amu/parser/para.go

package parser

import (
	"bytes"
	"strings"

	"go.arsenm.dev/amu/ast"
	"go.arsenm.dev/amu/scanner"
)

// parsePara attempts to parse a paragraph until untilEOLAmt
// newlines are encountered
func (p *Parser) parsePara(untilEOLAmt int) *ast.Para {
	// Create new empty para
	para := &ast.Para{}

parseLoop:
	for {
		// Scan token
		tok, lit := p.scan()

		switch tok {
		case scanner.WS:
			// Add whitespace to para
			para.Fragments = append(para.Fragments, ast.ParaFragment{Whitespace: &lit})
		case scanner.PUNCT:
			if lit == "[" {
				// Attempt to parse link
				link, _ := p.parseLink()
				// If successful
				if link != nil {
					// Add link to para
					para.Fragments = append(para.Fragments, ast.ParaFragment{Link: link})
					// Continue to next token
					continue
				}
			}
			// Add punctuation to para
			para.Fragments = append(para.Fragments, ast.ParaFragment{Punct: &lit})
		case scanner.WORD:
			if strings.HasPrefix(lit, "+") {
				// Attempt to parse function
				function := p.parseFunc(tok, lit)
				// If successful
				if function != nil {
					// Add function to para
					para.Fragments = append(para.Fragments, ast.ParaFragment{Func: function})
					// Continue to next token
					continue
				}
			}
			// Add word to para
			para.Fragments = append(para.Fragments, ast.ParaFragment{Word: &lit})
		case scanner.FORMAT:
			// Create new nil slice of ast.FormatType
			var types []ast.FormatType
			if strings.HasPrefix(lit, "_") {
				// Remove leading and trailing "_" 
				lit = strings.Trim(lit, "_")
				// Add italic format to slice
				types = append(types, ast.FormatTypeItalic)
			}
			if strings.HasPrefix(lit, "*") {
				// Remove leading and trailing "*"
				lit = strings.Trim(lit, "*")
				// Add bold format to slice
				types = append(types, ast.FormatTypeBold)
			}
			if strings.HasPrefix(lit, "$") {
				// Remove leading and trailing "$"
				lit = strings.Trim(lit, "$")
				// Add math format to slice
				types = append(types, ast.FormatTypeMath)
			}
			if strings.HasPrefix(lit, "`") {
				// Remove leading and trailing "`"
				lit = strings.Trim(lit, "`")
				// Add code format to slice
				types = []ast.FormatType{ast.FormatTypeCode}
			}
			if strings.HasPrefix(lit, "~") {
				// Remove leading and trailing "~"
				lit = strings.Trim(lit, "~")
				// Add strike format to slice
				types = []ast.FormatType{ast.FormatTypeStrike}
			}
			// Add format to para
			para.Fragments = append(para.Fragments, ast.ParaFragment{Format: &ast.Format{
				Types: types,
				Text:  lit,
			}})
		case scanner.EOL:
			// If untilEOLAmt or more newlines encountered
			if strings.Count(lit, "\n") >= untilEOLAmt {
				// Stop parsing
				break parseLoop
			}
			// Add EOL to para
			para.Fragments = append(para.Fragments, ast.ParaFragment{Whitespace: &lit})
		case scanner.EOF:
			// Stop parsing
			break parseLoop
		}
	}

	// If nothing in para
	if len(para.Fragments) == 0 {
		// Return nothing
		return nil
	}

	return para
}

// parseFunc appempts to parse a function call
func (p *Parser) parseFunc(tok scanner.Token, lit string) *ast.Func {
	// Create new function
	function := &ast.Func{}

	// If the token is not a word or does not have a prefix of "+"
	if tok != scanner.WORD || !strings.HasPrefix(lit, "+") {
		// Return nil as this is an invalid function call
		return nil
	}

	// Set function name to literal, trimming "+" prefix
	function.Name = strings.TrimPrefix(lit, "+")

	// Scan token
	tok, lit = p.scan()

	// If token is not punctuatuion or is not "["
	if tok != scanner.PUNCT || lit != "[" {
		// Unscan token
		p.unscan()
		// Return nil as this is an invalid function call
		return nil
	}

	// Parse arguments
	function.Args = p.parseArgs()

	return function
}

// Attempt to parse link
func (p *Parser) parseLink() (*ast.Link, bool) {
	// Create new link
	link := &ast.Link{}

	// Initialize buffers for link properties
	textBuf := &bytes.Buffer{}
	linkBuf := &bytes.Buffer{}
	// Set current buffer to text buffer
	currentBuf := textBuf

	// Declare variable for last literal
	var lastLit string

	// Define variable for amount of scans performed
	amtScans := 0
parseLoop:
	for {
		// Scan token
		tok, lit := p.scan()
		// Increment amtScans
		amtScans++

		switch tok {
		case scanner.WORD:
			// Write word to current buffer
			currentBuf.WriteString(lit)
		case scanner.WS:
			// Write word to current buffer
			currentBuf.WriteString(lit)
		case scanner.PUNCT:
			// If closing bracket found but no text stored	
			if lit == "]" && currentBuf.Len() == 0 {
				// Unscan token
				p.unscan()
				// Return nil as this is an invalid link
				return nil, false
			}
			// If last literal is "]" and current is "("
			if lastLit == "]" && lit == "(" {
				// Switch current buffer to link buffer
				currentBuf = linkBuf
				// Continue to next token
				continue
			}
			// If literal is ")"
			if lit == ")" {
				// Stop parsing
				break parseLoop
			}
			// If literal is not "]"
			if lit != "]" {
				// Write literal to current buffer
				currentBuf.WriteString(lit)
			}
		case scanner.EOL, scanner.EOF:
			// Unscan all performed scans
			p.unscanMulti(amtScans)
			// Return nil as this is an invalid link
			return nil, false
		}

		// Set last literal
		lastLit = lit
	}

	// If no text
	if textBuf.Len() == 0 {
		// Use link as text
		textBuf.WriteString(linkBuf.String())
	}

	// Set properties
	link.Text = textBuf.String()
	link.Link = linkBuf.String()

	return link, false
}