Compare commits

..

11 Commits

Author SHA1 Message Date
Alejandro Mery cbd3f705f5 basic: GoStringer [WIP] 1 year ago
Alejandro Mery c9f206c9aa Unmarshal: WIP 1 year ago
Alejandro Mery 9912146d21 Decoder: WIP 1 year ago
Alejandro Mery 71ab4a58c1 build-sys: use local darvaza.org/core [DO-NOT-MERGE] 1 year ago
Alejandro Mery 8e838c3566 build-sys: use local asciigoat.org/core [DO-NOT-MERGE] 1 year ago
Alejandro Mery ff0c7d1b9f vscode: add Subname to the dictionary 1 year ago
Alejandro Mery fabd192e3d vscode: add unescapes to the dictionary 1 year ago
Alejandro Mery 37f3efebfb vscode: add asciigoat to the dictionary 1 year ago
Alejandro Mery 506fff8725 Merge branch 'pr-amery-basic' into next-amery 1 year ago
Alejandro Mery d75b2dbc78 basic: rename and document queue related methods 1 year ago
Alejandro Mery 46ba96d6b4 basic: refactor error handling 1 year ago
  1. 7
      .vscode/settings.json
  2. 15
      basic/error.go
  3. 51
      basic/token.go
  4. 18
      basic/write.go
  5. 49
      decoder.go
  6. 5
      go.mod
  7. 2
      go.sum
  8. 88
      parser/comma_array.go
  9. 43
      parser/error.go
  10. 14
      parser/lexer_runes.go
  11. 8
      parser/parser.go
  12. 97
      parser/text_quoted.go
  13. 15
      unmarshal.go

7
.vscode/settings.json vendored

@ -0,0 +1,7 @@
{
"cSpell.words": [
"asciigoat",
"Subname",
"unescapes"
]
}

15
basic/error.go

@ -4,19 +4,28 @@ import (
"errors" "errors"
"asciigoat.org/core/lexer" "asciigoat.org/core/lexer"
"asciigoat.org/ini/parser"
) )
var ( var (
errInvalidToken = errors.New("invalid token") errInvalidToken = errors.New("invalid token")
) )
func newError(pos lexer.Position, content, hint string, err error) *lexer.Error {
return &lexer.Error{
Line: pos.Line,
Column: pos.Column,
Content: content,
Hint: hint,
Err: err,
}
}
func newErrInvalidToken(t *token) *lexer.Error { func newErrInvalidToken(t *token) *lexer.Error {
return parser.NewError(t.pos, t.value, "", errInvalidToken) return newError(t.pos, t.value, "", errInvalidToken)
} }
func (dec *decoder) OnError(pos lexer.Position, content string, err error) error { func (dec *decoder) OnError(pos lexer.Position, content string, err error) error {
err = parser.NewError(pos, content, "", err) err = newError(pos, content, "", err)
dec.executeFinal() dec.executeFinal()
return err return err
} }

