feat: implement table parser

plugin/gomark/ast/ast.go

@@ -15,6 +15,7 @@ const (
 	UnorderedListNode
 	TaskListNode
 	MathBlockNode
+	TableNode
 	// Inline nodes.
 	TextNode
 	BoldNode

plugin/gomark/ast/block.go

@@ -194,3 +194,42 @@ func (*MathBlock) Type() NodeType {
 func (n *MathBlock) Restore() string {
 	return fmt.Sprintf("$$\n%s\n$$", n.Content)
 }
+
+type Table struct {
+	BaseBlock
+
+	Header []string
+	// Delimiter is the list of delimiter counts.
+	Delimiter []int
+	Rows      [][]string
+}
+
+func (*Table) Type() NodeType {
+	return TableNode
+}
+
+func (n *Table) Restore() string {
+	var result string
+	for _, header := range n.Header {
+		result += fmt.Sprintf("| %s ", header)
+	}
+	result += "|\n"
+	for _, d := range n.Delimiter {
+		symbol := ""
+		for i := 0; i < d; i++ {
+			symbol += "-"
+		}
+		result += fmt.Sprintf("| %s ", symbol)
+	}
+	result += "|\n"
+	for index, row := range n.Rows {
+		for _, cell := range row {
+			result += fmt.Sprintf("| %s ", cell)
+		}
+		result += "|"
+		if index != len(n.Rows)-1 {
+			result += "\n"
+		}
+	}
+	return result
+}

plugin/gomark/parser/parser.go

@@ -31,6 +31,7 @@ func Parse(tokens []*tokenizer.Token) ([]ast.Node, error) {
 
 var defaultBlockParsers = []BlockParser{
 	NewCodeBlockParser(),
+	NewTableParser(),
 	NewHorizontalRuleParser(),
 	NewHeadingParser(),
 	NewBlockquoteParser(),

plugin/gomark/parser/table.go

+package parser
+
+import (
+	"errors"
+
+	"github.com/usememos/memos/plugin/gomark/ast"
+	"github.com/usememos/memos/plugin/gomark/parser/tokenizer"
+)
+
+type TableParser struct{}
+
+func NewTableParser() *TableParser {
+	return &TableParser{}
+}
+
+func (*TableParser) Match(tokens []*tokenizer.Token) (int, bool) {
+	headerTokens := []*tokenizer.Token{}
+	for _, token := range tokens {
+		if token.Type == tokenizer.Newline {
+			break
+		} else {
+			headerTokens = append(headerTokens, token)
+		}
+	}
+	if len(headerTokens) < 5 || len(tokens) < len(headerTokens)+3 {
+		return 0, false
+	}
+
+	alignTokens := []*tokenizer.Token{}
+	for _, token := range tokens[len(headerTokens)+1:] {
+		if token.Type == tokenizer.Newline {
+			break
+		} else {
+			alignTokens = append(alignTokens, token)
+		}
+	}
+	if len(alignTokens) < 5 || len(tokens) < len(headerTokens)+len(alignTokens)+3 {
+		return 0, false
+	}
+
+	rowTokens := []*tokenizer.Token{}
+	for index, token := range tokens[len(headerTokens)+len(alignTokens)+2:] {
+		temp := len(headerTokens) + len(alignTokens) + 2 + index
+		if token.Type == tokenizer.Newline && temp != len(tokens)-1 && tokens[temp+1].Type != tokenizer.Pipe {
+			break
+		} else {
+			rowTokens = append(rowTokens, token)
+		}
+	}
+	if len(rowTokens) < 5 {
+		return 0, false
+	}
+
+	// Check header.
+	if len(headerTokens) < 5 {
+		return 0, false
+	}
+	headerCells, ok := matchTableCellTokens(headerTokens)
+	if headerCells == 0 || !ok {
+		return 0, false
+	}
+
+	// Check align.
+	if len(alignTokens) < 5 {
+		return 0, false
+	}
+	alignCells, ok := matchTableCellTokens(alignTokens)
+	if alignCells != headerCells || !ok {
+		return 0, false
+	}
+	for _, t := range tokenizer.Split(alignTokens, tokenizer.Pipe) {
+		delimiterTokens := t[1 : len(t)-1]
+		if len(delimiterTokens) < 3 {
+			return 0, false
+		}
+		for _, token := range delimiterTokens {
+			if token.Type != tokenizer.Hyphen {
+				return 0, false
+			}
+		}
+	}
+
+	// Check rows.
+	if len(rowTokens) < 5 {
+		return 0, false
+	}
+	rows := tokenizer.Split(rowTokens, tokenizer.Newline)
+	if len(rows) == 0 {
+		return 0, false
+	}
+	for _, row := range rows {
+		cells, ok := matchTableCellTokens(row)
+		if cells != headerCells || !ok {
+			return 0, false
+		}
+	}
+
+	return len(headerTokens) + len(alignTokens) + len(rowTokens) + 2, true
+}
+
+func (p *TableParser) Parse(tokens []*tokenizer.Token) (ast.Node, error) {
+	size, ok := p.Match(tokens)
+	if size == 0 || !ok {
+		return nil, errors.New("not matched")
+	}
+
+	rawRows := tokenizer.Split(tokens[:size-1], tokenizer.Newline)
+	headerTokens := rawRows[0]
+	alignTokens := rawRows[1]
+	rowTokens := rawRows[2:]
+	header := make([]string, 0)
+	delimiter := make([]int, 0)
+	rows := make([][]string, 0)
+
+	for _, t := range tokenizer.Split(headerTokens, tokenizer.Pipe) {
+		header = append(header, tokenizer.Stringify(t[1:len(t)-1]))
+	}
+	for _, t := range tokenizer.Split(alignTokens, tokenizer.Pipe) {
+		delimiter = append(delimiter, len(t[1:len(t)-1]))
+	}
+	for _, row := range rowTokens {
+		cells := make([]string, 0)
+		for _, t := range tokenizer.Split(row, tokenizer.Pipe) {
+			cells = append(cells, tokenizer.Stringify(t[1:len(t)-1]))
+		}
+		rows = append(rows, cells)
+	}
+
+	return &ast.Table{
+		Header:    header,
+		Delimiter: delimiter,
+		Rows:      rows,
+	}, nil
+}
+
+func matchTableCellTokens(tokens []*tokenizer.Token) (int, bool) {
+	if len(tokens) == 0 {
+		return 0, false
+	}
+
+	pipes := 0
+	for _, token := range tokens {
+		if token.Type == tokenizer.Pipe {
+			pipes++
+		}
+	}
+	cells := tokenizer.Split(tokens, tokenizer.Pipe)
+	if len(cells) != pipes-1 {
+		return 0, false
+	}
+	for _, cellTokens := range cells {
+		if len(cellTokens) == 0 {
+			return 0, false
+		}
+		if cellTokens[0].Type != tokenizer.Space {
+			return 0, false
+		}
+		if cellTokens[len(cellTokens)-1].Type != tokenizer.Space {
+			return 0, false
+		}
+	}
+
+	return len(cells), true
+}

