902 lines
28 KiB
C++
902 lines
28 KiB
C++
#include <string>
|
|
#include <cstring>
|
|
#include <fstream>
|
|
#include <sstream>
|
|
#include <vector>
|
|
#include <regex>
|
|
#include <sqlite3.h>
|
|
#include <filesystem>
|
|
#include <memory>
|
|
#include <tree_sitter/api.h>
|
|
#include <spdlog/spdlog.h>
|
|
#include <spdlog/sinks/stdout_color_sinks.h>
|
|
#include <spdlog/sinks/basic_file_sink.h>
|
|
#include <CLI11.hpp>
|
|
|
|
extern "C" TSLanguage *tree_sitter_cpp();
|
|
|
|
// Global address regex pattern
|
|
const std::regex ADDRESS_REGEX(R"(//\s*([0-9a-fA-F]{8}))");
|
|
|
|
// Helper function to check if a comment contains an address
|
|
bool hasAddressPattern(const std::string &comment) {
|
|
return std::regex_search(comment, ADDRESS_REGEX);
|
|
}
|
|
|
|
// Helper function to extract text from a TSNode
|
|
std::string extractNodeText(TSNode node, const char *source_code) {
|
|
uint32_t start = ts_node_start_byte(node);
|
|
uint32_t end = ts_node_end_byte(node);
|
|
return std::string(source_code + start, end - start);
|
|
}
|
|
|
|
// Helper function to find first identifier in a node
|
|
std::string findIdentifierInNode(TSNode node, const char *source_code) {
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
if (strcmp(ts_node_type(child), "identifier") == 0) {
|
|
return extractNodeText(child, source_code);
|
|
}
|
|
}
|
|
return "";
|
|
}
|
|
|
|
struct FunctionInfo {
|
|
std::string name;
|
|
std::string address;
|
|
std::string filepath;
|
|
bool is_import;
|
|
};
|
|
|
|
struct GlobalInfo {
|
|
std::string name;
|
|
std::string address;
|
|
std::string filepath;
|
|
};
|
|
|
|
class PreparedStatements {
|
|
private:
|
|
sqlite3 *db;
|
|
sqlite3_stmt *delete_functions_stmt;
|
|
sqlite3_stmt *delete_imports_stmt;
|
|
sqlite3_stmt *insert_functions_stmt;
|
|
sqlite3_stmt *insert_imports_stmt;
|
|
sqlite3_stmt *delete_globals_stmt;
|
|
sqlite3_stmt *insert_globals_stmt;
|
|
|
|
void prepareStatement(const char *sql, sqlite3_stmt **stmt,
|
|
const std::string &error_msg) {
|
|
if (sqlite3_prepare_v2(db, sql, -1, stmt, nullptr) != SQLITE_OK) {
|
|
throw std::runtime_error(error_msg + ": " + sqlite3_errmsg(db));
|
|
}
|
|
}
|
|
|
|
public:
|
|
PreparedStatements(sqlite3 *database) : db(database) {
|
|
prepareStatement("DELETE FROM Functions WHERE filepath = ?",
|
|
&delete_functions_stmt,
|
|
"Failed to prepare delete functions statement");
|
|
prepareStatement("DELETE FROM Imports WHERE filepath = ?",
|
|
&delete_imports_stmt,
|
|
"Failed to prepare delete imports statement");
|
|
prepareStatement("INSERT OR REPLACE INTO Functions (filepath, name, "
|
|
"address) VALUES (?, ?, ?)",
|
|
&insert_functions_stmt,
|
|
"Failed to prepare insert functions statement");
|
|
prepareStatement("INSERT OR REPLACE INTO Imports (filepath, name, address) "
|
|
"VALUES (?, ?, ?)",
|
|
&insert_imports_stmt,
|
|
"Failed to prepare insert imports statement");
|
|
prepareStatement("DELETE FROM Globals WHERE filepath = ?",
|
|
&delete_globals_stmt,
|
|
"Failed to prepare delete globals statement");
|
|
prepareStatement("INSERT OR REPLACE INTO Globals (filepath, name, address) "
|
|
"VALUES (?, ?, ?)",
|
|
&insert_globals_stmt,
|
|
"Failed to prepare insert globals statement");
|
|
}
|
|
|
|
~PreparedStatements() {
|
|
sqlite3_finalize(delete_functions_stmt);
|
|
sqlite3_finalize(delete_imports_stmt);
|
|
sqlite3_finalize(insert_functions_stmt);
|
|
sqlite3_finalize(insert_imports_stmt);
|
|
sqlite3_finalize(delete_globals_stmt);
|
|