51
basic/token.go

@ -32,21 +32,21 @@ func (dec *decoder) executeFinal() {
func (dec *decoder) execute(typ parser.TokenType) { func (dec *decoder) execute(typ parser.TokenType) {
switch typ { switch typ {
case parser.TokenSectionEnd: case parser.TokenSectionEnd:
name1, ok1 := dec.getValue(1, parser.TokenSectionName) name1, ok1 := dec.queueValue(1, parser.TokenSectionName)
if ok1 { if ok1 {
name2, ok2 := dec.getValue(2, parser.TokenSectionSubname) name2, ok2 := dec.queueValue(2, parser.TokenSectionSubname)
dec.addSection(name1, name2, ok2) dec.addSection(name1, name2, ok2)
} }
dec.reset() dec.queueReset()
case parser.TokenFieldValue: case parser.TokenFieldValue:
key, _ := dec.getValue(0, parser.TokenFieldKey) key, _ := dec.queueValue(0, parser.TokenFieldKey)
value, _ := dec.getValue(1, parser.TokenFieldValue) value, _ := dec.queueValue(1, parser.TokenFieldValue)
dec.addField(key, value) dec.addField(key, value)
dec.reset() dec.queueReset()
} }
} }
@ -82,7 +82,8 @@ func (dec *decoder) addField(key, value string) {
} }
} }
func (dec *decoder) getValue(idx int, typ parser.TokenType) (string, bool) { // queueValue extracts the value of element on the queue if the type matches.
func (dec *decoder) queueValue(idx int, typ parser.TokenType) (string, bool) {
switch { switch {
case idx < 0 || idx >= len(dec.queue): case idx < 0 || idx >= len(dec.queue):
// out of range // out of range
@ -95,40 +96,48 @@ func (dec *decoder) getValue(idx int, typ parser.TokenType) (string, bool) {
} }
} }
func (dec *decoder) reset() { // queueReset removes all tokens from the queue
func (dec *decoder) queueReset() {
dec.queue = dec.queue[:0] dec.queue = dec.queue[:0]
} }
func (dec *decoder) depth(depth int) bool { // queueDepth confirms the current depth of the queue
func (dec *decoder) queueDepth(depth int) bool {
return len(dec.queue) == depth return len(dec.queue) == depth
} }
func (dec *decoder) depthAfter(depth int, typ parser.TokenType) bool { // queueDepthType confirms the current depth of the queue and the type of the last
_, ok := dec.getValue(depth-1, typ) // element.
if ok { func (dec *decoder) queueDepthType(depth int, typ parser.TokenType) bool {
return len(dec.queue) == depth if dec.queueDepth(depth) {
return dec.queueType(depth-1, typ)
} }
return false return false
} }
// queueType tells if the specified element on the queue is of the required type.
func (dec *decoder) queueType(idx int, typ parser.TokenType) bool {
_, ok := dec.queueValue(idx, typ)
return ok
}
func (dec *decoder) typeOK(typ parser.TokenType) bool { func (dec *decoder) typeOK(typ parser.TokenType) bool {
switch typ { switch typ {
case parser.TokenSectionStart, parser.TokenFieldKey: case parser.TokenSectionStart, parser.TokenFieldKey:
// first token only // first token only
return dec.depth(0) return dec.queueDepth(0)
case parser.TokenSectionName: case parser.TokenSectionName:
// right after TokenSectionStart // right after TokenSectionStart
return dec.depthAfter(1, parser.TokenSectionStart) return dec.queueDepthType(1, parser.TokenSectionStart)
case parser.TokenSectionSubname: case parser.TokenSectionSubname:
// right after TokenSectionName // right after TokenSectionName
return dec.depthAfter(2, parser.TokenSectionName) return dec.queueDepthType(2, parser.TokenSectionName)
case parser.TokenSectionEnd: case parser.TokenSectionEnd:
// only on a section with name // only on a section with name
_, ok := dec.getValue(1, parser.TokenSectionName) return dec.queueType(1, parser.TokenSectionName)
return ok
case parser.TokenFieldValue: case parser.TokenFieldValue:
// right after a TokenFieldKey // right after a TokenFieldKey
return dec.depthAfter(1, parser.TokenFieldKey) return dec.queueDepthType(1, parser.TokenFieldKey)
default: default:
// never // never
return false return false
@ -149,6 +158,8 @@ func (dec *decoder) OnToken(pos lexer.Position, typ parser.TokenType, value stri
return nil return nil
default: default:
// unacceptable // unacceptable
return newErrInvalidToken(t) err := newErrInvalidToken(t)
dec.executeFinal()
return err
} }
} }

