databricks-cli/libs/filer/workspace_files_extensions_...

346 lines
12 KiB
Go
Raw Normal View History

package filer
import (
"context"
"errors"
"fmt"
"io"
"io/fs"
"net/http"
"path"
"strings"
"github.com/databricks/cli/libs/log"
"github.com/databricks/cli/libs/notebook"
"github.com/databricks/databricks-sdk-go"
"github.com/databricks/databricks-sdk-go/apierr"
"github.com/databricks/databricks-sdk-go/client"
"github.com/databricks/databricks-sdk-go/marshal"
"github.com/databricks/databricks-sdk-go/service/workspace"
)
type workspaceFilesExtensionsClient struct {
workspaceClient *databricks.WorkspaceClient
apiClient *client.DatabricksClient
wsfs Filer
root string
}
var extensionsToLanguages = map[string]workspace.Language{
".py": workspace.LanguagePython,
".r": workspace.LanguageR,
".scala": workspace.LanguageScala,
".sql": workspace.LanguageSql,
".ipynb": workspace.LanguagePython,
}
// workspaceFileStatus defines a custom response body for the "/api/2.0/workspace/get-status" API.
// The "repos_export_format" field is not exposed by the SDK.
type workspaceFileStatus struct {
*workspace.ObjectInfo
// The export format of the notebook. This is not exposed by the SDK.
ReposExportFormat workspace.ExportFormat `json:"repos_export_format,omitempty"`
// Name of the file to be used in any API calls made using the workspace files
// filer. For notebooks this path does not include the extension.
nameForWorkspaceAPI string
}
// A custom unmarsaller for the workspaceFileStatus struct. This is needed because
// workspaceFileStatus embeds the workspace.ObjectInfo which itself has a custom
// unmarshaller.
// If a custom unmarshaller is not provided extra fields like ReposExportFormat
// will not have values set.
func (s *workspaceFileStatus) UnmarshalJSON(b []byte) error {
return marshal.Unmarshal(b, s)
}
func (s *workspaceFileStatus) MarshalJSON() ([]byte, error) {
return marshal.Marshal(s)
}
func (w *workspaceFilesExtensionsClient) stat(ctx context.Context, name string) (*workspaceFileStatus, error) {
stat := &workspaceFileStatus{
nameForWorkspaceAPI: name,
}
// Perform bespoke API call because "return_export_info" is not exposed by the SDK.
// We need "repos_export_format" to determine if the file is a py or a ipynb notebook.
// This is not exposed by the SDK so we need to make a direct API call.
err := w.apiClient.Do(
ctx,
http.MethodGet,
"/api/2.0/workspace/get-status",
nil,
map[string]string{
"path": path.Join(w.root, name),
"return_export_info": "true",
},
stat,
)
if err != nil {
// If we got an API error we deal with it below.
var aerr *apierr.APIError
if !errors.As(err, &aerr) {
return nil, err
}
// This API returns a 404 if the specified path does not exist.
if aerr.StatusCode == http.StatusNotFound {
return nil, FileDoesNotExistError{path.Join(w.root, name)}
}
}
return stat, err
}
// This function returns the stat for the provided notebook. The stat object itself contains the path
// with the extension since it is meant to be used in the context of a fs.FileInfo.
func (w *workspaceFilesExtensionsClient) getNotebookStatByNameWithExt(ctx context.Context, name string) (*workspaceFileStatus, error) {
ext := path.Ext(name)
nameWithoutExt := strings.TrimSuffix(name, ext)
// File name does not have an extension associated with Databricks notebooks, return early.
if _, ok := extensionsToLanguages[ext]; !ok {
return nil, nil
}
// If the file could be a notebook, check if it is and has the correct language.
stat, err := w.stat(ctx, nameWithoutExt)
if err != nil {
// If the file does not exist, return early.
if errors.As(err, &FileDoesNotExistError{}) {
return nil, nil
}
log.Debugf(ctx, "attempting to determine if %s could be a notebook. Failed to fetch the status of object at %s: %s", name, path.Join(w.root, nameWithoutExt), err)
return nil, err
}
// Not a notebook. Return early.
if stat.ObjectType != workspace.ObjectTypeNotebook {
log.Debugf(ctx, "attempting to determine if %s could be a notebook. Found an object at %s but it is not a notebook. It is a %s.", name, path.Join(w.root, nameWithoutExt), stat.ObjectType)
return nil, nil
}
// Not the correct language. Return early.
if stat.Language != extensionsToLanguages[ext] {
log.Debugf(ctx, "attempting to determine if %s could be a notebook. Found a notebook at %s but it is not of the correct language. Expected %s but found %s.", name, path.Join(w.root, nameWithoutExt), extensionsToLanguages[ext], stat.Language)
return nil, nil
}
// When the extension is .py we expect the export format to be source.
