作为学习 Ruby 的一部分,我尝试实现一个基本的解释器,它读取输入并进行基本的算术计算。到目前为止,基本算术运算可以工作,但运算符优先级存在问题。尚未处理。这是代码。我处于初级水平。这段代码中的任何错误都是由于我缺乏知识造成的。如何修改此代码以处理运算符优先级。 示例输出
2+2+2 = 6 #correct
10+10/2 = 10 # incorrect as in irb answer must be 15
=begin
Basic calculator Interpreter
can add, substract, multiply , divide with any number of operands at a time
Drawback : Lacks operator precedence
=end
class Interpreter
attr_accessor :input
def initialize
@input = gets.chomp
end
def intepret
first_operand = []
f = []
operator = '+'
array = Array.new
lc = 0
@input.split.join.split("").each_with_index.map do |i, index|
if i.is_number?
first_operand.push(i)
if index == @input.length-1
array.push(first_operand.join("").to_i)
end
elsif i.is_plus?
f = first_operand
first_operand = nil
first_operand = []
array.push(f.join("").to_i)
array.push("+")
elsif i.is_minus?
f = first_operand
first_operand = nil
first_operand = []
operator = '-'
array.push(f.join("").to_i)
array.push("-")
elsif i.is_multi?
f = first_operand
first_operand = nil
first_operand = []
operator = '*'
array.push(f.join("").to_i)
array.push("*")
elsif i.is_divide?
f = first_operand
first_operand = nil
first_operand = []
operator = '/'
array.push(f.join("").to_i)
array.push("/")
else
puts "Illegal input exiting.."
exit
end
lc = lc+1
end
#apply the appropriate operation on the inputs based on the operand
#puts "=======TOKENS======"
#puts array.inspect
result = 0
array.each_with_index.map do |x, key|
result = x if key == 0
if x == '+'
if key == 0
result = add(result, array[key+1])
else
result = add(result, array [key+1])
end
elsif x == '-'
if key == 0
result = minus(result, array[key+1])
else
result = minus(result, array [key+1])
end
elsif x == '*'
if key == 0
result = multi(result, array[key+1])
else
result = multi(result, array [key+1])
end
elsif x == '/'
begin
if key == 0
result = divide(result, array[key+1])
else
result = divide(result, array [key+1])
end
rescue
puts "Zero Divsion error"
exit
end
end
end
puts "Result is: "+result.to_s
end
def print_token(type, value)
puts type + ' '+ value
end
def add(f,s)
return f.to_i + s.to_i
end
def minus(f,s)
return f.to_i - s.to_i
end
def multi(f,s)
return f.to_i * s.to_i
end
def divide(f,s)
return f.to_i / s.to_i
end
end
# Override the string class, to directly use methods like obj.is_number? rather than is_number?(obj)
class String
def is_number?
true if Float(self) rescue false
end
def is_plus?
true if self == '+' rescue false
end
def is_minus?
true if self == '-' rescue false
end
def is_multi?
true if self == '*' rescue false
end
def is_divide?
true if self == '/' rescue false
end
end
#continue accepting inputs until exit CTRL + D
while true
print 'pck>:'
i_obj = Interpreter.new
i_obj.intepret
end
首先,使用 Shunting-yard 算法 处理输入。这应该给出一个采用 Reverse Polish notation (RPN) 的标记列表。然后你就可以计算 RPN 表达式了。
我知道这是一个老问题,但它首先出现在谷歌上,并且接受的答案确实没有给我我想要的东西。
我相信原始发帖者正在寻找的东西称为递归体面解析器。基本上实现了一个递归地解决自身问题的解析器。为了理解,让我们看看我们正在尝试做的事情的语法。
EXPR: factor(add|sub factor)*
FACTOR: number
如果我们更新语法,将数字设置为数字或左右括号,我们就可以简单地解决问题:
EXPR: factor(add|sub factor)*
FACTOR: number | lparen expr rparen
这样,如果我们遇到括号,我们只需求解其中的表达式,然后返回其结果。因此我们递归地调用 EXPR。
这里用代码更详细地解释了这一点, 博客:https://ruslanspivak.com/lsbasi-part5/ 代码:https://github.com/rspivak/lsbasi/blob/master/part6/calc6.py#L126-L136