sqlite3_finalize(insert_globals_stmt);
|
|
}
|
|
|
|
void clearEntriesForFile(const std::string &filepath) {
|
|
for (auto stmt : {delete_functions_stmt, delete_imports_stmt}) {
|
|
sqlite3_reset(stmt);
|
|
sqlite3_bind_text(stmt, 1, filepath.c_str(), -1, SQLITE_STATIC);
|
|
sqlite3_step(stmt);
|
|
}
|
|
}
|
|
|
|
void clearGlobalsForFile(const std::string &filepath) {
|
|
sqlite3_reset(delete_globals_stmt);
|
|
sqlite3_bind_text(delete_globals_stmt, 1, filepath.c_str(), -1,
|
|
SQLITE_STATIC);
|
|
sqlite3_step(delete_globals_stmt);
|
|
}
|
|
|
|
void insertFunction(const FunctionInfo &func) {
|
|
sqlite3_stmt *stmt =
|
|
func.is_import ? insert_imports_stmt : insert_functions_stmt;
|
|
sqlite3_reset(stmt);
|
|
sqlite3_bind_text(stmt, 1, func.filepath.c_str(), -1, SQLITE_STATIC);
|
|
sqlite3_bind_text(stmt, 2, func.name.c_str(), -1, SQLITE_STATIC);
|
|
sqlite3_bind_text(stmt, 3, func.address.c_str(), -1, SQLITE_STATIC);
|
|
sqlite3_step(stmt);
|
|
}
|
|
|
|
void insertGlobal(const GlobalInfo &global) {
|
|
sqlite3_reset(insert_globals_stmt);
|
|
sqlite3_bind_text(insert_globals_stmt, 1, global.filepath.c_str(), -1,
|
|
SQLITE_STATIC);
|
|
sqlite3_bind_text(insert_globals_stmt, 2, global.name.c_str(), -1,
|
|
SQLITE_STATIC);
|
|
sqlite3_bind_text(insert_globals_stmt, 3, global.address.c_str(), -1,
|
|
SQLITE_STATIC);
|
|
sqlite3_step(insert_globals_stmt);
|
|
}
|
|
};
|
|
|
|
class DatabaseManager {
|
|
private:
|
|
sqlite3 *db;
|
|
std::unique_ptr<PreparedStatements> prepared_stmts;
|
|
|
|
public:
|
|
DatabaseManager(const std::string &db_path) : db(nullptr) {
|
|
if (sqlite3_open(db_path.c_str(), &db) != SQLITE_OK) {
|
|
spdlog::error("Can't open database: {}", sqlite3_errmsg(db));
|
|
sqlite3_close(db);
|
|
throw std::runtime_error("Failed to open database");
|
|
}
|
|
|
|
const char *create_tables = R"(
|
|
CREATE TABLE IF NOT EXISTS Functions (filepath TEXT, name TEXT, address TEXT, PRIMARY KEY (name, filepath));
|
|
CREATE TABLE IF NOT EXISTS Imports (filepath TEXT, name TEXT, address TEXT, PRIMARY KEY (name, filepath));
|
|
CREATE TABLE IF NOT EXISTS Globals (filepath TEXT, name TEXT, address TEXT);
|
|
)";
|
|
|
|
sqlite3_exec(db, create_tables, nullptr, nullptr, nullptr);
|
|
prepared_stmts = std::make_unique<PreparedStatements>(db);
|
|
}
|
|
|
|
~DatabaseManager() {
|
|
if (db)
|
|
sqlite3_close(db);
|
|
}
|
|
|
|
void clearEntriesForFile(const std::string &filepath) {
|
|
prepared_stmts->clearEntriesForFile(filepath);
|
|
}
|
|
void insertFunction(const FunctionInfo &func) {
|
|
prepared_stmts->insertFunction(func);
|
|
}
|
|
void clearGlobalsForFile(const std::string &filepath) {
|
|
prepared_stmts->clearGlobalsForFile(filepath);
|
|
}
|
|
void insertGlobal(const GlobalInfo &global) {
|
|
prepared_stmts->insertGlobal(global);
|
|
}
|
|
void beginTransaction() {
|
|
sqlite3_exec(db, "BEGIN TRANSACTION", nullptr, nullptr, nullptr);
|
|
}
|
|
void commitTransaction() {
|
|
sqlite3_exec(db, "COMMIT", nullptr, nullptr, nullptr);
|
|
}
|
|
void rollbackTransaction() {
|
|
sqlite3_exec(db, "ROLLBACK", nullptr, nullptr, nullptr);
|
|
}
|
|
|
|
// New methods for duplicate checking
|
|
bool checkDuplicateAddresses() {
|
|
const char *sql = R"(
|
|
WITH all_addresses AS (
|
|
SELECT 'Functions' as table_name, name, address, filepath FROM Functions WHERE address != ''
|
|
UNION ALL
|
|
SELECT 'Imports' as table_name, name, address, filepath FROM Imports WHERE address != ''
|
|
UNION ALL
|
|
SELECT 'Globals' as table_name, name, address, filepath FROM Globals WHERE address != ''
