github.com/pgavlin/text@v0.0.0-20240419000839-8438d0a47805/reader.go (about)

     1  // Copyright 2009 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  package text
     6  
     7  import (
     8  	"errors"
     9  	"io"
    10  
    11  	"github.com/pgavlin/text/internal/bytealg"
    12  	"github.com/pgavlin/text/utf8"
    13  )
    14  
    15  // A Reader implements the io.Reader, io.ReaderAt, io.ByteReader, io.ByteScanner,
    16  // io.RuneReader, io.RuneScanner, io.Seeker, and io.WriterTo interfaces by reading
    17  // from a string.
    18  // The zero value for Reader operates like a Reader of an empty string.
    19  type Reader[S String] struct {
    20  	s        S
    21  	i        int64 // current reading index
    22  	prevRune int   // index of previous rune; or < 0
    23  }
    24  
    25  // Len returns the number of bytes of the unread portion of the
    26  // string.
    27  func (r *Reader[S]) Len() int {
    28  	if r.i >= int64(len(r.s)) {
    29  		return 0
    30  	}
    31  	return int(int64(len(r.s)) - r.i)
    32  }
    33  
    34  // Size returns the original length of the underlying string.
    35  // Size is the number of bytes available for reading via ReadAt.
    36  // The returned value is always the same and is not affected by calls
    37  // to any other method.
    38  func (r *Reader[S]) Size() int64 { return int64(len(r.s)) }
    39  
    40  // Read implements the io.Reader interface.
    41  func (r *Reader[S]) Read(b []byte) (n int, err error) {
    42  	if r.i >= int64(len(r.s)) {
    43  		return 0, io.EOF
    44  	}
    45  	r.prevRune = -1
    46  	n = copy(b, r.s[r.i:])
    47  	r.i += int64(n)
    48  	return
    49  }
    50  
    51  // ReadAt implements the io.ReaderAt interface.
    52  func (r *Reader[S]) ReadAt(b []byte, off int64) (n int, err error) {
    53  	// cannot modify state - see io.ReaderAt
    54  	if off < 0 {
    55  		return 0, errors.New("text.Reader.ReadAt: negative offset")
    56  	}
    57  	if off >= int64(len(r.s)) {
    58  		return 0, io.EOF
    59  	}
    60  	n = copy(b, r.s[off:])
    61  	if n < len(b) {
    62  		err = io.EOF
    63  	}
    64  	return
    65  }
    66  
    67  // ReadByte implements the io.ByteReader interface.
    68  func (r *Reader[S]) ReadByte() (byte, error) {
    69  	r.prevRune = -1
    70  	if r.i >= int64(len(r.s)) {
    71  		return 0, io.EOF
    72  	}
    73  	b := r.s[r.i]
    74  	r.i++
    75  	return b, nil
    76  }
    77  
    78  // UnreadByte implements the io.ByteScanner interface.
    79  func (r *Reader[S]) UnreadByte() error {
    80  	if r.i <= 0 {
    81  		return errors.New("text.Reader.UnreadByte: at beginning of string")
    82  	}
    83  	r.prevRune = -1
    84  	r.i--
    85  	return nil
    86  }
    87  
    88  // ReadRune implements the io.RuneReader interface.
    89  func (r *Reader[S]) ReadRune() (ch rune, size int, err error) {
    90  	if r.i >= int64(len(r.s)) {
    91  		r.prevRune = -1
    92  		return 0, 0, io.EOF
    93  	}
    94  	r.prevRune = int(r.i)
    95  	if c := r.s[r.i]; c < utf8.RuneSelf {
    96  		r.i++
    97  		return rune(c), 1, nil
    98  	}
    99  	ch, size = utf8.DecodeRune(r.s[r.i:])
   100  	r.i += int64(size)
   101  	return
   102  }
   103  
   104  // UnreadRune implements the io.RuneScanner interface.
   105  func (r *Reader[S]) UnreadRune() error {
   106  	if r.i <= 0 {
   107  		return errors.New("text.Reader.UnreadRune: at beginning of string")
   108  	}
   109  	if r.prevRune < 0 {
   110  		return errors.New("text.Reader.UnreadRune: previous operation was not ReadRune")
   111  	}
   112  	r.i = int64(r.prevRune)
   113  	r.prevRune = -1
   114  	return nil
   115  }
   116  
   117  // Seek implements the io.Seeker interface.
   118  func (r *Reader[S]) Seek(offset int64, whence int) (int64, error) {
   119  	r.prevRune = -1
   120  	var abs int64
   121  	switch whence {
   122  	case io.SeekStart:
   123  		abs = offset
   124  	case io.SeekCurrent:
   125  		abs = r.i + offset
   126  	case io.SeekEnd:
   127  		abs = int64(len(r.s)) + offset
   128  	default:
   129  		return 0, errors.New("text.Reader.Seek: invalid whence")
   130  	}
   131  	if abs < 0 {
   132  		return 0, errors.New("text.Reader.Seek: negative position")
   133  	}
   134  	r.i = abs
   135  	return abs, nil
   136  }
   137  
   138  // WriteTo implements the io.WriterTo interface.
   139  func (r *Reader[S]) WriteTo(w io.Writer) (n int64, err error) {
   140  	r.prevRune = -1
   141  	if r.i >= int64(len(r.s)) {
   142  		return 0, nil
   143  	}
   144  	s := r.s[r.i:]
   145  	m, err := io.WriteString(w, bytealg.AsString(s))
   146  	if m > len(s) {
   147  		panic("text.Reader.WriteTo: invalid WriteString count")
   148  	}
   149  	r.i += int64(m)
   150  	n = int64(m)
   151  	if m != len(s) && err == nil {
   152  		err = io.ErrShortWrite
   153  	}
   154  	return
   155  }
   156  
   157  // Reset resets the Reader to be reading from s.
   158  func (r *Reader[S]) Reset(s S) { *r = Reader[S]{s, 0, -1} }
   159  
   160  // NewReader returns a new Reader reading from s.
   161  // It is similar to bytes.NewBufferString but more efficient and read-only.
   162  func NewReader[S String](s S) *Reader[S] { return &Reader[S]{s, 0, -1} }