// If it's not, return early.
if ext == ".py" && stat.ReposExportFormat != workspace.ExportFormatSource {
log.Debugf(ctx, "attempting to determine if %s could be a notebook. Found a notebook at %s but it is not exported as a source notebook. Its export format is %s.", name, path.Join(w.root, nameWithoutExt), stat.ReposExportFormat)
return nil, nil
}
// When the extension is .ipynb we expect the export format to be Jupyter.
// If it's not, return early.
if ext == ".ipynb" && stat.ReposExportFormat != workspace.ExportFormatJupyter {
log.Debugf(ctx, "attempting to determine if %s could be a notebook. Found a notebook at %s but it is not exported as a Jupyter notebook. Its export format is %s.", name, path.Join(w.root, nameWithoutExt), stat.ReposExportFormat)
return nil, nil
}
// Modify the stat object path to include the extension. This stat object will be used
// to return the fs.FileInfo object in the stat method.
stat.Path = stat.Path + ext
return stat, nil
}
func (w *workspaceFilesExtensionsClient) getNotebookStatByNameWithoutExt(ctx context.Context, name string) (*workspaceFileStatus, error) {
stat, err := w.stat(ctx, name)
if err != nil {
return nil, err
}
// We expect this internal function to only be called from [ReadDir] when we are sure
// that the object is a notebook. Thus, this should never happen.
if stat.ObjectType != workspace.ObjectTypeNotebook {
return nil, fmt.Errorf("expected object at %s to be a notebook but it is a %s", path.Join(w.root, name), stat.ObjectType)
}
// Get the extension for the notebook.
ext := notebook.GetExtensionByLanguage(stat.ObjectInfo)
// If the notebook was exported as a Jupyter notebook, the extension should be .ipynb.
if stat.Language == workspace.LanguagePython && stat.ReposExportFormat == workspace.ExportFormatJupyter {
ext = ".ipynb"
}
// Modify the stat object path to include the extension. This stat object will be used
// to return the fs.DirEntry object in the ReadDir method.
stat.Path = stat.Path + ext
return stat, nil
}
type DuplicatePathError struct {
oi1 workspace.ObjectInfo
oi2 workspace.ObjectInfo
commonName string
}
func (e DuplicatePathError) Error() string {
return fmt.Sprintf("failed to read files from the workspace file system. Duplicate paths encountered. Both %s at %s and %s at %s resolve to the same name %s. Changing the name of one of these objects will resolve this issue", e.oi1.ObjectType, e.oi1.Path, e.oi2.ObjectType, e.oi2.Path, e.commonName)
}
// This is a filer for the workspace file system that allows you to pretend the
// workspace file system is a traditional file system. It allows you to list, read, write,
// delete, and stat notebooks (and files in general) in the workspace, using their paths
// with the extension included.
//
// The ReadDir method returns a DuplicatePathError if this traditional file system view is
// not possible. For example, a Python notebook called foo and a Python file called `foo.py`
// would resolve to the same path `foo.py` in a tradition file system.
//
// Users of this filer should be careful when using the Write and Mkdir methods.
// The underlying import API we use to upload notebooks and files returns opaque internal
// errors for namespace clashes (e.g. a file and a notebook or a directory and a notebook).
