Add: Baseline C++ class

ashvardanian · Dec 21, 2023 · 4bec620 · 4bec620
1 parent 44ef989
commit 4bec620
Show file tree

Hide file tree

Showing 11 changed files with 970 additions and 325 deletions.
diff --git a/.vscode/launch.json b/.vscode/launch.json
@@ -5,10 +5,10 @@
     "version": "0.2.0",
     "configurations": [
         {
-            "name": "Test",
+            "name": "Debug Unit Tests",
             "type": "cppdbg",
             "request": "launch",
-            "program": "${workspaceFolder}/build_debug/stringzilla_bench",
+            "program": "${workspaceFolder}/build_debug/stringzilla_test_substring",
             "cwd": "${workspaceFolder}",
             "environment": [
                 {
@@ -18,11 +18,33 @@
             ],
             "stopAtEntry": false,
             "linux": {
-                "preLaunchTask": "Linux Build C++ Test Debug",
+                "preLaunchTask": "Build for Linux: Debug",
                 "MIMode": "gdb"
             },
             "osx": {
-                "preLaunchTask": "MacOS Build C++ Test Debug",
+                "preLaunchTask": "Build for MacOS: Debug",
+                "MIMode": "lldb"
+            }
+        },
+        {
+            "name": "Debug Benchmarks",
+            "type": "cppdbg",
+            "request": "launch",
+            "program": "${workspaceFolder}/build_debug/stringzilla_bench_substring",
+            "cwd": "${workspaceFolder}",
+            "environment": [
+                {
+                    "name": "ASAN_OPTIONS",
+                    "value": "detect_leaks=0:atexit=1:strict_init_order=1:strict_string_checks=1"
+                }
+            ],
+            "stopAtEntry": false,
+            "linux": {
+                "preLaunchTask": "Build for Linux: Debug",
+                "MIMode": "gdb"
+            },
+            "osx": {
+                "preLaunchTask": "Build for MacOS: Debug",
                 "MIMode": "lldb"
             }
         }

diff --git a/.vscode/tasks.json b/.vscode/tasks.json
@@ -2,32 +2,32 @@
     "version": "2.0.0",
     "tasks": [
         {
-            "label": "Linux Build C++ Test Debug",
-            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DCMAKE_CXX_COMPILER=gcc-12 -DCMAKE_CXX_COMPILER=g++-12 -DCMAKE_BUILD_TYPE=Debug -B ./build_debug && make -C ./build_debug",
+            "label": "Build for Linux: Debug",
+            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DSTRINGZILLA_BUILD_BENCHMARK=1 -DCMAKE_CXX_COMPILER=gcc-12 -DCMAKE_CXX_COMPILER=g++-12 -DCMAKE_BUILD_TYPE=Debug -B ./build_debug && make -C ./build_debug",
             "args": [],
             "type": "shell",
             "problemMatcher": [
                 "$gcc"
             ]
         },
         {
-            "label": "Linux Build C++ Test Release",
-            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DCMAKE_CXX_COMPILER=gcc-12 -DCMAKE_CXX_COMPILER=g++-12 -DCMAKE_BUILD_TYPE=RelWithDebInfo -B ./build_release && make -C ./build_release",
+            "label": "Build for Linux: Release",
+            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DSTRINGZILLA_BUILD_BENCHMARK=1 -DCMAKE_CXX_COMPILER=gcc-12 -DCMAKE_CXX_COMPILER=g++-12 -DCMAKE_BUILD_TYPE=RelWithDebInfo -B ./build_release && make -C ./build_release",
             "args": [],
             "type": "shell",
             "problemMatcher": [
                 "$gcc"
             ]
         },
         {
-            "label": "MacOS Build C++ Test Debug",
-            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DCMAKE_C_COMPILER=/opt/homebrew/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/opt/homebrew/opt/llvm/bin/clang++ -DCMAKE_BUILD_TYPE=Debug -B ./build_debug && make -C ./build_debug",
+            "label": "Build for MacOS: Debug",
+            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DSTRINGZILLA_BUILD_BENCHMARK=1 -DCMAKE_C_COMPILER=/opt/homebrew/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/opt/homebrew/opt/llvm/bin/clang++ -DCMAKE_BUILD_TYPE=Debug -B ./build_debug && make -C ./build_debug",
             "args": [],
             "type": "shell",
         },
         {
-            "label": "MacOS Build C++ Test Release",
-            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DCMAKE_C_COMPILER=/opt/homebrew/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/opt/homebrew/opt/llvm/bin/clang++ -DCMAKE_BUILD_TYPE=RelWithDebInfo -B ./build_release && make -C ./build_release",
+            "label": "Build for MacOS: Release",
+            "command": "cmake -DSTRINGZILLA_BUILD_TEST=1 -DSTRINGZILLA_BUILD_BENCHMARK=1 -DCMAKE_C_COMPILER=/opt/homebrew/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/opt/homebrew/opt/llvm/bin/clang++ -DCMAKE_BUILD_TYPE=RelWithDebInfo -B ./build_release && make -C ./build_release",
             "args": [],
             "type": "shell"
         }

diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -71,35 +71,40 @@ if(STRINGZILLA_INSTALL)
     DESTINATION ${STRINGZILLA_INCLUDE_INSTALL_DIR})
 endif()
 
-if(${STRINGZILLA_BUILD_TEST} OR ${STRINGZILLA_BUILD_BENCHMARK})
-  add_executable(stringzilla_bench scripts/bench_substring.cpp)
-  target_link_libraries(stringzilla_bench PRIVATE ${STRINGZILLA_TARGET_NAME})
-  set_target_properties(stringzilla_bench PROPERTIES RUNTIME_OUTPUT_DIRECTORY
+if(${CMAKE_VERSION} VERSION_EQUAL 3.13 OR ${CMAKE_VERSION} VERSION_GREATER
+  3.13)
+  include(CTest)
+  enable_testing()
+endif()
+
+# Function to set compiler-specific flags
+function(set_compiler_flags target)
+  target_link_libraries(${target} PRIVATE ${STRINGZILLA_TARGET_NAME})
+  set_target_properties(${target} PROPERTIES RUNTIME_OUTPUT_DIRECTORY
     ${CMAKE_BINARY_DIR})
-  target_link_options(stringzilla_bench PRIVATE
-    "-Wl,--unresolved-symbols=ignore-all")
 
-  # Check for compiler and set flags for stringzilla_bench
   if(${CMAKE_CXX_COMPILER_ID} STREQUAL "GNU" OR ${CMAKE_CXX_COMPILER_ID} MATCHES "Clang")
-    # Set -march=native and -fmax-errors=1 for all build types
-    target_compile_options(stringzilla_bench PRIVATE "-march=native")
-    target_compile_options(stringzilla_bench PRIVATE "-fmax-errors=1")
-
-    # Set -O3 for Release build, and -g for Debug and RelWithDebInfo
-    target_compile_options(stringzilla_bench PRIVATE
+    target_compile_options(${target} PRIVATE "-march=native")
+    target_compile_options(${target} PRIVATE "-fmax-errors=1")
+    target_compile_options(${target} PRIVATE
       "$<$<CONFIG:Release>:-O3>"
       "$<$<OR:$<CONFIG:Debug>,$<CONFIG:RelWithDebInfo>>:-g>")
   elseif(${CMAKE_CXX_COMPILER_ID} STREQUAL "Intel")
-    # Intel specific flags
-    target_compile_options(stringzilla_bench PRIVATE "-xHost")
+    target_compile_options(${target} PRIVATE "-xHost")
   elseif(${CMAKE_CXX_COMPILER_ID} STREQUAL "MSVC")
     # MSVC specific flags or other settings
   endif()
+endfunction()
 
-  if(${CMAKE_VERSION} VERSION_EQUAL 3.13 OR ${CMAKE_VERSION} VERSION_GREATER
-    3.13)
-    include(CTest)
-    enable_testing()
-    add_test(NAME stringzilla_bench COMMAND stringzilla_bench)
-  endif()
+if(${STRINGZILLA_BUILD_BENCHMARK})
+  add_executable(stringzilla_bench_substring scripts/bench_substring.cpp)
+  set_compiler_flags(stringzilla_bench_substring)
+  add_test(NAME stringzilla_bench_substring COMMAND stringzilla_bench_substring)
+endif()
+
+if(${STRINGZILLA_BUILD_TEST})
+  # Test target
+  add_executable(stringzilla_test_substring scripts/test_substring.cpp)
+  set_compiler_flags(stringzilla_test_substring)
+  add_test(NAME stringzilla_test_substring COMMAND stringzilla_test_substring)
 endif()
diff --git a/README.md b/README.md
@@ -9,6 +9,17 @@ StringZilla is the Godzilla of string libraries, searching, splitting, sorting,
 - ✅ [Radix](https://en.wikipedia.org/wiki/Radix_sort)-like sorting faster than C++ `std::sort`
 - ✅ [Memory-mapping](https://en.wikipedia.org/wiki/Memory-mapped_file) to work with larger-than-RAM datasets
 
