitbucket-cloud/docs/repository-access-tokens/)
+
+[1] Currently, this works under the assumption that the token can authenticate the devfile and the parent devfile; both devfiles are in the same repository.
+
+[2] In this scenario, the token will be used to authenticate the main devfile.
 
 ## Usage
 
@@ -35,7 +71,6 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
    devfile, variableWarning, err := devfilePkg.ParseDevfileAndValidate(parserArgs)
    ```
 
-
 2. To override the HTTP request and response timeouts for a devfile with a parent reference from a registry URL, specify the HTTPTimeout value in the parser arguments
    ```go
       // specify the timeout in seconds  
@@ -45,7 +80,6 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
 	  }
    ```
 
-
 3. To get specific content from devfile
    ```go
    // To get all the components from the devfile
@@ -77,7 +111,7 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
 		},
    })
    ```
-   
+
 4. To get the Kubernetes objects from the devfile, visit [generators.go source file](pkg/devfile/generator/generators.go)
    ```go
     // To get a slice of Kubernetes containers of type corev1.Container from the devfile component containers
@@ -94,7 +128,7 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
 	}
 	deployment := generator.GetDeployment(deployParams)
    ```
-   
+
 5. To update devfile content
    ```go
    // To update an existing component in devfile object
@@ -131,20 +165,19 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
    ```go
    // If the devfile object has been created with devfile path already set, can simply call WriteYamlDevfile to writes the devfile
    err := devfile.WriteYamlDevfile()
-
-
+
    // To write to a devfile from scratch
    // create a new DevfileData with a specific devfile version
    devfileData, err := data.NewDevfileData(devfileVersion)
 
    // set schema version
    devfileData.SetSchemaVersion(devfileVersion)
-   
+
    // add devfile content use library APIs
    devfileData.AddComponents([]v1.Component{...})
    devfileData.AddCommands([]v1.Commands{...})
    ......
-   
+
    // create a new DevfileCtx
    ctx := devfileCtx.NewDevfileCtx(devfilePath)
    err = ctx.SetAbsPath()
@@ -154,10 +187,11 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
 		Ctx:  ctx,
 		Data: devfileData,
    }
-    
+
    // write to the devfile on disk
    err = devfile.WriteYamlDevfile()
    ```
+
 7. To parse the outerloop Kubernetes/OpenShift component's uri or inline content, call the read and parse functions
    ```go
    // Read the YAML content
@@ -166,6 +200,7 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
    // Get the Kubernetes resources
    resources, err := ParseKubernetesYaml(values)
    ```
+
 8. By default, the parser will set all unset boolean properties to their spec defined default values.  Clients can override this behaviour by specifiying the parser argument `SetBooleanDefaults` to false
    ```go
    setDefaults := false
@@ -174,6 +209,15 @@ The function documentation can be accessed via [pkg.go.dev](https://pkg.go.dev/g
    }
    ```
 
+9. When parsing a devfile that contains a parent reference, if the parent uri is a supported git provider repo url with the correct personal access token, all resources from the parent git repo excluding the parent devfile.yaml will be downloaded to the location of the devfile being parsed. **Note: The URL must point to a devfile.yaml**
+   ```yaml
+   schemaVersion: 2.2.0
+   ...
+   parent:
+      uri: <uri-to-parent-devfile>/devfile.yaml
+   ...
+   ```
+
 ## Projects using devfile/library
 
 The following projects are consuming this library as a Golang dependency

pkg/devfile/parser/context/content.go

@@ -1,5 +1,5 @@
 //
-// Copyright 2022 Red Hat, Inc.
+// Copyright 2022-2023 Red Hat, Inc.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -69,6 +69,9 @@
 	if d.url != "" {
 		// set the client identifier for telemetry
 		params := util.HTTPRequestParams{URL: d.url, TelemetryClientName: util.TelemetryClientName}
+		if d.token != "" {
+			params.Token = d.token
+		}
 		data, err = util.DownloadInMemory(params)
 		if err != nil {
 			return errors.Wrap(err, "error getting devfile info from url")

pkg/devfile/parser/context/context.go

@@ -1,5 +1,5 @@
 //
-// Copyright 2022 Red Hat, Inc.
+// Copyright 2022-2023 Red Hat, Inc.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -45,13 +45,16 @@ type DevfileCtx struct {
 	// devfile json schema
 	jsonSchema string
 