|
|
)
|
|
SELECT address, COUNT(*) as count,
|
|
GROUP_CONCAT(table_name || ':' || name || ' (' || filepath || ')', '; ') as entries
|
|
FROM all_addresses
|
|
GROUP BY address
|
|
HAVING COUNT(*) > 1
|
|
ORDER BY address;
|
|
)";
|
|
|
|
sqlite3_stmt *stmt;
|
|
if (sqlite3_prepare_v2(db, sql, -1, &stmt, nullptr) != SQLITE_OK) {
|
|
spdlog::error("Failed to prepare duplicate address query: {}",
|
|
sqlite3_errmsg(db));
|
|
return false;
|
|
}
|
|
|
|
bool found_duplicates = false;
|
|
while (sqlite3_step(stmt) == SQLITE_ROW) {
|
|
found_duplicates = true;
|
|
const char *address = (const char *)sqlite3_column_text(stmt, 0);
|
|
int count = sqlite3_column_int(stmt, 1);
|
|
const char *entries = (const char *)sqlite3_column_text(stmt, 2);
|
|
|
|
spdlog::error("DUPLICATE ADDRESS: {} appears {} times in: {}", address,
|
|
count, entries);
|
|
}
|
|
|
|
sqlite3_finalize(stmt);
|
|
return found_duplicates;
|
|
}
|
|
|
|
bool checkDuplicateNames() {
|
|
bool found_duplicates = false;
|
|
|
|
// Check Functions table
|
|
const char *functions_sql = R"(
|
|
SELECT name, COUNT(*) as count,
|
|
GROUP_CONCAT(filepath, '; ') as filepaths
|
|
FROM Functions
|
|
GROUP BY name
|
|
HAVING COUNT(*) > 1
|
|
ORDER BY name;
|
|
)";
|
|
|
|
sqlite3_stmt *stmt;
|
|
if (sqlite3_prepare_v2(db, functions_sql, -1, &stmt, nullptr) ==
|
|
SQLITE_OK) {
|
|
while (sqlite3_step(stmt) == SQLITE_ROW) {
|
|
found_duplicates = true;
|
|
const char *name = (const char *)sqlite3_column_text(stmt, 0);
|
|
int count = sqlite3_column_int(stmt, 1);
|
|
const char *filepaths = (const char *)sqlite3_column_text(stmt, 2);
|
|
|
|
spdlog::error(
|
|
"DUPLICATE FUNCTION NAME: '{}' appears {} times in files: {}", name,
|
|
count, filepaths);
|
|
}
|
|
sqlite3_finalize(stmt);
|
|
}
|
|
|
|
// Check Imports table
|
|
const char *imports_sql = R"(
|
|
SELECT name, COUNT(*) as count,
|
|
GROUP_CONCAT(filepath, '; ') as filepaths
|
|
FROM Imports
|
|
GROUP BY name
|
|
HAVING COUNT(*) > 1
|
|
ORDER BY name;
|
|
)";
|
|
|
|
if (sqlite3_prepare_v2(db, imports_sql, -1, &stmt, nullptr) == SQLITE_OK) {
|
|
while (sqlite3_step(stmt) == SQLITE_ROW) {
|
|
found_duplicates = true;
|
|
const char *name = (const char *)sqlite3_column_text(stmt, 0);
|
|
int count = sqlite3_column_int(stmt, 1);
|
|
const char *filepaths = (const char *)sqlite3_column_text(stmt, 2);
|
|
|
|
spdlog::error(
|
|
"DUPLICATE IMPORT NAME: '{}' appears {} times in files: {}", name,
|
|
count, filepaths);
|
|
}
|
|
sqlite3_finalize(stmt);
|
|
}
|
|
|
|
// Check Globals table
|
|
const char *globals_sql = R"(
|
|
SELECT name, COUNT(*) as count,
|
|
GROUP_CONCAT(filepath, '; ') as filepaths
|
|
FROM Globals
|
|
GROUP BY name
|
|
HAVING COUNT(*) > 1
|
|
ORDER BY name;
|
|
)";
|
|
|
|
if (sqlite3_prepare_v2(db, globals_sql, -1, &stmt, nullptr) == SQLITE_OK) {
|
|
while (sqlite3_step(stmt) == SQLITE_ROW) {
|
|
found_duplicates = true;
|
|
const char *name = (const char *)sqlite3_column_text(stmt, 0);
|
|
int count = sqlite3_column_int(stmt, 1);
|
|
const char *filepaths = (const char *)sqlite3_column_text(stmt, 2);
|
|
|
|
spdlog::error(
|
|
"DUPLICATE GLOBAL NAME: '{}' appears {} times in files: {}", name,
|
|
count, filepaths);
|
|
}
|
|
sqlite3_finalize(stmt);
|
|
}
|
|
|
|
return found_duplicates;
|
|
}
|
|
};
|
|
|
|
std::string extractAddress(const std::string &comment) {
|
|
std::smatch match;
|
|
return std::regex_search(comment, match, ADDRESS_REGEX) ? match[1].str() : "";
|
|
