| 26 | 26 | |
| 27 | 27 | |
| 28 | 28 | def parse(input_string: str) -> SgfTree: |
| 29 | | pass |
| 29 | # Edge Case: Empty input or doesn't start with '(' |
| 30 | if not input_string or input_string[0] != '(': |
| 31 | raise ValueError("tree missing") |
| 32 | |
| 33 | # Edge Case: Doesn't end with ')' |
| 34 | if input_string[-1] != ')': |
| 35 | raise ValueError("tree missing") |
| 36 | |
| 37 | # Parse the content inside the outermost parentheses |
| 38 | content = input_string[1:-1] |
| 39 | |
| 40 | # Edge Case: Empty content after removing parentheses |
| 41 | if not content: |
| 42 | raise ValueError("tree with no nodes") |
| 43 | |
| 44 | # Check if it starts with a semicolon (node) |
| 45 | if content[0] != ';': |
| 46 | raise ValueError("tree with no nodes") |
| 47 | |
| 48 | # Parse the first node and any subsequent variations |
| 49 | i = 1 # Start after the initial semicolon |
| 50 | properties = {} |
| 51 | |
| 52 | # Parse properties of the current node |
| 53 | while i < len(content) and content[i] != '(' and content[i] != ';': |
| 54 | # Parse property key |
| 55 | key_start = i |
| 56 | while i < len(content) and content[i].isalpha(): |
| 57 | i += 1 |
| 58 | |
| 59 | # Edge Case: Key is not all uppercase |
| 60 | key = content[key_start:i] |
| 61 | if not key.isupper(): |
| 62 | raise ValueError("property must be in uppercase") |
| 63 | |
| 64 | # Edge Case: No property values |
| 65 | if i >= len(content) or content[i] != '[': |
| 66 | raise ValueError("properties without delimiter") |
| 67 | |
| 68 | # Parse property values |
| 69 | values = [] |
| 70 | while i < len(content) and content[i] == '[': |
| 71 | i += 1 # Skip '[' |
| 72 | value_start = i |
| 73 | |
| 74 | # Parse value, handling escapes |
| 75 | value_chars = [] |
| 76 | while i < len(content) and content[i] != ']': |
| 77 | if content[i] == '\\': |
| 78 | i += 1 |
| 79 | if i >= len(content): |
| 80 | raise ValueError("properties without delimiter") |
| 81 | |
| 82 | # Handle escaped characters according to SGF Text type rules |
| 83 | if content[i] == '\n': |
| 84 | # If newline comes immediately after backslash, remove it |
| 85 | pass |
| 86 | elif content[i].isspace(): |
| 87 | # Any whitespace after \ is converted to space |
| 88 | value_chars.append(' ') |
| 89 | else: |
| 90 | # Any non-whitespace character after \ is inserted as-is |
| 91 | value_chars.append(content[i]) |
| 92 | elif content[i] == '\n': |
| 93 | # Newlines remain as newlines unless escaped |
| 94 | value_chars.append(content[i]) |
| 95 | elif content[i].isspace() and content[i] != '\n': |
| 96 | # All whitespace other than newline converted to space |
| 97 | value_chars.append(' ') |
| 98 | else: |
| 99 | value_chars.append(content[i]) |
| 100 | i += 1 |
| 101 | |
| 102 | # Edge Case: Unclosed property value |
| 103 | if i >= len(content) or content[i] != ']': |
| 104 | raise ValueError("properties without delimiter") |
| 105 | |
| 106 | value = ''.join(value_chars) |
| 107 | values.append(value) |
| 108 | i += 1 # Skip ']' |
| 109 | |
| 110 | properties[key] = values |
| 111 | |
| 112 | # Parse children |
| 113 | children = [] |
| 114 | |
| 115 | # Handle variations |
| 116 | while i < len(content) and content[i] == '(': |
| 117 | # Find matching closing parenthesis |
| 118 | paren_count = 1 |
| 119 | start = i + 1 |
| 120 | i += 1 |
| 121 | |
| 122 | while i < len(content) and paren_count > 0: |
| 123 | if content[i] == '(': |
| 124 | paren_count += 1 |
| 125 | elif content[i] == ')': |
| 126 | paren_count -= 1 |
| 127 | i += 1 |
| 128 | |
| 129 | # Edge Case: Unmatched parenthesis |
| 130 | if paren_count != 0: |
| 131 | raise ValueError("tree missing") |
| 132 | |
| 133 | # Parse the subtree |
| 134 | subtree_content = content[start:i-1] |
| 135 | # Prepend ';' to make it a valid node |
| 136 | subtree_string = '(' + subtree_content + ')' |
| 137 | child_tree = parse(subtree_string) |
| 138 | children.append(child_tree) |
| 139 | |
| 140 | # Handle linear children (after variations) |
| 141 | if i < len(content) and content[i] == ';': |
| 142 | # Parse the rest as a single child |
| 143 | rest_content = content[i:] |
| 144 | rest_string = '(' + rest_content + ')' |
| 145 | child_tree = parse(rest_string) |
| 146 | children.append(child_tree) |
| 147 | |
| 148 | return SgfTree(properties, children) |
| 149 | # Handled Edge Cases: Empty input, missing parentheses, empty content, missing semicolon, |
| 150 | # lowercase property keys, missing delimiters, unclosed property values, |
| 151 | # unmatched parentheses |