brokk_bifrost_cpp/
clones.rs

1//! Token and AST-label normalization for C++ structural-clone candidates.
2//!
3//! `CloneCandidateData` and `compact_clone_excerpt` are analysis-owned and the
4//! declaration source comes from the analyzer, so `analyzer/cpp/clones.rs` keeps
5//! the ~12-LOC entry point; everything that knows C++ is here.
6
7use brokk_bifrost_core::analyzer::tree_walk::push_children_reversed;
8use brokk_bifrost_core::clones::{
9    CloneNormalizationRules, normalize_clone_ast_label as normalize_cpp_clone_ast_label,
10    normalize_clone_leaf_token as normalize_cpp_clone_leaf_token,
11};
12use tree_sitter::{Node, Parser};
13
14const CPP_CLONE_AST_IDENTIFIER_TYPES: &[&str] = &[
15    "identifier",
16    "field_identifier",
17    "namespace_identifier",
18    "type_identifier",
19];
20const CPP_CLONE_AST_STRING_TYPES: &[&str] = &["string_literal", "raw_string_literal"];
21const CPP_CLONE_AST_NUMBER_TYPES: &[&str] = &["number_literal"];
22const CPP_CLONE_NORMALIZATION_RULES: CloneNormalizationRules = CloneNormalizationRules {
23    identifier_kinds: CPP_CLONE_AST_IDENTIFIER_TYPES,
24    string_kinds: CPP_CLONE_AST_STRING_TYPES,
25    number_kinds: CPP_CLONE_AST_NUMBER_TYPES,
26    boolean_kinds: &[],
27    boolean_values: &["true", "false"],
28    null_kinds: &[],
29    null_values: &[],
30    comment_kinds: &["comment"],
31    ignored_ast_kinds: &[],
32};
33
34pub fn cpp_clone_parser() -> Parser {
35    let mut parser = Parser::new();
36    parser
37        .set_language(&tree_sitter_cpp::LANGUAGE.into())
38        .expect("failed to load cpp parser");
39    parser
40}
41
42/// The normalized token stream and the joined AST-label signature for one C++
43/// declaration body, in a single walk.
44pub fn cpp_clone_profile(parser: &mut Parser, source: &str) -> (Vec<String>, String) {
45    let Some(tree) = parser.parse(source, None) else {
46        return (Vec::new(), String::new());
47    };
48    let mut normalized_tokens = Vec::new();
49    let mut ast_labels = Vec::new();
50    let mut stack = vec![tree.root_node()];
51    while let Some(node) = stack.pop() {
52        if cpp_is_ignorable_clone_logging_node(node, source) {
53            continue;
54        }
55        ast_labels.push(normalize_cpp_clone_ast_label(
56            node,
57            source,
58            CPP_CLONE_NORMALIZATION_RULES,
59        ));
60        if node.named_child_count() == 0 {
61            let token = normalize_cpp_clone_leaf_token(node, source, CPP_CLONE_NORMALIZATION_RULES);
62            if !token.is_empty() {
63                normalized_tokens.push(token);
64            }
65        }
66        push_children_reversed(node, &mut stack);
67    }
68    (normalized_tokens, ast_labels.join("|"))
69}
70
71fn cpp_is_ignorable_clone_logging_node(node: Node<'_>, source: &str) -> bool {
72    if node.kind() != "expression_statement" {
73        return false;
74    }
75    let text = source
76        .get(node.start_byte()..node.end_byte())
77        .unwrap_or("")
78        .trim();
79    text.contains("std::cout")
80        || text.contains("std::cerr")
81        || text.contains("std::clog")
82        || text.starts_with("printf(")
83}
brokk_bifrost_cpp/clones.rs

brokk_bifrost_cpp/
clones.rs