}
|
|
|
|
std::string getFunctionName(TSNode node, const char *source_code) {
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
const char *type = ts_node_type(child);
|
|
|
|
if (strcmp(type, "function_declarator") == 0) {
|
|
std::string name = findIdentifierInNode(child, source_code);
|
|
if (!name.empty())
|
|
return name;
|
|
} else if (strcmp(type, "identifier") == 0) {
|
|
return extractNodeText(child, source_code);
|
|
} else if (strcmp(type, "pointer_declarator") == 0) {
|
|
std::string name = getFunctionName(child, source_code);
|
|
if (!name.empty())
|
|
return name;
|
|
}
|
|
}
|
|
return "";
|
|
}
|
|
|
|
std::string getComment(TSNode node, const char *source_code,
|
|
uint32_t source_length, bool search_before) {
|
|
TSNode current = node;
|
|
|
|
if (search_before) {
|
|
// Look for comments before the current node
|
|
while (!ts_node_is_null(current)) {
|
|
TSNode prev_sibling = ts_node_prev_sibling(current);
|
|
|
|
while (!ts_node_is_null(prev_sibling)) {
|
|
const char *type = ts_node_type(prev_sibling);
|
|
|
|
if (strcmp(type, "comment") == 0) {
|
|
std::string comment_text = extractNodeText(prev_sibling, source_code);
|
|
|
|
// Check if it contains an address pattern
|
|
if (hasAddressPattern(comment_text)) {
|
|
return comment_text;
|
|
}
|
|
}
|
|
// Skip whitespace and continue looking
|
|
else if (strcmp(type, "ERROR") != 0) {
|
|
// If we hit non-comment, non-whitespace content, stop searching
|
|
break;
|
|
}
|
|
|
|
prev_sibling = ts_node_prev_sibling(prev_sibling);
|
|
}
|
|
|
|
// Move up to parent and continue searching
|
|
current = ts_node_parent(current);
|
|
}
|
|
} else {
|
|
// Look for comments after the current node
|
|
TSNode next_sibling = ts_node_next_sibling(node);
|
|
|
|
while (!ts_node_is_null(next_sibling)) {
|
|
const char *type = ts_node_type(next_sibling);
|
|
|
|
if (strcmp(type, "comment") == 0) {
|
|
std::string comment_text = extractNodeText(next_sibling, source_code);
|
|
|
|
// Check if it contains an address pattern
|
|
if (hasAddressPattern(comment_text)) {
|
|
return comment_text;
|
|
}
|
|
}
|
|
// Skip whitespace and continue looking
|
|
else if (strcmp(type, "ERROR") != 0) {
|
|
// If we hit non-comment, non-whitespace content, stop searching
|
|
break;
|
|
}
|
|
|
|
next_sibling = ts_node_next_sibling(next_sibling);
|
|
}
|
|
}
|
|
|
|
return "";
|
|
}
|
|
|
|
bool hasFunctionBody(TSNode node) {
|
|
if (strcmp(ts_node_type(node), "function_definition") != 0)
|
|
return false;
|
|
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
if (strcmp(ts_node_type(ts_node_child(node, i)), "compound_statement") ==
|
|
0) {
|
|
return true;
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
|
|
void findFunctions(TSNode node, const char *source_code, uint32_t source_length,
|
|
std::vector<FunctionInfo> &functions) {
|
|
const char *type = ts_node_type(node);
|
|
|
|
if (strcmp(type, "function_definition") == 0 ||
|
|
strcmp(type, "declaration") == 0) {
|
|
std::string func_name = getFunctionName(node, source_code);
|
|
if (!func_name.empty()) {
|
|
std::string address =
|
|
extractAddress(getComment(node, source_code, source_length, false));
|
|
|
|
if (address.empty() && strcmp(type, "function_definition") == 0) {
|
|
address =
|
|
extractAddress(getComment(node, source_code, source_length, true));
|
|
}
|
|
|
|
if (!address.empty()) {
|
|
FunctionInfo func{func_name, address, "",
|
|
strcmp(type, "function_definition") == 0
|
|
? !hasFunctionBody(node)
|
|
: true};
|
|
