Python - 分子式的语法检查

Question

这将是一个很长的问题，希望大家耐心等待。

我正在编写一个程序来检查分子式的语法是否正确。

我有一个 BNF 语法：

<formel>::= <mol> \n
<mol>   ::= <group> | <group><mol>
<group> ::= <atom> |<atom><num> | (<mol>) <num>
<atom>  ::= <LETTER> | <LETTER><letter>
<LETTER>::= A | B | C | ... | Z
<letter>::= a | b | c | ... | z
<num>   ::= 2 | 3 | 4 | ...

这是我的代码：

from linkedQFile import LinkedQ
import string
import sys

ATOMER = ["H","He","Li","Be","B","C","N","O","F","Ne","Na","Mg","Al","Si","P","S","Cl","Ar"]

class FormelError(Exception):
    pass
class Gruppfel(Exception):
    pass

q = LinkedQ()
formel= "(Cl)2)3"

for symbol in formel:
    q.put(symbol)


def readNum():
    """Reads digits larger than 1. Raises exception if condition is not fulfilled."""
    try:
        if int(q.peek()) >= 2:
            print(q.peek())
            q.get()
            return
        else:
            q.get()
            print("Too small digit at the end of row: "+getRest())
            sys.exit()
    except (ValueError,TypeError):
        raise FormelError("Not a number.")

def readletter():
    """Reads lowercase letters and returns them."""
    if q.peek() in string.ascii_lowercase:
        print(q.peek())
        return q.get()
    else:
        raise FormelError("Expected lowercase letter.")

def readLetter():
    """Reads capital letters and returns them."""
    if q.peek() in string.ascii_uppercase:
        print(q.peek())
        return q.get()
    else:
        raise FormelError("Expected capital letter.")

def readAtom():
    """Reads atoms on the form X and Xx. Raises Exception if the format for an atom is not fulfilled or if the atom does not exist."""
    X = ""
    try:
        X += readLetter()
    except FormelError:
        print("Missing capital letter at end of row: "+getRest())
        sys.exit()  
        return

    try:
        x = readletter()
        atom = X+x
    except (FormelError, TypeError):
        atom = X

    if atom in ATOMER:
        return
    else:
        raise FormelError("Unknown atom.")

def readGroup():
    if q.peek() in string.ascii_uppercase or q.peek() in string.ascii_lowercase:
        try:
            readAtom()
        except:
            print("Unknown atom at end of row: "+getRest())
            sys.exit()
        try:
            while True:
                readNum()
        except FormelError:
            pass
        return
    if q.peek() == "(":
        print(q.peek())
        q.get()
        try:
            readMol()
        except FormelError:
            pass
        if q.peek() == ")":
            print(q.peek())
            q.get()
        else:
            print("Missing right parenthesis at end of row: "+ getRest())
            sys.exit()
            return
        digitfound = False
        try:
            while True:
                readNum()
                digitfound = True
        except:
            if digitfound:
                return
            print("Missing digit at end of row: "+getRest())
            sys.exit()
            return
    raise FormelError("Incorrect start of group")

def readMol():
    try:
        readGroup()
    except FormelError:
        print("Incorrect start of group at end of row: "+getRest()) 
        raise FormelError
    if q.peek() == None:
        return
    if not q.peek() == ")": 
        try:
            readMol()
        except FormelError:
            pass

def readFormel():
    try:
        readMol()
    except:
        return
    print("Correct formula")

def getRest():
    rest = ""
    while not q.isEmpty():
        rest += q.get()
    return rest

readFormel()

现在代码应该接受一些给定的公式，并为一些给定的不正确公式提供错误代码。让我们看看这些给定的公式：

正确： Si(C3(COOH)2)4(H2O)7

不正确： H2O)Fe

(Cl)2)3

程序接受正确的公式，但不幸的是也接受不正确的公式。原因在于：

中的if语句

if not q.peek() == ")": 
    try:
        readMol()
    except FormelError:
        pass

使右侧的括号不平衡（右侧有一个或多个括号过多）滑过代码，而不是被检测为“组”的不正确开头。我该如何解决这个问题，同时仍然让 Si(C3(COOH)2)4(H2O)7 被接受为语法正确？

感谢您的耐心等待:)

Answer 1

您的 readMol 代码对“)”进行了这个错误的测试（您甚至告诉过我们）。如果您正在编码（如您所愿）recursive descent parser，则您的语法表明不需要进行此类测试。

事实上，你的语法对 mol 有一个奇怪的规则：

<mol>   ::= <group> | <group><mol>

这不适用于递归下降解析器。您已重构此类规则以共享每个规则中的公共前缀。在这种情况下，很容易：

<mol>   ::= <group> ( <mol> | empty ) ;

然后你直接从语法规则编写代码（见上面的链接） [除了“）”检查之外，您确实这样做了。] 它应该看起来像这样（我不是 python 专家）：

def readMol():
    try:
        readGroup()
    except FormelError:
        print("Incorrect start of group at end of row: "+getRest()) 
        raise FormelError
    try:
       readMol()
    except FormelError:
       pass

在编写递归下降解析器时很有帮助，首先将语法按摩成最兼容的形式（就像我对你的 mol 规则所做的那样）。然后对各个识别器进行编码是一项很难出错的纯机械任务。