gdocs

Documentation

Index

• Constants
• func GetEntities(ctx context.Context, client *language.Client, doc *docs.Document) ([]*languagepb.Entity, error)
• func ReadText(doc *docs.Document) (string, error)
• type Client
  • func NewClient(c *http.Client, log logr.Logger) (*Client, error)
  • func (c *Client) Search(query string, driveId string, corpora string, resultFunc ResultFunc) error
• type DriveSearch
• type FakeSearch
  • func (f *FakeSearch) Search(query string, driveId string, corpora string, resultFunc ResultFunc) error
• type GoogleDocUri
  • func ParseGoogleDocUri(u string) (*GoogleDocUri, error)
• type HyperLink
  • func GetAllLinks(doc *docs.Document) ([]*HyperLink, error)
• type Indexer
  • func NewIndexer(searcher DriveSearch, docsService *docs.Service, store *datastore.Datastore, ...) (*Indexer, error)
  • func (idx *Indexer) Index(driveId string) error
  • func (idx *Indexer) IndexDocument(docId string) error
  • func (idx *Indexer) ProcessDoc(r *datastore.DocReference)
  • func (idx *Indexer) ProcessDocLinks(r *datastore.DocReference, d *docs.Document) error
  • func (idx *Indexer) ProcessEntities(r *datastore.DocReference, d *docs.Document) error
• type IndexerOption
  • func IndexerWithHTTPClient(c *http.Client) IndexerOption
  • func IndexerWithLogger(log logr.Logger) IndexerOption
• type QueryStats
• type ResultFunc
  • func NewStatsBuilder(s *QueryStats) (ResultFunc, error)

Constants

const (
	// DocumentMimeType is the mime type for Google Documents.
	DocumentMimeType = "application/vnd.google-apps.document"
)

const (
	GoogleDocsHost = "docs.google.com"
)

Functions

func GetEntities

func GetEntities(ctx context.Context, client *language.Client, doc *docs.Document) ([]*languagepb.Entity, error)

GetEntities gets the entities from the document.

N.B. The current implementation doesn't keep track of

func ReadText

func ReadText(doc *docs.Document) (string, error)

ReadText reads all the text from the provided document. It is based on https://developers.google.com/docs/api/samples/extract-text#python.

TODO(https://github.com/jlewi/p22h/issues/1): Linearize text so as to preserve positioning.

Types

type Client

type Client struct {
	// contains filtered or unexported fields
}

Client is a high level client for interacting with gdrive

func NewClient

func NewClient(c *http.Client, log logr.Logger) (*Client, error)

func (*Client) Search

func (c *Client) Search(query string, driveId string, corpora string, resultFunc ResultFunc) error

Search runs the provided search query.

type DriveSearch

type DriveSearch interface {
	Search(query string, driveId string, corpora string, resultFunc ResultFunc) error
}

type FakeSearch

type FakeSearch struct {
	Docs []*drive.File
}

FakeSearch implements the search interface for an in memory set of drive documents. FakeSearch is intended for testing.

func (*FakeSearch) Search

func (f *FakeSearch) Search(query string, driveId string, corpora string, resultFunc ResultFunc) error

type GoogleDocUri

type GoogleDocUri struct {
	ID      string
	Heading string
}

func ParseGoogleDocUri

func ParseGoogleDocUri(u string) (*GoogleDocUri, error)

ParseGoogleDocUri parses a google document URI Return nil if not a googledocument.

type HyperLink

type HyperLink struct {
	Url        string
	Text       string
	StartIndex int64
	EndIndex   int64
}

func GetAllLinks

func GetAllLinks(doc *docs.Document) ([]*HyperLink, error)

GetAllLinks gets all the links from the document.

type Indexer

type Indexer struct {
	// contains filtered or unexported fields
}

Indexer indexes Google Drive.

func NewIndexer

func NewIndexer(searcher DriveSearch, docsService *docs.Service, store *datastore.Datastore, nlpClient *language.Client, log logr.Logger, opts ...IndexerOption) (*Indexer, error)

NewIndexer creates a new indexer

func (*Indexer) Index

func (idx *Indexer) Index(driveId string) error

TODO(jeremy): Should rename this IndexFolder or IndexDrive

func (*Indexer) IndexDocument

func (idx *Indexer) IndexDocument(docId string) error

IndexDocument indexes a specific document

func (*Indexer) ProcessDoc

func (idx *Indexer) ProcessDoc(r *datastore.DocReference)

ProcessDoc processes the referenced doc.

TODO(jeremy): This function should really return an error. Originally it wasn't returning an error because it was only being called from Index which just continued but now its being called from IndexDocument and we should propogate the error to that.

func (*Indexer) ProcessDocLinks

func (idx *Indexer) ProcessDocLinks(r *datastore.DocReference, d *docs.Document) error

ProcessDocLinks processes all the docs for the doc referenced by r and represented by d.

func (*Indexer) ProcessEntities

func (idx *Indexer) ProcessEntities(r *datastore.DocReference, d *docs.Document) error

ProcessEntities gets all the entities in the document

type IndexerOption

type IndexerOption func(*Indexer)

func IndexerWithHTTPClient

func IndexerWithHTTPClient(c *http.Client) IndexerOption

func IndexerWithLogger

func IndexerWithLogger(log logr.Logger) IndexerOption

type QueryStats

type QueryStats struct {
	Count int64
	// Size in bytes
	Size   int64
	ByType map[string]float64
}

QueryStats contains statistics about the results of a search query

type ResultFunc

type ResultFunc func(file *drive.File) error

ResultFunc is invoked by search to process each result A non nil error causes result processing to stop.

func NewStatsBuilder

func NewStatsBuilder(s *QueryStats) (ResultFunc, error)

NewStatsBuilder returns a ResultFunc that will aggregate statistics.