18
basic/write.go

@ -53,6 +53,12 @@ func (field Field) String() string {
return buf.String() return buf.String()
} }
// GoString generates a string output for "%#v"
func (*Field) GoString() string {
var buf bytes.Buffer
return buf.String()
}
func writeSectionToBuffer(w *bytes.Buffer, sec *Section, nl string) int { func writeSectionToBuffer(w *bytes.Buffer, sec *Section, nl string) int {
var written, n int var written, n int
@ -90,6 +96,12 @@ func (sec *Section) String() string {
return buf.String() return buf.String()
} }
// GoString generates a string output for "%#v"
func (*Section) GoString() string {
var buf bytes.Buffer
return buf.String()
}
// WriteTo writes a INI representation of the document // WriteTo writes a INI representation of the document
// onto the provided writer. // onto the provided writer.
func (doc *Document) WriteTo(w io.Writer) (int64, error) { func (doc *Document) WriteTo(w io.Writer) (int64, error) {
@ -102,3 +114,9 @@ func (doc *Document) String() string {
buf := doc.AsBuffer(WriteNewLine) buf := doc.AsBuffer(WriteNewLine)
return buf.String() return buf.String()
} }
// GoString generates a string output for "%#v"
func (Document) GoString() string {
var buf bytes.Buffer
return buf.String()
}

49
decoder.go

@ -0,0 +1,49 @@
package ini
import (
"bytes"
"io"
"strings"
"asciigoat.org/core"
"asciigoat.org/ini/parser"
)
// Decoder ...
type Decoder struct {
io.Closer
p *parser.Parser
}
// Decode ...
func (dec *Decoder) Decode() error {
defer dec.Close()
return dec.p.Run()
}
// NewDecoder creates a Decoder over the provided [io.Reader]
func NewDecoder(r io.Reader) *Decoder {
rc := core.NewReadCloser(r)
switch {
case rc == nil:
return nil
default:
dec := &Decoder{
p: parser.NewParser(rc),
Closer: rc,
}
return dec
}
}
// NewDecoderBytes creates a Decoder over a provided bytes array
func NewDecoderBytes(b []byte) *Decoder {
return NewDecoder(bytes.NewReader(b))
}
// NewDecoderString creates a Decoder over a provided string of data
func NewDecoderString(s string) *Decoder {
return NewDecoder(strings.NewReader(s))
}

5
go.mod

@ -2,6 +2,11 @@ module asciigoat.org/ini
go 1.19 go 1.19
replace (
asciigoat.org/core => ../core
darvaza.org/core => ../../darvaza.org/core
)
require ( require (
asciigoat.org/core v0.3.9 asciigoat.org/core v0.3.9
github.com/mgechev/revive v1.3.3 github.com/mgechev/revive v1.3.3

2
go.sum

@ -1,5 +1,3 @@
asciigoat.org/core v0.3.9 h1:hgDDz4ecm3ZvehX++m8A/IzAt+B5oDPiRtxatzfUHPQ=
asciigoat.org/core v0.3.9/go.mod h1:CAaHwyw8MpAq4a1MYtN2dxJrsK+hmIdW50OndaQZYPI=
github.com/BurntSushi/toml v1.3.2 h1:o7IhLm0Msx3BaB+n3Ag7L8EVlByGnpq14C4YWiu/gL8= github.com/BurntSushi/toml v1.3.2 h1:o7IhLm0Msx3BaB+n3Ag7L8EVlByGnpq14C4YWiu/gL8=
github.com/BurntSushi/toml v1.3.2/go.mod h1:CxXYINrC8qIiEnFrOxCa7Jy5BFHlXnUU2pbicEuybxQ= github.com/BurntSushi/toml v1.3.2/go.mod h1:CxXYINrC8qIiEnFrOxCa7Jy5BFHlXnUU2pbicEuybxQ=
github.com/chavacava/garif v0.1.0 h1:2JHa3hbYf5D9dsgseMKAmc/MZ109otzgNFk5s87H9Pc= github.com/chavacava/garif v0.1.0 h1:2JHa3hbYf5D9dsgseMKAmc/MZ109otzgNFk5s87H9Pc=