functions.push_back(func);
|
|
}
|
|
// We'll never nest function declarations
|
|
return;
|
|
} else {
|
|
spdlog::error("Failed to get function name for {}",
|
|
extractNodeText(node, source_code));
|
|
}
|
|
}
|
|
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
findFunctions(ts_node_child(node, i), source_code, source_length,
|
|
functions);
|
|
}
|
|
}
|
|
|
|
std::vector<std::string> readFileList(const std::string &list_file) {
|
|
std::vector<std::string> files;
|
|
std::ifstream file(list_file);
|
|
if (!file.is_open()) {
|
|
spdlog::error("Could not open list file {}", list_file);
|
|
return files;
|
|
}
|
|
|
|
std::string line;
|
|
while (std::getline(file, line)) {
|
|
if (line.empty() || line[0] == '#')
|
|
continue;
|
|
|
|
if (line.find('*') != std::string::npos) {
|
|
spdlog::info("Skipping wildcard pattern: {}", line);
|
|
continue;
|
|
}
|
|
|
|
if (std::filesystem::exists(line)) {
|
|
files.push_back(line);
|
|
} else {
|
|
spdlog::warn("File not found: {}", line);
|
|
}
|
|
}
|
|
return files;
|
|
}
|
|
|
|
bool processFile(const std::string &filepath, DatabaseManager &db) {
|
|
std::ifstream file(filepath);
|
|
if (!file.is_open()) {
|
|
spdlog::error("Could not open file {}", filepath);
|
|
return false;
|
|
}
|
|
|
|
std::string file_content((std::istreambuf_iterator<char>(file)),
|
|
std::istreambuf_iterator<char>());
|
|
|
|
TSParser *parser = ts_parser_new();
|
|
ts_parser_set_language(parser, tree_sitter_cpp());
|
|
|
|
TSTree *tree = ts_parser_parse_string(parser, nullptr, file_content.c_str(),
|
|
file_content.length());
|
|
TSNode root_node = ts_tree_root_node(tree);
|
|
|
|
if (ts_node_is_null(root_node)) {
|
|
spdlog::error("Failed to parse file {}", filepath);
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return false;
|
|
}
|
|
|
|
db.clearEntriesForFile(filepath);
|
|
|
|
std::vector<FunctionInfo> functions;
|
|
findFunctions(root_node, file_content.c_str(), file_content.length(),
|
|
functions);
|
|
|
|
for (auto &func : functions) {
|
|
func.filepath = filepath;
|
|
db.insertFunction(func);
|
|
spdlog::debug("{}: {} @ {} in {}", func.is_import ? "Import" : "Function",
|
|
func.name, func.address, filepath);
|
|
}
|
|
|
|
spdlog::info("Processed {} functions/imports from {}", functions.size(),
|
|
filepath);
|
|
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return true;
|
|
}
|
|
|
|
// Helper function to recursively find identifier in any declarator
|
|
std::string findIdentifierInDeclarator(TSNode node, const char *source_code) {
|
|
const char *type = ts_node_type(node);
|
|
|
|
// If this is an identifier, return it
|
|
if (strcmp(type, "identifier") == 0) {
|
|
return extractNodeText(node, source_code);
|
|
}
|
|
|
|
// Recursively search all children
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
std::string result = findIdentifierInDeclarator(child, source_code);
|
|
if (!result.empty()) {
|
|
return result;
|
|
}
|
|
}
|
|
|
|
return "";
|
|
}
|
|
|
|
std::string getGlobalName(TSNode node, const char *source_code) {
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
const char *type = ts_node_type(child);
|
|
|
|
// Look for any kind of declarator and recursively search for identifier
|
|
if (strcmp(type, "init_declarator") == 0 ||
|
|
strcmp(type, "declarator") == 0 ||
|
|
strcmp(type, "reference_declarator") == 0 ||
|
|
strcmp(type, "pointer_declarator") == 0 ||
|
|
strcmp(type, "parenthesized_declarator") == 0 ||
|
|
strcmp(type, "array_declarator") == 0) {
