Setup v1 of borded spam parser

borded_cpp/.gitignore

@@ -0,0 +1,97 @@
+*.d
+*.slo
+*.lo
+*.o
+*.obj
+*.gch
+*.pch
+*.so
+*.dylib
+*.dll
+*.mod
+*.smod
+*.lai
+*.la
+*.a
+*.lib
+*.exe
+*.out
+*.app
+.idea/**/workspace.xml
+.idea/**/tasks.xml
+.idea/**/usage.statistics.xml
+.idea/**/dictionaries
+.idea/**/shelf
+.idea/**/aws.xml
+.idea/**/contentModel.xml
+.idea/**/dataSources/
+.idea/**/dataSources.ids
+.idea/**/dataSources.local.xml
+.idea/**/sqlDataSources.xml
+.idea/**/dynamic.xml
+.idea/**/uiDesigner.xml
+.idea/**/dbnavigator.xml
+.idea/**/gradle.xml
+.idea/**/libraries
+.idea
+cmake-build-*/
+.idea/**/mongoSettings.xml
+*.iws
+out/
+.idea_modules/
+atlassian-ide-plugin.xml
+.idea/replstate.xml
+.idea/sonarlint/
+com_crashlytics_export_strings.xml
+crashlytics.properties
+crashlytics-build.properties
+fabric.properties
+.idea/httpRequests
+.idea/caches/build_file_checksums.ser
+*~
+.fuse_hidden*
+.directory
+.Trash-*
+.nfs*
+CMakeLists.txt.user
+CMakeCache.txt
+CMakeFiles
+CMakeScripts
+Testing
+Makefile
+cmake_install.cmake
+install_manifest.txt
+compile_commands.json
+CTestTestfile.cmake
+_deps
+CMakeUserPresets.json
+Thumbs.db
+Thumbs.db:encryptable
+ehthumbs.db
+ehthumbs_vista.db
+*.stackdump
+[Dd]esktop.ini
+$RECYCLE.BIN/
+*.cab
+*.msi
+*.msix
+*.msm
+*.msp
+*.lnk
+.DS_Store
+.AppleDouble
+.LSOverride
+Icon
+._*
+.DocumentRevisions-V100
+.fseventsd
+.Spotlight-V100
+.TemporaryItems
+.Trashes
+.VolumeIcon.icns
+.com.apple.timemachine.donotpresent
+.AppleDB
+.AppleDesktop
+Network Trash Folder
+Temporary Items
+.apdisk

borded_cpp/CMakeLists.txt

@@ -0,0 +1,21 @@
+cmake_minimum_required(VERSION 3.30)
+project(isspam)
+
+set(CMAKE_CXX_STANDARD 26)
+if (MSVC)
+    add_compile_options(/W4)
+    add_compile_options(/WX)
+    add_compile_options(/external:anglebrackets)
+    add_compile_options(/external:W0)
+    add_compile_options(/wd4100)
+    add_compile_options(/wd5050)
+    add_definitions(-DWIN32_LEAN_AND_MEAN -DVC_EXTRALEAN)
+    add_compile_definitions(WIN32_LEAN_AND_MEAN NOMINMAX)
+else ()
+    add_compile_options(-Wall)
+    add_compile_options(-Wextra)
+    add_compile_options(-Wpedantic)
+    add_compile_options(-Werror)
+endif ()
+
+add_executable(${PROJECT_NAME} src/main.cpp)

borded_cpp/src/main.cpp

@@ -0,0 +1,140 @@
+import std;
+
+const std::vector<std::wstring_view> BAD_WORDS = {
+    L"recovery",
+    L"techie",
+    L"http",
+    L"https",
+    L"digital",
+    L"hack",
+    L"::",
+    L"//",
+    L"com",
+    L"@",
+    L"crypto",
+    L"bitcoin",
+    L"wallet",
+    L"hacker",
+    L"welcome",
+    L"whatsapp",
+    L"email",
+    L"cryptocurrency",
+    L"stolen",
+    L"freeze",
+    L"quick",
+    L"crucial",
+    L"tracing",
+    L"scammers",
+    L"expers",
+    L"hire",
+    L"century",
+    L"transaction",
+    L"essential",
+    L"managing",
+    L"contact",
+    L"contacting",
+    L"understanding",
+    L"assets",
+    L"funds",
+};
+
+
+struct AnalysisResult {
+    std::size_t totalWordCount = 0;
+    std::size_t totalCapitalizedCount = 0;
+    std::size_t totalSentenceCount = 0;
+    std::size_t totalNumberCount = 0;
+    std::size_t totalForbiddenCount = 0;
+
+    operator std::string() const {
+        return std::format(
+            "Word Count: {}\nCapitalized Count: {}\nSentence Count: {}\nNumber Count: {}\nForbidden Count: {}",
+            totalWordCount, totalCapitalizedCount, totalSentenceCount, totalNumberCount, totalForbiddenCount
+            );
+    }
+
+    AnalysisResult operator+(const AnalysisResult &other) const {
+        return {
+            totalWordCount + other.totalWordCount,
+            totalCapitalizedCount + other.totalCapitalizedCount,
+            totalSentenceCount + other.totalSentenceCount,
+            totalNumberCount + other.totalNumberCount,
+            totalForbiddenCount + other.totalForbiddenCount
+        };
+    }
+};
+
+void check_word(const std::wstring &word, std::size_t &forbiddenCount) {
+    if (std::find(BAD_WORDS.begin(), BAD_WORDS.end(), word) != BAD_WORDS.end()) {
+        forbiddenCount++;
+    }
+}
+
+AnalysisResult parseFile(const std::string_view &filename) {
+    std::wifstream file;
+
+    file.open(filename);
+    if (!file.is_open()) {
+        std::println("File doesn't exist: {}", filename);
+        return { };
+    }
+
+    AnalysisResult result{ };
+
+    bool inWord = false;
+    wchar_t c;
+
+    std::wstring word;
+    while (file.get(c)) {
+        if (c == '.') {
+            result.totalSentenceCount++;
+        }
+
+        if (std::isspace(c) || c == '.' || c == '?' || c == '!' || c == ';' || c == ':') {
+            inWord = false;
+
+            check_word(word, result.totalForbiddenCount);
+            word.clear();
+            continue;
+        } else {
+            if (!inWord) {
+                result.totalWordCount++;
+                if (std::isupper(c)) {
+                    result.totalCapitalizedCount++;
+                }
+            }
+            inWord = true;
+
+            if (std::isdigit(c)) {
+                result.totalNumberCount++;
+                while (file.get(c) && std::isdigit(c)) { }
+                file.unget();
+                continue;
+            }
+
+            word.push_back(c);
+        }
+    };
+
+    if (!word.empty()) {
+        check_word(word, result.totalForbiddenCount);
+    }
+
+    file.close();
+
+    return result;
+}
+
+int main(const int argc, char *argv[]) {
+    if (argc < 2) {
+        std::println("Usage: {} <file1> <file2> ... <fileN>", argv[0]);
+        return 1;
+    }
+
+    const AnalysisResult result = std::transform_reduce(std::execution::par, std::next(argv), argv + argc, AnalysisResult{ }, std::plus{ },
+                                                        parseFile
+        );
+
+    std::println("{}", std::string(result));
+    return 0;
+}