parser

package
v0.0.0-...-d54656b Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Apr 7, 2026 License: MIT Imports: 23 Imported by: 0

Documentation

Index

Constants

View Source 
const (
	FileTypeUnknown = "unknown"
	FileTypeTXT     = "txt"
	FileTypeMD      = "md"
	FileTypeCSV     = "csv"
	FileTypeHTML    = "html"
	FileTypeJSON    = "json"
	FileTypeYAML    = "yaml"
	FileTypeYML     = "yml"
	FileTypeEML     = "eml"
	FileTypeRTF     = "rtf"
	FileTypePDF     = "pdf"
	FileTypePNG     = "png"
	FileTypeJPG     = "jpg"
	FileTypeJPEG    = "jpeg"
	FileTypeDOC     = "doc"
	FileTypeDOCX    = "docx"
	FileTypePPTX    = "pptx"
	FileTypeXLSX    = "xlsx"
)

Variables

View Source 
var (
	ErrUnsupportedFileType = errors.New("unsupported file type")
	ErrEmptyInput          = errors.New("empty input")
)

Functions

func DetectFileType 

func DetectFileType(req *ParseRequest) string

Types

type CSVParser 

type CSVParser struct{}

func (*CSVParser) Parse 

func (p *CSVParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*CSVParser) Provider 

func (p *CSVParser) Provider() string

func (*CSVParser) SupportedTypes 

func (p *CSVParser) SupportedTypes() []string

type DOCParser 

type DOCParser struct{}

func (*DOCParser) Parse 

func (p *DOCParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*DOCParser) Provider 

func (p *DOCParser) Provider() string

func (*DOCParser) SupportedTypes 

func (p *DOCParser) SupportedTypes() []string

type DOCXParser 

type DOCXParser struct{}

func (*DOCXParser) Parse 

func (p *DOCXParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*DOCXParser) Provider 

func (p *DOCXParser) Provider() string

func (*DOCXParser) SupportedTypes 

func (p *DOCXParser) SupportedTypes() []string

type EMLParser 

type EMLParser struct{}

func (*EMLParser) Parse 

func (p *EMLParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*EMLParser) Provider 

func (p *EMLParser) Provider() string

func (*EMLParser) SupportedTypes 

func (p *EMLParser) SupportedTypes() []string

type HTMLParser 

type HTMLParser struct{}

func (*HTMLParser) Parse 

func (p *HTMLParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*HTMLParser) Provider 

func (p *HTMLParser) Provider() string

func (*HTMLParser) SupportedTypes 

func (p *HTMLParser) SupportedTypes() []string

type JSONParser 

type JSONParser struct{}

func (*JSONParser) Parse 

func (p *JSONParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*JSONParser) Provider 

func (p *JSONParser) Provider() string

func (*JSONParser) SupportedTypes 

func (p *JSONParser) SupportedTypes() []string

type OCRParser 

type OCRParser struct {
	Language string
}

func (*OCRParser) Parse 

func (p *OCRParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*OCRParser) Provider 

func (p *OCRParser) Provider() string

func (*OCRParser) SupportedTypes 

func (p *OCRParser) SupportedTypes() []string

type PDFParser 

type PDFParser struct{}

func (*PDFParser) Parse 

func (p *PDFParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*PDFParser) Provider 

func (p *PDFParser) Provider() string

func (*PDFParser) SupportedTypes 

func (p *PDFParser) SupportedTypes() []string

type PPTXParser 

type PPTXParser struct{}

func (*PPTXParser) Parse 

func (p *PPTXParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*PPTXParser) Provider 

func (p *PPTXParser) Provider() string

func (*PPTXParser) SupportedTypes 

func (p *PPTXParser) SupportedTypes() []string

type ParseOptions 

type ParseOptions struct {
	MaxTextLength      int
	IncludeTables      bool
	IncludeHidden      bool
	PreserveLineBreaks bool
}

type ParseRequest 

type ParseRequest struct {
	FileType    string
	FileName    string
	Path        string
	ContentType string
	Content     []byte
	Reader      io.Reader
	Metadata    map[string]any
}

type ParseResult 

type ParseResult struct {
	FileType string
	FileName string
	Text     string
	Sections []Section
	Metadata map[string]any
	ParsedAt time.Time
}

func ParseAuto 

func ParseAuto(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func ParseBytes 

func ParseBytes(ctx context.Context, fileName string, content []byte, opts *ParseOptions) (*ParseResult, error)

func ParsePath 

func ParsePath(ctx context.Context, path string, opts *ParseOptions) (*ParseResult, error)

type Parser 

type Parser interface {
	Provider() string

	SupportedTypes() []string

	Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)
}

type RTFParser 

type RTFParser struct{}

func (*RTFParser) Parse 

func (p *RTFParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*RTFParser) Provider 

func (p *RTFParser) Provider() string

func (*RTFParser) SupportedTypes 

func (p *RTFParser) SupportedTypes() []string

type Router 

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

func DefaultRouter 

func DefaultRouter() *Router

func NewRouter 

func NewRouter(parsers ...Parser) *Router

func (*Router) Parse 

func (r *Router) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*Router) Register 

func (r *Router) Register(p Parser) error

type Section 

type Section struct {
	Type     SectionType
	Index    int
	Title    string
	Text     string
	Metadata map[string]any
}

type SectionType 

type SectionType string
const (
	SectionTypeUnknown  SectionType = "unknown"
	SectionTypeDocument SectionType = "document"
	SectionTypePage     SectionType = "page"
	SectionTypeSheet    SectionType = "sheet"
	SectionTypeSlide    SectionType = "slide"
)

type TXTParser 

type TXTParser struct{}

func (*TXTParser) Parse 

func (p *TXTParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*TXTParser) Provider 

func (p *TXTParser) Provider() string

func (*TXTParser) SupportedTypes 

func (p *TXTParser) SupportedTypes() []string

type XLSXParser 

type XLSXParser struct{}

func (*XLSXParser) Parse 

func (p *XLSXParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*XLSXParser) Provider 

func (p *XLSXParser) Provider() string

func (*XLSXParser) SupportedTypes 

func (p *XLSXParser) SupportedTypes() []string

type YAMLParser 

type YAMLParser struct{}

func (*YAMLParser) Parse 

func (p *YAMLParser) Parse(ctx context.Context, req *ParseRequest, opts *ParseOptions) (*ParseResult, error)

func (*YAMLParser) Provider 

func (p *YAMLParser) Provider() string

func (*YAMLParser) SupportedTypes 

func (p *YAMLParser) SupportedTypes() []string

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.