|
|
std::string name = findIdentifierInDeclarator(child, source_code);
|
|
if (!name.empty()) {
|
|
return name;
|
|
}
|
|
}
|
|
// Direct identifier child
|
|
else if (strcmp(type, "identifier") == 0) {
|
|
return extractNodeText(child, source_code);
|
|
}
|
|
}
|
|
return "";
|
|
}
|
|
|
|
void findGlobals(TSNode node, const char *source_code, uint32_t source_length,
|
|
std::vector<GlobalInfo> &globals) {
|
|
const char *type = ts_node_type(node);
|
|
|
|
// Look for extern declarations
|
|
if (strcmp(type, "declaration") == 0) {
|
|
// Check if this is an extern declaration
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
bool is_extern = false;
|
|
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
if (strcmp(ts_node_type(child), "storage_class_specifier") == 0) {
|
|
std::string storage_class = extractNodeText(child, source_code);
|
|
if (storage_class == "extern") {
|
|
is_extern = true;
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
|
|
if (is_extern) {
|
|
std::string global_name = getGlobalName(node, source_code);
|
|
if (!global_name.empty()) {
|
|
// Look for address comment after the declaration
|
|
std::string address =
|
|
extractAddress(getComment(node, source_code, source_length, false));
|
|
|
|
if (!address.empty()) {
|
|
GlobalInfo global{global_name, address, ""};
|
|
globals.push_back(global);
|
|
}
|
|
} else {
|
|
std::string src = extractNodeText(node, source_code);
|
|
SPDLOG_ERROR("Failed to get global name for {}", src);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Recursively search child nodes
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
findGlobals(ts_node_child(node, i), source_code, source_length, globals);
|
|
}
|
|
}
|
|
|
|
bool processGlobalsFile(const std::string &filepath, DatabaseManager &db) {
|
|
std::ifstream file(filepath);
|
|
if (!file.is_open()) {
|
|
spdlog::error("Could not open file {}", filepath);
|
|
return false;
|
|
}
|
|
|
|
std::string file_content((std::istreambuf_iterator<char>(file)),
|
|
std::istreambuf_iterator<char>());
|
|
|
|
TSParser *parser = ts_parser_new();
|
|
ts_parser_set_language(parser, tree_sitter_cpp());
|
|
|
|
TSTree *tree = ts_parser_parse_string(parser, nullptr, file_content.c_str(),
|
|
file_content.length());
|
|
TSNode root_node = ts_tree_root_node(tree);
|
|
|
|
if (ts_node_is_null(root_node)) {
|
|
spdlog::error("Failed to parse file {}", filepath);
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return false;
|
|
}
|
|
|
|
db.clearGlobalsForFile(filepath);
|
|
|
|
std::vector<GlobalInfo> globals;
|
|
findGlobals(root_node, file_content.c_str(), file_content.length(), globals);
|
|
|
|
for (auto &global : globals) {
|
|
global.filepath = filepath;
|
|
db.insertGlobal(global);
|
|
spdlog::debug("Global: {} @ {} in {}", global.name, global.address,
|
|
filepath);
|
|
}
|
|
|
|
spdlog::info("Processed {} globals from {}", globals.size(), filepath);
|
|
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return true;
|
|
}
|
|
|
|
// Helper function to dump Tree-sitter AST
|
|
void dumpTreeSitterAST(TSNode node, const char *source_code, int depth = 0) {
|
|
std::string indent(depth * 2, ' ');
|
|
const char *type = ts_node_type(node);
|
|
|
|
uint32_t start = ts_node_start_byte(node);
|
|
uint32_t end = ts_node_end_byte(node);
|
|
|
|
// Get the text content for leaf nodes or small nodes
|
|
std::string content;
|
|
if (end - start < 100) { // Only show content for small nodes
|
|
content = extractNodeText(node, source_code);