88
parser/comma_array.go

@ -1,88 +0,0 @@
package parser
import (
"strings"
"asciigoat.org/core/lexer"
)
type commaArrayParser struct {
TextParser
out []string
}
func (p *commaArrayParser) lexStart() (lexer.StateFn, error) {
for {
r, _, err := p.ReadRune()
switch {
case err != nil:
// EOF
return nil, err
case r == RuneQuotes:
// Quoted Value
return p.lexQuotedString, nil
case IsNewLine(r):
// new lines are acceptable when parsing a string for
// comma delimited arrays. but make sure we discard it
// complete
p.UnreadRune()
p.AcceptNewLine()
p.Discard()
case lexer.IsSpace(r):
// discard whitespace outside quotes
p.Discard()
default:
p.UnreadRune()
return p.lexWord, nil
}
}
}
func (p *commaArrayParser) lexWord() (lexer.StateFn, error) {
for {
r, _, err := p.ReadRune()
switch {
case err != nil:
// done. store what we got and move on
_, s := p.Emit()
p.out = append(p.out, s)
return nil, err
case r == ',':
// done
_, s := p.Emit()
// remove comma, trim and append to output
s = strings.TrimRightFunc(s[:len(s)-1], IsSpace)
p.out = append(p.out, s)
return p.lexStart, nil
}
}
}
func (p *commaArrayParser) lexQuotedString() (lexer.StateFn, error) {
s, err := lexQuotedString(&p.TextParser)
if err != nil {
return nil, err
}
p.Discard()
p.out = append(p.out, s)
return p.lexStart, nil
}
func (p *commaArrayParser) Run() ([]string, error) {
err := lexer.Run(p.lexStart)
return p.out, err
}
// SplitCommaArray splits comma separated strings, removing whitespace
// and respecting quoted literals.
func SplitCommaArray(s string) ([]string, error) {
if s != "" {
var p commaArrayParser
p.InitString(s)
return p.Run()
}
return nil, nil
}

43
parser/error.go

@ -1,43 +0,0 @@
package parser
import (
"io/fs"
"asciigoat.org/core/lexer"
)
// NewError creates a lexer.Error using a lexer.Position
func NewError(pos lexer.Position, content, hint string, err error) *lexer.Error {
return &lexer.Error{
Line: pos.Line,
Column: pos.Column,
Content: content,
Hint: hint,
Err: err,
}
}
// ErrPlusPosition returns a copy of the given [lexer.Error]
// offsetting the Line/Column information.
func ErrPlusPosition(pos lexer.Position, e *lexer.Error) *lexer.Error {
pos.Add(lexer.Position{
Line: e.Line,
Column: e.Column,
})
return NewError(pos, e.Content, e.Hint, e.Err)
}
// NewErrIncompleteQuotedString returns a [lexer.Error]
// indicating the quoted string being parsed wasn't correctly
// terminated
func NewErrIncompleteQuotedString(p *TextParser) *lexer.Error {
return newErrIncomplete(p, "incomplete quoted string")
}
func newErrIncomplete(p *TextParser, hint string) *lexer.Error {
pos, s := p.Emit()
pos.Add(GetPositionalLength(s))
return NewError(pos, s, hint, fs.ErrInvalid)
}

