添加了编译原理h的内容

大三上/编译原理和技术(H)/lab2-pre.md

@@ -0,0 +1,29 @@
+lab2-预热实验
+0.配置环境
+    在之前实验的基础上，将llvm的路径llvm-install/bin添加到PATH中.
+1.人工翻译
+    通过观察clang -S -emit-llvm输出的结果，推理并理解llvm IR翻译过程.
+    首先，每个函数对应一个独立的计数过程.每个label占用一个计数值.
+    一个函数的开始，参数先占据计数变量，之后其实label占用一个，之后存储return值占用一个，之后局部变量等开始占用.
+    正如上课所讲，每个基本块(BasicBlock)对应没有分支的,顺序的过程.
+    具体指令有store load add new br等。每个基本块的结束(除了最后一个基本块)都需要一个br指令标志结束.
+    对于多个return情况,将多个return值都存储在同一个地方,并跳转到最后一个基本块读取return值并返回.
+    由此,得到了手工翻译的结果.
+2.编写gen_fib.cpp
+    这一步,要求利用IRBuilder编写一个gen_fib.cpp,来构建LLVM IR并打印.
+    1.首先,要理解通过IRBuilder进行代码生成的系统框架,理解BasicBuilder.这一点可以通过阅读Kaleidoscope的例子来初步理解;
+    2.其次要清楚我们所需要用到的指令所对应的API.这一点可以通过文档,和VS的AutoComplete功能来推测.
+    3.另外还要清楚一些运用API的细节,比如如何获得函数参数,如何获得整形常量,如何store,load,align等等.
+    4.这次还要编写Makefile或者CMakeLists.txt.我都试了一下,发现Makefile在子程序返回的不是0的时候会误报为error,并且终止make,这里我的解决方式是中间通过echo生成一个.sh文件,运行.sh文件来执行lli,输出返回结果.而至于CMakeLists.txt我修改之后爆出了一长串的错误,精力有限,没有继续.希望下次能好好学一下cmake.
+
+3.遇到的问题：
+    本次实验，我很早开始，没有拖到ddl，是基于上一次实验超时的改变，很好.不过还是遇到了一些意想不到的问题:
+```1.实验环境崩溃.
+    不知是什么原因,ubuntu虚拟机突然在实验过程中磁盘损坏,通过v2v工具也没能修复.
+    重装新ubuntu之后没更新完再次崩溃.通过磁盘检测工具发现物理磁盘没有损坏,推测是VMware的问题.
+    在咨询助教之后决定迁移环境到wsl中.还好私钥已备份.
+2.阅读文档能力较差.
+    实验中找了很久INSTRUCTION对应的函数,对应的描述很少,只能通过猜测.这里收李嘉豪同学的启发,知道了IntellliSence对变成效率的重要性,安装了VSCode,VS等.最后由于迁移到了wsl,通过VS写代码,变成效率得到了很大的提升.提升的地方主要体现在IntelliSence和AutoComplete上.
+    更好的生产工具,意味着更强的生产力.这是我这次试验最大的收获.
+    同时,文档阅读能力,阅读习惯需要加强.
+```

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/README.md

@@ -0,0 +1,6 @@
+# Compiler at USTC in 2017 Fall
+
+This is a repository to store project toolkit for Compiler class at USTC in 2017 Fall.
+
+[Course homepage](http://staff.ustc.edu.cn/~yuzhang/compiler)
+

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/antlr-example/.gitignore

@@ -0,0 +1,4 @@
+*
+!expr.g4
+!expr_calculator.py
+!.gitignore

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/antlr-example/expr.g4

@@ -0,0 +1,14 @@
+grammar expr;
+
+Identifier: [a-zA-Z_] [a-zA-Z_0-9]*;
+Number: [1-9] [0-9]*;
+Plus: '+';
+Multiply: '*';
+WhiteSpace: [ \t\n\r]+ -> skip;
+
+expr: expr Multiply expr # Mult
+    | expr Plus expr # Plus
+    | '(' expr ')' # Brac
+    | Identifier # Id
+    | Number # Num
+    ;

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/antlr-example/expr_calculator.py