|
|
// Replace newlines with \n for better readability
|
|
std::regex newline_regex("\n");
|
|
content = std::regex_replace(content, newline_regex, "\\n");
|
|
// Truncate if still too long
|
|
if (content.length() > 50) {
|
|
content = content.substr(0, 47) + "...";
|
|
}
|
|
}
|
|
|
|
if (!content.empty()) {
|
|
spdlog::info("{}{}[{}:{}] \"{}\"", indent, type, start, end, content);
|
|
} else {
|
|
spdlog::info("{}{}[{}:{}]", indent, type, start, end);
|
|
}
|
|
|
|
// Recursively dump children
|
|
uint32_t child_count = ts_node_child_count(node);
|
|
for (uint32_t i = 0; i < child_count; i++) {
|
|
TSNode child = ts_node_child(node, i);
|
|
dumpTreeSitterAST(child, source_code, depth + 1);
|
|
}
|
|
}
|
|
|
|
bool dumpTreeFile(const std::string &filepath) {
|
|
std::ifstream file(filepath);
|
|
if (!file.is_open()) {
|
|
spdlog::error("Could not open file {}", filepath);
|
|
return false;
|
|
}
|
|
|
|
std::string file_content((std::istreambuf_iterator<char>(file)),
|
|
std::istreambuf_iterator<char>());
|
|
|
|
TSParser *parser = ts_parser_new();
|
|
ts_parser_set_language(parser, tree_sitter_cpp());
|
|
|
|
TSTree *tree = ts_parser_parse_string(parser, nullptr, file_content.c_str(),
|
|
file_content.length());
|
|
TSNode root_node = ts_tree_root_node(tree);
|
|
|
|
if (ts_node_is_null(root_node)) {
|
|
spdlog::error("Failed to parse file {}", filepath);
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return false;
|
|
}
|
|
|
|
spdlog::info("=== Tree-sitter AST for {} ===", filepath);
|
|
dumpTreeSitterAST(root_node, file_content.c_str());
|
|
spdlog::info("=== End of AST dump ===");
|
|
|
|
ts_tree_delete(tree);
|
|
ts_parser_delete(parser);
|
|
return true;
|
|
}
|
|
|
|
bool processDuplicates(DatabaseManager &db) {
|
|
spdlog::info("=== Checking for duplicate addresses ===");
|
|
bool found_address_duplicates = db.checkDuplicateAddresses();
|
|
|
|
spdlog::info("=== Checking for duplicate names ===");
|
|
bool found_name_duplicates = db.checkDuplicateNames();
|
|
|
|
if (!found_address_duplicates && !found_name_duplicates) {
|
|
spdlog::info("No duplicates found in the database.");
|
|
return true;
|
|
}
|
|
|
|
if (found_address_duplicates) {
|
|
spdlog::error("Found duplicate addresses in the database!");
|
|
}
|
|
|
|
if (found_name_duplicates) {
|
|
spdlog::error("Found duplicate names in the database!");
|
|
}
|
|
|
|
return false; // Return false to indicate errors were found
|
|
}
|
|
|
|
int main(int argc, char *argv[]) {
|
|
// Initialize spdlog
|
|
auto console = spdlog::stdout_color_mt("console");
|
|
spdlog::set_default_logger(console);
|
|
spdlog::set_level(spdlog::level::info); // Default to info level
|
|
spdlog::set_pattern("[%H:%M:%S] [%^%l%$] %v");
|
|
|
|
CLI::App app{"C++ Function/Global Parser - Extracts function addresses or "
|
|
"global variable addresses from C++ files"};
|
|
|
|
std::vector<std::string> input_files;
|
|
std::string list_file;
|
|
std::string db_path = "gh.db";
|
|
std::string mode = "functions";
|
|
std::string log_file = "";
|
|
bool verbose = false;
|
|
|
|
app.add_option("files", input_files,
|
|
"Input C++ files to parse (supports @listfile.txt syntax)");
|
|
app.add_option("-l,--list", list_file,
|
|
"File containing list of files to process");
|
|
app.add_option("-d,--database", db_path, "SQLite database path")
|
|
->default_val("gh.db");
|
|
app.add_option("-m,--mode", mode,
|
|
"Processing mode: 'functions', 'globals', 'duplicates', or "
|
|
"'dump-tree'")
|
|