// Thus users of these methods should be careful to avoid such clashes.
func NewWorkspaceFilesExtensionsClient(w *databricks.WorkspaceClient, root string) (Filer, error) {
apiClient, err := client.New(w.Config)
if err != nil {
return nil, err
}
filer, err := NewWorkspaceFilesClient(w, root)
if err != nil {
return nil, err
}
return &workspaceFilesExtensionsClient{
workspaceClient: w,
apiClient: apiClient,
wsfs: filer,
root: root,
}, nil
}
func (w *workspaceFilesExtensionsClient) ReadDir(ctx context.Context, name string) ([]fs.DirEntry, error) {
entries, err := w.wsfs.ReadDir(ctx, name)
if err != nil {
return nil, err
}
seenPaths := make(map[string]workspace.ObjectInfo)
for i := range entries {
info, err := entries[i].Info()
if err != nil {
return nil, err
}
sysInfo := info.Sys().(workspace.ObjectInfo)
// If the object is a notebook, include an extension in the entry.
if sysInfo.ObjectType == workspace.ObjectTypeNotebook {
stat, err := w.getNotebookStatByNameWithoutExt(ctx, entries[i].Name())
if err != nil {
return nil, err
}
// Replace the entry with the new entry that includes the extension.
entries[i] = wsfsDirEntry{wsfsFileInfo{oi: *stat.ObjectInfo}}
}
// Error if we have seen this path before in the current directory.
// If not seen before, add it to the seen paths.
if _, ok := seenPaths[entries[i].Name()]; ok {
return nil, DuplicatePathError{
oi1: seenPaths[entries[i].Name()],
oi2: sysInfo,
commonName: path.Join(name, entries[i].Name()),
}
}
seenPaths[entries[i].Name()] = sysInfo
}
return entries, nil
}
// Note: The import API returns opaque internal errors for namespace clashes
// (e.g. a file and a notebook or a directory and a notebook). Thus users of this
// method should be careful to avoid such clashes.
func (w *workspaceFilesExtensionsClient) Write(ctx context.Context, name string, reader io.Reader, mode ...WriteMode) error {
return w.wsfs.Write(ctx, name, reader, mode...)
}
// Try to read the file as a regular file. If the file is not found, try to read it as a notebook.
func (w *workspaceFilesExtensionsClient) Read(ctx context.Context, name string) (io.ReadCloser, error) {
r, err := w.wsfs.Read(ctx, name)
// If the file is not found, it might be a notebook.
if errors.As(err, &FileDoesNotExistError{}) {
stat, serr := w.getNotebookStatByNameWithExt(ctx, name)
if serr != nil {
// Unable to stat. Return the stat error.
return nil, serr
}
if stat == nil {
// Not a notebook. Return the original error.
return nil, err
}
// The workspace files filer performs an additional stat call to make sure
// the path is not a directory. We can skip this step since we already have
// the stat object and know that the path is a notebook.
return w.workspaceClient.Workspace.Download(
ctx,
path.Join(w.root, stat.nameForWorkspaceAPI),
workspace.DownloadFormat(stat.ReposExportFormat),
)
}
return r, err
}
// Try to delete the file as a regular file. If the file is not found, try to delete it as a notebook.
func (w *workspaceFilesExtensionsClient) Delete(ctx context.Context, name string, mode ...DeleteMode) error {
err := w.wsfs.Delete(ctx, name, mode...)
// If the file is not found, it might be a notebook.
if errors.As(err, &FileDoesNotExistError{}) {
stat, serr := w.getNotebookStatByNameWithExt(ctx, name)
if serr != nil {
// Unable to stat. Return the stat error.
return serr
}
if stat == nil {
// Not a notebook. Return the original error.
return err
}
return w.wsfs.Delete(ctx, stat.nameForWorkspaceAPI, mode...)
}
return err
}
// Try to stat the file as a regular file. If the file is not found, try to stat it as a notebook.
func (w *workspaceFilesExtensionsClient) Stat(ctx context.Context, name string) (fs.FileInfo, error) {
info, err := w.wsfs.Stat(ctx, name)
// If the file is not found, it might be a notebook.
if errors.As(err, &FileDoesNotExistError{}) {
stat, serr := w.getNotebookStatByNameWithExt(ctx, name)
if serr != nil {
// Unable to stat. Return the stat error.
return nil, serr
}
if stat == nil {
// Not a notebook. Return the original error.
return nil, err
}
return wsfsFileInfo{oi: *stat.ObjectInfo}, nil
}
return info, err
}
// Note: The import API returns opaque internal errors for namespace clashes
// (e.g. a file and a notebook or a directory and a notebook). Thus users of this
// method should be careful to avoid such clashes.
func (w *workspaceFilesExtensionsClient) Mkdir(ctx context.Context, name string) error {
return w.wsfs.Mkdir(ctx, name)
}