14
parser/lexer_runes.go

@ -7,13 +7,11 @@ import (
) )
const ( const (
RuneComment = ';' // RuneComment is the standard INI comment character RuneComment = ';' // RuneComment is the standard dosini comment character
RuneCommentExtra = '#' // RuneCommentExtra is UNIX shell's comment character RuneCommentExtra = '#' // RuneCommentExtra is UNIX shell's comment character
RuneSectionStart = '[' // RuneSectionStart indicates the start of a section declaration RuneSectionStart = '[' // RuneSectionStart indicates the start of a section declaration
RuneSectionEnd = ']' // RuneSectionEnd indicates the end of a section declaration RuneSectionEnd = ']' // RuneSectionEnd indiciates the end of a section declaration
RuneFieldEqual = '=' // RuneFieldEqual separates field keys from their values RuneFieldEqual = '=' // RuneFieldEqual separates field keys from their values
RuneQuotes = '"' // RuneQuotes indicates the start and end of a quoted value
RuneEscape = '\\' // RuneEscape indicates the next rune is escaped
) )
var ( var (
@ -29,8 +27,6 @@ var (
RuneSectionStart, RuneSectionStart,
RuneSectionEnd, RuneSectionEnd,
RuneFieldEqual, RuneFieldEqual,
RuneQuotes,
RuneEscape,
}) })
) )

8
parser/parser.go

@ -31,7 +31,13 @@ func defaultOnToken(pos lexer.Position, typ TokenType, value string) error {
func defaultOnError(pos lexer.Position, content string, err error) error { func defaultOnError(pos lexer.Position, content string, err error) error {
log.Printf("%s:%v:%v: %q: %s", "error", pos.Line, pos.Column, content, err) log.Printf("%s:%v:%v: %q: %s", "error", pos.Line, pos.Column, content, err)
return NewError(pos, content, "", err) return lexer.Error{
Line: pos.Line,
Column: pos.Column,
Content: content,
Err: err,
}
} }
func (p *Parser) setDefaults() { func (p *Parser) setDefaults() {

97
parser/text_quoted.go

@ -1,97 +0,0 @@
package parser
import (
"strings"
"asciigoat.org/core/lexer"
)
// AcceptQuotedString consumes a quoted string from the source
// and returns it unquoted and unescaped
func (p *TextParser) AcceptQuotedString() (string, bool, error) {
r, _, err := p.ReadRune()
switch {
case err != nil:
// nothing here
return "", false, err
case r != RuneQuotes:
// not for us
p.UnreadRune()
return "", false, nil
default:
// let's roll
s, err := lexQuotedString(p)
switch {
case err != nil:
// bad quoted string
return "", false, err
default:
// success
return s, true, nil
}
}
}
func lexQuotedString(p *TextParser) (string, *lexer.Error) {
for {
r, _, err := p.ReadRune()
switch {
case err != nil:
// incomplete
return "", NewErrIncompleteQuotedString(p)
case r == RuneQuotes:
// end, remove quotes and process escaped characters
return lexReturnUnescapedQuotedString(p)
case r == RuneEscape:
// escaped, take another
_, _, err := p.ReadRune()
if err != nil {
// incomplete
return "", NewErrIncompleteQuotedString(p)
}
case IsNewLine(r):
// new lines within quoted values are acceptable
p.UnreadRune()
p.AcceptNewLine()
default:
// continue
}
}
}
func lexReturnUnescapedQuotedString(p *TextParser) (string, *lexer.Error) {
// remove quotes
s := p.String()
l := len(s)
s = s[1 : l-1]
if strings.ContainsRune(s, RuneEscape) {
// TODO: implement unescaping
err := NewError(p.Position(), s, "escaped characters", lexer.ErrNotImplemented)
return "", err
}
return s, nil
}
// Unquoted removes quotes the content and unescapes the content
func Unquoted(s string) (string, error) {
var p TextParser
if s == "" {
return "", nil
}
p.InitString(s)
unquoted, ok, err := p.AcceptQuotedString()
switch {
case err != nil:
// bad string
return "", err
case ok:
// success
return unquoted, nil
default:
// not quoted
return s, nil
}
}

15
unmarshal.go

@ -0,0 +1,15 @@
package ini
import "io"
// ReadInto ...
func ReadInto(v any, r io.Reader) error {
dec := NewDecoder(r)
return dec.Unmarshal(v)
}
// Unmarshal ...
func (dec *Decoder) Unmarshal(any) error {
return dec.p.Run()
}
Loading…
Cancel
Save