mirror of https://gitlab.com/cppit/libclangmm
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
83 lines
3.7 KiB
83 lines
3.7 KiB
#include "Tokens.h" |
|
#include "Utility.h" |
|
#include <unordered_set> |
|
#include <cstring> |
|
|
|
clangmm::Tokens::Tokens(CXTranslationUnit &cx_tu, const SourceRange &range): cx_tu(cx_tu) { |
|
unsigned num_tokens; |
|
clang_tokenize(cx_tu, range.cx_range, &cx_tokens, &num_tokens); |
|
cx_cursors=std::unique_ptr<CXCursor[]>(new CXCursor[num_tokens]); // To avoid allocation with initialization |
|
clang_annotateTokens(cx_tu, cx_tokens, num_tokens, cx_cursors.get()); |
|
for (unsigned i = 0; i < num_tokens; i++) { |
|
if(cx_cursors[i].kind==CXCursor_DeclRefExpr) { // Temporary fix to a libclang bug |
|
auto real_cursor=clang_getCursor(cx_tu, clang_getTokenLocation(cx_tu, cx_tokens[i])); |
|
cx_cursors[i]=real_cursor; |
|
} |
|
// Corrects: when getting tokens from a header, FieldDecl tokens are getting ClassDecl or StructDecl cursors |
|
else if(cx_cursors[i].kind==CXCursor_ClassDecl || cx_cursors[i].kind==CXCursor_StructDecl) { |
|
Token token(cx_tu, cx_tokens[i], cx_cursors[i]); |
|
auto cursor=token.get_cursor(); |
|
auto token_offsets=token.get_source_range().get_offsets(); |
|
if(token_offsets.second!=cursor.get_source_range().get_offsets().second && token_offsets.first!=cursor.get_source_location().get_offset() && token.is_identifier()) { |
|
class VisitorData { |
|
public: |
|
std::string path; |
|
bool found_path; |
|
std::pair<Offset, Offset> offsets; |
|
CXCursor found_cursor; |
|
}; |
|
VisitorData data{token.get_source_location().get_path(), false, token_offsets, clang_getNullCursor()}; |
|
auto translation_unit_cursor = clang_getTranslationUnitCursor(cx_tu); |
|
clang_visitChildren(translation_unit_cursor, [](CXCursor cx_cursor, CXCursor cx_parent, CXClientData data_) { |
|
auto data=static_cast<VisitorData*>(data_); |
|
Cursor cursor(cx_cursor); |
|
if(data->found_path || cursor.get_source_location().get_path()==data->path) { |
|
data->found_path=true; |
|
if(cursor.get_source_range().get_offsets().second==data->offsets.second && cursor.get_source_location().get_offset()==data->offsets.first) { |
|
data->found_cursor=cx_cursor; |
|
return CXChildVisit_Break; |
|
} |
|
return CXChildVisit_Recurse; |
|
} |
|
return CXChildVisit_Continue; |
|
}, &data); |
|
|
|
if(!clang_Cursor_isNull(data.found_cursor)) |
|
cx_cursors[i]=data.found_cursor; |
|
} |
|
} |
|
emplace_back(Token(cx_tu, cx_tokens[i], cx_cursors[i])); |
|
} |
|
} |
|
|
|
clangmm::Tokens::~Tokens() { |
|
clang_disposeTokens(cx_tu, cx_tokens, size()); |
|
} |
|
|
|
//This works across TranslationUnits. Similar tokens defined as tokens with equal canonical cursors. |
|
std::vector<std::pair<clangmm::Offset, clangmm::Offset> > clangmm::Tokens::get_similar_token_offsets(const std::string &spelling, |
|
const std::unordered_set<std::string> &usrs) { |
|
std::vector<std::pair<Offset, Offset> > offsets; |
|
for(auto &token: *this) { |
|
if(token.is_identifier()) { |
|
auto referenced=token.get_cursor().get_referenced(); |
|
if(referenced) { |
|
const char *strstr_ptr=nullptr; |
|
auto cx_string=clang_getTokenSpelling(cx_tu, token.cx_token); |
|
if(cx_string.data) |
|
strstr_ptr=std::strstr(static_cast<const char*>(cx_string.data), spelling.c_str()); |
|
clang_disposeString(cx_string); |
|
if(strstr_ptr) { |
|
auto referenced_usrs=referenced.get_all_usr_extended(); |
|
for(auto &usr: referenced_usrs) { |
|
if(usrs.count(usr)) { |
|
offsets.emplace_back(token.get_source_range().get_offsets()); |
|
break; |
|
} |
|
} |
|
} |
|
} |
|
} |
|
} |
|
return offsets; |
|
}
|
|
|