+Putting this into a table:
+
+| Feature \ Library    |  STL | LibC |      StringZilla |
+| :------------------- | ---: | ---: | ---------------: |
+| Substring Search     |      |      |                  |
+| Reverse Order Search |      |    ❌ |                  |
+| Fuzzy Search         |    ❌ |    ❌ |                  |
+| Edit Distance        |    ❌ |    ❌ |                  |
+| Interface            |  C++ |    C | C , C++ , Python |
+
+
 Who is this for?
 
 - you want to process strings faster than default strings in Python, C, or C++
@@ -22,7 +33,6 @@ Limitations:
 - Assumes ASCII or UTF-8 encoding
 - Assumes 64-bit address space
 
-
 This library saved me tens of thousands of dollars pre-processing large datasets for machine learning, even on the scale of a single experiment.
 So if you want to process the 6 Billion images from [LAION](https://laion.ai/blog/laion-5b/), or the 250 Billion web pages from the [CommonCrawl](https://commoncrawl.org/), or even just a few million lines of server logs, and haunted by Python's `open(...).readlines()` and `str().splitlines()` taking forever, this should help 😊
 
@@ -216,15 +226,15 @@ Running benchmarks:
 ```sh
 cmake -DCMAKE_BUILD_TYPE=Release -DSTRINGZILLA_BUILD_TEST=1 -B ./build_release
 cmake --build build_release --config Release
-./build_release/stringzilla_bench
+./build_release/stringzilla_bench_substring
 ```
 
 Running tests:
 
 ```sh
 cmake -DCMAKE_BUILD_TYPE=Debug -DSTRINGZILLA_BUILD_TEST=1 -B ./build_debug
 cmake --build build_debug --config Debug
-./build_debug/stringzilla_bench
+./build_debug/stringzilla_bench_substring
 ```
 
 On MacOS it's recommended to use non-default toolchain:
@@ -240,7 +250,7 @@ cmake -B ./build_release \
     -DSTRINGZILLA_USE_OPENMP=1 \
     -DSTRINGZILLA_BUILD_TEST=1 \
     && \
-    make -C ./build_release -j && ./build_release/stringzilla_bench
+    make -C ./build_release -j && ./build_release/stringzilla_bench_substring
 ```
 
 ## License 📜
@@ -257,3 +267,34 @@ If you like this project, you may also enjoy [USearch][usearch], [UCall][ucall],
 [ustore]: https://github.com/unum-cloud/ustore
 [simsimd]: https://github.com/ashvardanian/simsimd
 [tenpack]: https://github.com/ashvardanian/tenpack
+
+
+# The weirdest interfaces of C++23 strings:
+
+## Third `std::basic_string_view<CharT,Traits>::find`
+
+constexpr size_type find( basic_string_view v, size_type pos = 0 ) const noexcept;
+(1)	(since C++17)
+constexpr size_type find( CharT ch, size_type pos = 0 ) const noexcept;
+(2)	(since C++17)
+constexpr size_type find( const CharT* s, size_type pos, size_type count ) const;
+(3)	(since C++17)
+constexpr size_type find( const CharT* s, size_type pos = 0 ) const;
+(4)	(since C++17)
+
+
+## HTML Parsing
+
+```txt
+<tag>       Isolated tag start
+<tag\w      Tag start with attributes
+<tag/>      Self-closing tag
+</tag>      Tag end
+```
+
+In any case, the tag name is always followed by whitespace, `/` or `>`.
+And is always preceded by whitespace. `/` or `<`.
+
+Important distinctions between XML and HTML:
+
+- XML does not truncate multiple white-spaces, while HTML does.