->default_val("functions")
|
|
->check(
|
|
CLI::IsMember({"functions", "globals", "duplicates", "dump-tree"}));
|
|
app.add_flag("-v,--verbose", verbose, "Enable verbose logging (debug level)");
|
|
app.add_flag("--log-file", log_file, "Enable logging to file");
|
|
|
|
CLI11_PARSE(app, argc, argv);
|
|
|
|
// Set log level based on verbose flag
|
|
if (verbose) {
|
|
spdlog::set_level(spdlog::level::debug);
|
|
}
|
|
|
|
spdlog::set_pattern(std::string("[%^%l%$] %v"));
|
|
|
|
if (!log_file.empty()) {
|
|
auto log_sink =
|
|
std::make_shared<spdlog::sinks::basic_file_sink_mt>(log_file, true);
|
|
spdlog::get("console")->sinks().push_back(log_sink);
|
|
}
|
|
|
|
std::vector<std::string> files_to_process;
|
|
bool needFiles = mode != "duplicates";
|
|
|
|
if (needFiles) {
|
|
if (!list_file.empty()) {
|
|
auto list_files = readFileList(list_file);
|
|
files_to_process.insert(files_to_process.end(), list_files.begin(),
|
|
list_files.end());
|
|
}
|
|
|
|
for (const auto &input : input_files) {
|
|
if (input.starts_with("@")) {
|
|
auto list_files = readFileList(input.substr(1));
|
|
files_to_process.insert(files_to_process.end(), list_files.begin(),
|
|
list_files.end());
|
|
} else if (std::filesystem::exists(input)) {
|
|
files_to_process.push_back(input);
|
|
} else {
|
|
spdlog::warn("File not found: {}", input);
|
|
}
|
|
}
|
|
|
|
if (files_to_process.empty()) {
|
|
spdlog::error("No files to process. Use --help for usage information.");
|
|
return 1;
|
|
}
|
|
}
|
|
|
|
try {
|
|
int processed_count = 0;
|
|
|
|
// For dump-tree mode, we don't need database operations
|
|
if (mode == "dump-tree") {
|
|
for (const auto &filepath : files_to_process) {
|
|
spdlog::info("=== Processing: {} ===", filepath);
|
|
if (dumpTreeFile(filepath)) {
|
|
processed_count++;
|
|
}
|
|
}
|
|
} else if (mode == "duplicates") {
|
|
DatabaseManager db(db_path);
|
|
// For duplicates mode, we only check the database, no file processing
|
|
spdlog::info("=== Checking database for duplicates ===");
|
|
bool has_duplicates = !processDuplicates(db);
|
|
spdlog::info("=== Summary ===");
|
|
spdlog::info("Mode: {}", mode);
|
|
spdlog::info("Database: {}", db_path);
|
|
return has_duplicates ? 1 : 0; // Return 1 if duplicates found, 0 if none
|
|
} else {
|
|
DatabaseManager db(db_path);
|
|
|
|
const size_t batch_size = 50;
|
|
size_t current_batch = 0;
|
|
|
|
db.beginTransaction();
|
|
|
|
for (const auto &filepath : files_to_process) {
|
|
spdlog::info("=== Processing: {} ===", filepath);
|
|
bool success = false;
|
|
if (mode == "functions") {
|
|
success = processFile(filepath, db);
|
|
} else if (mode == "globals") {
|
|
success = processGlobalsFile(filepath, db);
|
|
}
|
|
|
|
if (success)
|
|
processed_count++;
|
|
|
|
if (++current_batch >= batch_size) {
|
|
db.commitTransaction();
|
|
spdlog::info("Committed batch of {} files to database",
|
|
current_batch);
|
|
db.beginTransaction();
|
|
current_batch = 0;
|
|
}
|
|
}
|
|
|
|
if (current_batch > 0) {
|
|
db.commitTransaction();
|
|
spdlog::info("Committed final batch of {} files to database",
|
|
current_batch);
|
|
}
|
|
}
|
|
|
|
spdlog::info("=== Summary ===");
|
|
spdlog::info("Processed {} files successfully", processed_count);
|
|
spdlog::info("Mode: {}", mode);
|
|
if (mode != "dump-tree") {
|
|
spdlog::info("Database saved to: {}", db_path);
|
|
}
|
|
|
|
} catch (const std::exception &e) {
|
|
spdlog::error("Database error: {}", e.what());
|
|
return 1;
|
|
}
|
|
|
|
return 0;
|
|
}
|