output

Documentation

Index

• func CSVHeader(withVerification bool, sep rune) string
• type Meta
• type Name
• type OddsDatum
• type Output
  • func TokensToOutput(ts []token.TokenSN, text []rune, version string, cfg config.Config) Output
  • func (o *Output) Format(f gnfmt.Format) string
  • func (o *Output) MergeVerification(v map[string]vlib.Verification, dur float32)
  • func (o *Output) UniqueNameStrings() []string

Functions

func CSVHeader

func CSVHeader(withVerification bool, sep rune) string

CSVHeader returns the header string for CSV output format.

Types

type Meta

type Meta struct {
	// InputFile is the name of the source file.
	InputFile string `json:"inputFile,omitempty"`

	// TextExtractionSec is the time spent on converting the file
	// into UTF8-encoded text.
	TextExtractionSec float32 `json:"textExtractSec,omitempty"`

	// NameFindingSec is the time spent on name-finding.
	NameFindingSec float32 `json:"nameFindingSec"`

	// NameVerifSec is the time spent on name-verification.
	NameVerifSec float32 `json:"nameVerifSec,omitempty"`

	// TotalSec is time spent for the whole process
	TotalSec float32 `json:"totalSec,omitempty"`

	// Date represents time when output was generated.
	Date time.Time `json:"date"`

	// FinderVersion the version of gnfinder.
	FinderVersion string `json:"gnfinderVersion"`

	// WithBayes use of bayes during name-finding
	WithBayes bool `json:"withBayes"`

	// WithBytesOffset names position in bytes instead of UTF-8 chars.
	WithBytesOffset bool

	// WithOddsAdjustment to adjust prior odds according to the dencity of
	// scientific names in the text.
	WithOddsAdjustment bool `json:"withOddsAdjustment"`

	// WithVerification is true if results are checked by verification service.
	WithVerification bool `json:"withVerification"`

	// WordsAround shows the number of tokens preserved before and after
	// a name-string candidate.
	WordsAround int `json:"wordsAround"`

	// Language inside name-finding algorithm
	Language string `json:"language"`

	// LanguageDetected automatically for the text
	LanguageDetected string `json:"languageDetected,omitempty"`

	// LanguageForced by language option
	DetectLanguage bool `json:"detectLanguage"`

	// TotalTokens is a number of 'normalized' words in the text
	TotalTokens int `json:"totalWords"`

	// TotalNameCandidates is a number of words that might be a start of
	// a scientific name
	TotalNameCandidates int `json:"totalCandidates"`

	// TotalNames is a number of scientific names found
	TotalNames int `json:"totalNames"`

	// CurrentName (optional) is the index of the names array that designates a
	// "position of a cursor". It is used by programs like gntagger that allow
	// to work on the list of found names interactively.
	CurrentName int `json:"currentIndex,omitempty"`
}

Meta contains meta-information of name-finding result.

type Name

type Name struct {
	// Cardinality depicts number of elements in a name. 0 - Cannot determine
	// cardinality, 1 - Uninomial, 2 - Binomial, 3 - Trinomial.
	Cardinality int `json:"cardinality"`
	// Verbatim shows name the way it was in the text.
	Verbatim string `json:"verbatim,omitempty"`
	// Name is a normalized version of a name.
	Name string `json:"name"`
	// Odds show a probability that name detection was correct.
	Odds float64 `json:"-"`
	// OddsLog10 show a Log10 of Odds.
	OddsLog10 float64 `json:"oddsLog10,omitempty"`
	// OddsDetails descibes how Odds were calculated.
	OddsDetails token.OddsDetails `json:"oddsDetails,omitempty"`
	// OffsetStart is a start of a name on a page.
	OffsetStart int `json:"start"`
	// OffsetEnd is the end of the name on a page.
	OffsetEnd int `json:"end"`
	// AnnotNomen is a nomenclatural annotation for new species or combination.
	AnnotNomen string `json:"annotationNomen,omitempty"`
	// AnnotNomenType is normalized nomenclatural annotation.
	AnnotNomenType string `json:"annotationNomenType,omitempty"`
	// Annotation is a placeholder to add more information about name.
	Annotation string `json:"annotation,omitempty"`
	// WordsBefore are words that happened before the name.
	WordsBefore []string `json:"wordsBefore,omitempty"`
	// WordsAfter are words that happened right after the name.
	WordsAfter []string `json:"wordsAfter,omitempty"`
	// Verification gives results of verification process of the name.
	Verification *vlib.Verification `json:"verification,omitempty"`
}

Name represents one found name.

type OddsDatum

type OddsDatum struct {
	Name bool
	Odds float64
}

OddsDatum is a simplified version of a name, that stores boolean decision (Name/NotName), and corresponding odds of the name.

type Output

type Output struct {
	Meta      `json:"metadata"`
	InputText string `json:"inputText,omitempty"`
	Names     []Name `json:"names"`
}

Output type is the result of name-finding.

func TokensToOutput

func TokensToOutput(
	ts []token.TokenSN,
	text []rune,
	version string,
	cfg config.Config) Output

TokensToOutput takes tagged tokens and assembles output out of them.

func (*Output) Format

func (o *Output) Format(f gnfmt.Format) string

func (*Output) MergeVerification

func (o *Output) MergeVerification(
	v map[string]vlib.Verification,
	dur float32,
)

MergeVerification takes a map with verified names and incorporates into output.

func (*Output) UniqueNameStrings

func (o *Output) UniqueNameStrings() []string

UniqueNameStrings takes a list of names, and returns a list of unique name-strings