File size: 5,008 Bytes
6256eb9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
"""
HTML cleaning utilities for problem text.

Provides functions to clean HTML content while preserving:
- Inline MathML (fractions, superscripts, subscripts)
- Wiris math images (extracted from data-mathml attribute)
- Table structure (formatted with | separators)
- Image placeholders
"""

import html
import pandas as pd
from bs4 import BeautifulSoup, NavigableString


def parse_mathml_element(elem):
    """Recursively parse a MathML element to text."""
    if elem.name is None:
        return str(elem).strip()

    if elem.name == "mfrac":
        children = [c for c in elem.children if c.name]
        if len(children) >= 2:
            num = parse_mathml_element(children[0])
            denom = parse_mathml_element(children[1])
            return f"({num}/{denom})"
        return elem.get_text(strip=True)

    elif elem.name == "msup":
        children = [c for c in elem.children if c.name]
        if len(children) >= 2:
            base = parse_mathml_element(children[0])
            exp = parse_mathml_element(children[1])
            return f"{base}^{exp}"
        return elem.get_text(strip=True)

    elif elem.name == "msub":
        children = [c for c in elem.children if c.name]
        if len(children) >= 2:
            base = parse_mathml_element(children[0])
            sub = parse_mathml_element(children[1])
            return f"{base}_{sub}"
        return elem.get_text(strip=True)

    elif elem.name == "msqrt":
        content = parse_mathml_element_children(elem)
        return f"√({content})"

    elif elem.name == "mo":
        op = elem.get_text(strip=True)
        if op in ["÷", "×", "·", "+", "-", "=", "<", ">", "≤", "≥", "≠"]:
            return f" {op} "
        return op

    elif elem.name in ["mn", "mi", "mtext"]:
        return elem.get_text(strip=True)

    elif elem.name in ["mrow", "math", "mpadded", "mstyle"]:
        return parse_mathml_element_children(elem)

    else:
        return elem.get_text(strip=True)


def parse_mathml_element_children(elem):
    """Parse all children of a MathML element."""
    parts = []
    for child in elem.children:
        if isinstance(child, NavigableString):
            text = str(child).strip()
            if text:
                parts.append(text)
        elif child.name:
            parts.append(parse_mathml_element(child))
    return "".join(parts)


def clean_problem_body(text):
    """
    Clean HTML problem body with full MathML handling.

    Handles:
    - Inline MathML (<math>, <mfrac>, <msup>, etc.) → (4/3), x^2
    - Wiris math images (data-mathml attribute) → [15÷12]
    - Tables → [Table: Col1 | Col2 ...]
    - Regular images → [image]
    - HTML entities → decoded properly
    """
    if pd.isna(text) or text == "":
        return ""
    soup = BeautifulSoup(str(text), "html.parser")

    # 1. Handle inline MathML
    for math in soup.find_all("math"):
        parsed = parse_mathml_element(math)
        math.replace_with(f" {parsed} ")

    # 2. Handle Wiris images
    for img in soup.find_all("img"):
        alt = img.get("alt", "")
        src = img.get("src", "")
        data_mathml = img.get("data-mathml", "")

        if "wiris" in src.lower() or "pluginwiris" in src:
            if alt and alt.strip() and alt not in ["NO ALT", "NONE"]:
                img.replace_with(f" [{alt.strip()}] ")
            elif data_mathml:
                math_str = (
                    data_mathml.replace("«", "<").replace("»", ">").replace("¨", '"')
                )
                msoup = BeautifulSoup(math_str, "html.parser")
                math_elem = msoup.find("math")
                if math_elem:
                    mtext = parse_mathml_element(math_elem)
                else:
                    mtext = msoup.get_text(separator="")
                mtext = mtext.replace("§#247;", "÷").replace("§#215;", "×")
                mtext = (
                    mtext.replace("§#8722;", "-")
                    .replace("§#160;", " ")
                    .replace("§#183;", "·")
                )
                mtext = mtext.replace("§#", "&#")
                mtext = html.unescape(mtext).strip()
                img.replace_with(f" [{mtext}] " if mtext else " [math] ")
            else:
                img.replace_with(" [math] ")
        elif alt and alt.strip():
            img.replace_with(f" [Image: {alt.strip()[:100]}] ")
        else:
            img.replace_with(" [image] ")

    # 3. Handle tables
    for table in soup.find_all("table"):
        rows = []
        for tr in table.find_all("tr"):
            cells = [td.get_text(strip=True) for td in tr.find_all(["td", "th"])]
            if any(cells):
                rows.append(" | ".join(cells))
        if rows:
            table.replace_with(f"\n[Table:\n{chr(10).join(rows)}]\n")
        else:
            table.decompose()

    text = soup.get_text(separator=" ")
    text = html.unescape(text)
    text = " ".join(text.split())
    return text.strip()