plugin/gomark/parser/table_test.go

+package parser
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/require"
+	"github.com/usememos/memos/plugin/gomark/ast"
+	"github.com/usememos/memos/plugin/gomark/parser/tokenizer"
+	"github.com/usememos/memos/plugin/gomark/restore"
+)
+
+func TestTableParser(t *testing.T) {
+	tests := []struct {
+		text  string
+		table ast.Node
+	}{
+		{
+			text: "| header |\n| --- |\n| cell |\n",
+			table: &ast.Table{
+				Header:    []string{"header"},
+				Delimiter: []int{3},
+				Rows: [][]string{
+					{"cell"},
+				},
+			},
+		},
+		{
+			text: "| header1 | header2 |\n| --- | ---- |\n| cell1 | cell2 |\n| cell3 | cell4 |",
+			table: &ast.Table{
+				Header:    []string{"header1", "header2"},
+				Delimiter: []int{3, 4},
+				Rows: [][]string{
+					{"cell1", "cell2"},
+					{"cell3", "cell4"},
+				},
+			},
+		},
+	}
+
+	for _, test := range tests {
+		tokens := tokenizer.Tokenize(test.text)
+		node, _ := NewTableParser().Parse(tokens)
+		require.Equal(t, restore.Restore([]ast.Node{test.table}), restore.Restore([]ast.Node{node}))
+	}
+}

plugin/gomark/parser/tokenizer/tokenizer.go

@@ -20,6 +20,7 @@ const (
 	GreaterThan        TokenType = ">"
 	DollarSign         TokenType = "$"
 	EqualSign          TokenType = "="
+	Pipe               TokenType = "|"
 	Backslash          TokenType = "\\"
 	Newline            TokenType = "\n"
 	Space              TokenType = " "
@@ -80,6 +81,8 @@ func Tokenize(text string) []*Token {
 			tokens = append(tokens, NewToken(DollarSign, "$"))
 		case '=':
 			tokens = append(tokens, NewToken(EqualSign, "="))
+		case '|':
+			tokens = append(tokens, NewToken(Pipe, "|"))
 		case '\\':
 			tokens = append(tokens, NewToken(Backslash, `\`))
 		case '\n':
@@ -121,3 +124,22 @@ func Stringify(tokens []*Token) string {
 	}
 	return text
 }
+
+func Split(tokens []*Token, delimiter TokenType) [][]*Token {
+	result := make([][]*Token, 0)
+	current := make([]*Token, 0)
+	for _, token := range tokens {
+		if token.Type == delimiter {
+			if len(current) > 0 {
+				result = append(result, current)
+				current = make([]*Token, 0)
+			}
+		} else {
+			current = append(current, token)
+		}
+	}
+	if len(current) > 0 {
+		result = append(result, current)
+	}
+	return result
+}
\ No newline at end of file