/// 公式记号。
pub enum Token {
/// 数字字面量(含科学计数)。
Number(Double)
/// 字符串字面量(`""` 转义已解出)。
Str(String)
/// TRUE/FALSE(大小写不敏感)。
BoolLit(Bool)
/// 错误字面量(如 `#DIV/0!`)。
ErrLit(@value.XlError)
/// 标识符(函数名或单元格引用,由解析器消歧)。
Ident(String)
/// 运算符(`+ - * / ^ & = <> < <= > >=`)。
Op(String)
/// `(`
LParen
/// `)`
RParen
/// `,`
Comma
/// `:`(区域)
Colon
/// `%`(后缀)
Percent
} derive(Show, Debug, Eq)
fn is_ident_start(c : Char) -> Bool {
@value.is_ascii_letter(c) || c == '$' || c == '_'
}
fn is_ident_char(c : Char) -> Bool {
@value.is_ascii_letter(c) || @value.is_ascii_digit(c) || c == '$' || c == '_'
}
fn is_space(c : Char) -> Bool {
c == ' ' || c == '\t' || c == '\n' || c == '\r'
}
/// 把公式源码切成记号序列(前导 `=` 会被跳过);词法错误(如未闭合字符串)抛 ParseError。
pub fn lex(input : String) -> Array[Token] raise ParseError {
let chars : Array[Char] = Array::from_iter(input.iter())
let n = chars.length()
let toks : Array[Token] = []
let mut i = 0
while i < n && is_space(chars[i]) {
i += 1
}
if i < n && chars[i] == '=' {
i += 1
}
while i < n {
let c = chars[i]
if is_space(c) {
i += 1
continue
}
if @value.is_ascii_digit(c) ||
(c == '.' && i + 1 < n && @value.is_ascii_digit(chars[i + 1])) {
let start = i
while i < n && @value.is_ascii_digit(chars[i]) {
i += 1
}
if i < n && chars[i] == '.' {
i += 1
while i < n && @value.is_ascii_digit(chars[i]) {
i += 1
}
}
if i < n && (chars[i] == 'e' || chars[i] == 'E') {
i += 1
if i < n && (chars[i] == '+' || chars[i] == '-') {
i += 1
}
if i >= n || !@value.is_ascii_digit(chars[i]) {
raise ParseError("bad exponent at position " + i.to_string())
}
while i < n && @value.is_ascii_digit(chars[i]) {
i += 1
}
}
let text = String::from_array(chars[start:i])
let d = @string.parse_double(text[:]) catch {
_ => raise ParseError("bad number: " + text)
}
toks.push(Number(@value.excel_precision(d)))
continue
}
if c == '"' {
i += 1
let buf : Array[Char] = []
let mut closed = false
while i < n {
let ch = chars[i]
if ch == '"' {
if i + 1 < n && chars[i + 1] == '"' {
buf.push('"')
i += 2
} else {
i += 1
closed = true
break
}
} else {
buf.push(ch)
i += 1
}
}
if !closed {
raise ParseError("unterminated string literal")
}
toks.push(Str(String::from_array(buf[:])))
continue
}
if c == '#' {
let start = i
i += 1
while i < n &&
(@value.is_ascii_letter(chars[i]) ||
@value.is_ascii_digit(chars[i]) ||
chars[i] == '/' ||
chars[i] == '!' ||
chars[i] == '?') {
i += 1
}
let text = @value.ascii_upper(String::from_array(chars[start:i]))
match @value.XlError::parse(text) {
Some(e) => toks.push(ErrLit(e))
None => raise ParseError("unknown error literal: " + text)
}
continue
}
if is_ident_start(c) {
let start = i
while i < n && is_ident_char(chars[i]) {
i += 1
}
let raw = String::from_array(chars[start:i])
let up = @value.ascii_upper(raw)
if up == "TRUE" {
toks.push(BoolLit(true))
} else if up == "FALSE" {
toks.push(BoolLit(false))
} else {
toks.push(Ident(raw))
}
continue
}
if c == '<' {
if i + 1 < n && chars[i + 1] == '=' {
toks.push(Op("<="))
i += 2
} else if i + 1 < n && chars[i + 1] == '>' {
toks.push(Op("<>"))
i += 2
} else {
toks.push(Op("<"))
i += 1
}
continue
}
if c == '>' {
if i + 1 < n && chars[i + 1] == '=' {
toks.push(Op(">="))
i += 2
} else {
toks.push(Op(">"))
i += 1
}
continue
}
match c {
'=' => toks.push(Op("="))
'+' => toks.push(Op("+"))
'-' => toks.push(Op("-"))
'*' => toks.push(Op("*"))
'/' => toks.push(Op("/"))
'^' => toks.push(Op("^"))
'&' => toks.push(Op("&"))
'(' => toks.push(LParen)
')' => toks.push(RParen)
',' => toks.push(Comma)
':' => toks.push(Colon)
'%' => toks.push(Percent)
_ => raise ParseError("unexpected character: " + c.to_string())
}
i += 1
}
toks
}