-	//url path of the devfile
+	// url path of the devfile
 	url string
 
+	// token is a personal access token used with a private git repo URL
+	token string
+
 	// filesystem for devfile
 	fs filesystem.Filesystem
 
-	// devfile kubernetes components has been coverted from uri to inlined in memory
+	// devfile kubernetes components has been converted from uri to inlined in memory
 	convertUriToInlined bool
 }
 
@@ -150,6 +153,16 @@ func (d *DevfileCtx) GetURL() string {
 	return d.url
 }
 
+// GetToken func returns current devfile token
+func (d *DevfileCtx) GetToken() string {
+	return d.token
+}
+
+// SetToken sets the token for the devfile
+func (d *DevfileCtx) SetToken(token string) {
+	d.token = token
+}
+
 // SetAbsPath sets absolute file path for devfile
 func (d *DevfileCtx) SetAbsPath() (err error) {
 	// Set devfile absolute path

pkg/devfile/parser/context/context_test.go

@@ -1,5 +1,5 @@
 //
-// Copyright 2022 Red Hat, Inc.
+// Copyright 2022-2023 Red Hat, Inc.
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -83,6 +83,20 @@ func TestPopulateFromInvalidURL(t *testing.T) {
 	})
 }
 
+func TestNewURLDevfileCtx(t *testing.T) {
+	var (
+		token = "fake-token"
+		url   = "https://github.com/devfile/registry/blob/main/stacks/go/2.0.0/devfile.yaml"
+	)
+	{
+		d := NewURLDevfileCtx(url)
+		assert.Equal(t, "https://github.com/devfile/registry/blob/main/stacks/go/2.0.0/devfile.yaml", d.GetURL())
+		assert.Equal(t, "", d.GetToken())
+		d.SetToken(token)
+		assert.Equal(t, "fake-token", d.GetToken())
+	}
+}
+
 func invalidJsonRawContent200() []byte {
 	return []byte(InvalidDevfileContent)
 }

pkg/devfile/parser/parse.go

@@ -19,6 +19,8 @@
 	"context"
 	"encoding/json"
 	"fmt"
+	"github.com/devfile/library/v2/pkg/git"
+	"github.com/hashicorp/go-multierror"
 	"io/ioutil"
 	"net/url"
 	"os"
@@ -46,6 +48,57 @@
 	"github.com/pkg/errors"
 )
 
+// downloadGitRepoResources is exposed as a global variable for the purpose of running mock tests
+var downloadGitRepoResources = func(url string, destDir string, httpTimeout *int, token string) error {
+	var returnedErr error
+
+	gitUrl, err := git.NewGitUrlWithURL(url)
+	if err != nil {
+		return err
+	}
+
+	if gitUrl.IsGitProviderRepo() {
+		if !gitUrl.IsFile || gitUrl.Revision == "" || !strings.Contains(gitUrl.Path, OutputDevfileYamlPath) {
+			return fmt.Errorf("error getting devfile from url: failed to retrieve %s", url)
+		}
+
+		stackDir, err := os.MkdirTemp("", fmt.Sprintf("git-resources"))
+		if err != nil {
+			return fmt.Errorf("failed to create dir: %s, error: %v", stackDir, err)
+		}
+
+		defer func(path string) {
+			err := os.RemoveAll(path)
+			if err != nil {
+				returnedErr = multierror.Append(returnedErr, err)
+			}
+		}(stackDir)
+
+		if !gitUrl.IsPublic(httpTimeout) {
+			err = gitUrl.SetToken(token, httpTimeout)
+			if err != nil {
+				returnedErr = multierror.Append(returnedErr, err)
+				return returnedErr
+			}
+		}
+
+		err = gitUrl.CloneGitRepo(stackDir)
+		if err != nil {
+			returnedErr = multierror.Append(returnedErr, err)
+			return returnedErr
+		}
+
+		dir := path.Dir(path.Join(stackDir, gitUrl.Path))
+		err = git.CopyAllDirFiles(dir, destDir)
+		if err != nil {
+			returnedErr = multierror.Append(returnedErr, err)
+			return returnedErr
+		}
+	}
+
+	return nil
+}
+
 // ParseDevfile func validates the devfile integrity.
 // Creates devfile context and runtime objects
 func parseDevfile(d DevfileObj, resolveCtx *resolutionContextTree, tool resolverTools, flattenedDevfile bool) (DevfileObj, error) {
@@ -97,6 +150,8 @@
 	// RegistryURLs is a list of registry hosts which parser should pull parent devfile from.
 	// If registryUrl is defined in devfile, this list will be ignored.
 	RegistryURLs []string
+	// Token is a GitHub, GitLab, or Bitbucket personal access token used with a private git repo URL
+	Token string
 	// DefaultNamespace is the default namespace to use
 	// If namespace is defined under devfile's parent kubernetes object, this namespace will be ignored.
 	DefaultNamespace string
@@ -129,6 +184,10 @@
 		return d, errors.Wrap(err, "the devfile source is not provided")
 	}
 
+	if args.Token != "" {
+		d.Ctx.SetToken(args.Token)
+	}
+
 	tool := resolverTools{
 		defaultNamespace: args.DefaultNamespace,
 		registryURLs:     args.RegistryURLs,
@@ -431,17 +490,16 @@
 			return DevfileObj{}, fmt.Errorf("failed to resolve parent uri, devfile context is missing absolute url and path to devfile. %s", resolveImportReference(importReference))
 		}
 
+		token := curDevfileCtx.GetToken()
 		d.Ctx = devfileCtx.NewURLDevfileCtx(newUri)
-		if strings.Contains(newUri, "raw.githubusercontent.com") {
-			urlComponents, err := util.GetGitUrlComponentsFromRaw(newUri)
-			if err != nil {
-				return DevfileObj{}, err
-			}
-			destDir := path.Dir(curDevfileCtx.GetAbsPath())
-			err = getResourcesFromGit(urlComponents, destDir)
-			if err != nil {
-				return DevfileObj{}, err
-			}
+		if token != "" {
+			d.Ctx.SetToken(token)
+		}
+
+		destDir := path.Dir(curDevfileCtx.GetAbsPath())
+		err = downloadGitRepoResources(newUri, destDir, tool.httpTimeout, token)
+		if err != nil {
+			return DevfileObj{}, err
 		}
 	}
 	importReference.Uri = newUri
@@ -450,27 +508,6 @@
 	return populateAndParseDevfile(d, newResolveCtx, tool, true)
 }
 
-func getResourcesFromGit(gitUrlComponents map[string]string, destDir string) error {
-	stackDir, err := ioutil.TempDir(os.TempDir(), fmt.Sprintf("git-resources"))
-	if err != nil {
-		return fmt.Errorf("failed to create dir: %s, error: %v", stackDir, err)
-	}
-	defer os.RemoveAll(stackDir)
-
-	err = util.CloneGitRepo(gitUrlComponents, stackDir)
-	if err != nil {
-		return err
-	}
-
-	dir := path.Dir(path.Join(stackDir, gitUrlComponents["file"]))
-	err = util.CopyAllDirFiles(dir, destDir)
-	if err != nil {
-		return err
-	}
-
-	return nil
-}
-
 func parseFromRegistry(importReference v1.ImportReference, resolveCtx *resolutionContextTree, tool resolverTools) (d DevfileObj, err error) {
 	id := importReference.Id
 	registryURL := importReference.RegistryUrl
@@ -839,6 +876,9 @@
 			newUri = uri
 		}
 		params := util.HTTPRequestParams{URL: newUri}
+		if d.GetToken() != "" {
+			params.Token = d.GetToken()
+		}
 		data, err = util.DownloadInMemory(params)
 		if err != nil {
 			return nil, errors.Wrapf(err, "error getting kubernetes resources definition information")