@@ -0,0 +1,52 @@
+'''Expression recognizer.
+Handles expression described in expr.g4.
+Grammar file should be compiled by antlr4 with option '-Dlanguage=Python3' before executing this.
+Module 'antlr4' is required.
+'''
+import antlr4
+import exprLexer
+import exprParser
+import exprListener
+from typing import Mapping
+
+
+class Listener(exprListener.exprListener):
+    '''Listener doing calculation based on recognized input.
+    '''
+
+    def __init__(self, var_value_source: Mapping[str, int]):
+        self.var_value_source = var_value_source
+
+    def exitMult(self, ctx: exprParser.exprParser.MultContext):
+        ctx.value = ctx.getChild(0).value * ctx.getChild(2).value
+
+    def exitNum(self, ctx: exprParser.exprParser.NumContext):
+        ctx.value = int(str(ctx.getChild(0)))
+
+    def exitId(self, ctx: exprParser.exprParser.IdContext):
+        ctx.value = self.var_value_source[str(ctx.getChild(0))]
+
+    def exitPlus(self, ctx: exprParser.exprParser.PlusContext):
+        ctx.value = ctx.getChild(0).value + ctx.getChild(2).value
+
+    def exitBrac(self, ctx: exprParser.exprParser.BracContext):
+        ctx.value = ctx.getChild(1).value
+
+
+class LazyInputDict(dict):
+    '''A lazy dictionary asking for input when a new item is queried.'''
+
+    def __getitem__(self, key):
+        try:
+            return dict.__getitem__(self, key)
+        except KeyError:
+            self[key] = int(
+                input('Please enter value for variable \'{}\': '.format(key)))
+        return dict.__getitem__(self, key)
+
+
+if __name__ == '__main__':
+    PARSER = exprParser.exprParser(antlr4.CommonTokenStream(exprLexer.exprLexer(
+        antlr4.InputStream(input('Please enter a expression: ')))))
+    PARSER.addParseListener(Listener(LazyInputDict()))
+    print(PARSER.expr().value)

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/c1recognizer/CMakeLists.txt

@@ -0,0 +1,71 @@
+cmake_minimum_required(VERSION 3.5)
+
+add_definitions(-DFULL_IMPL)
+
+list(APPEND CMAKE_MODULE_PATH ${CMAKE_CURRENT_SOURCE_DIR}/cmake)
+
+# compiler must be 11 or 14
+set(CMAKE_CXX_STANDARD 14)
+
+# set variable pointing to the antlr tool that supports C++
+set(ANTLR4CPP_JAR_LOCATION "NOT-FOUND" CACHE STRING "ANTLR v4 JAR file location. Used by antlr4cpp.")
+
+# add external build for antlrcpp
+include(ExternalAntlr4Cpp)
+
+# add antrl4cpp artifacts to project environment
+include_directories(${ANTLR4CPP_INCLUDE_DIRS})
+link_directories(${ANTLR4CPP_LIBS})
+message(STATUS "Found antlr4cpp libs: ${ANTLR4CPP_LIBS} and includes: ${ANTLR4CPP_INCLUDE_DIRS} ")
+
+# Call macro to add lexer and grammar to your build dependencies.
+antlr4cpp_process_grammar(c1_recognizer c1_recognizer
+  ${CMAKE_CURRENT_SOURCE_DIR}/grammar/C1Lexer.g4
+  ${CMAKE_CURRENT_SOURCE_DIR}/grammar/C1Parser.g4)
+
+include(ExternalProject)
+ExternalProject_Add(rapidjson
+  PREFIX ${CMAKE_BINARY_DIR}/externals/rapidjson
+  GIT_REPOSITORY https://github.com/Tencent/rapidjson.git
+  GIT_TAG v1.1.0
+  GIT_SUBMODULES ""
+  CONFIGURE_COMMAND ""
+  BUILD_COMMAND ""
+  UPDATE_COMMAND ""
+  INSTALL_COMMAND "")
+ExternalProject_Get_Property(rapidjson DOWNLOAD_DIR)
+set(rapidjson_include_dirs ${DOWNLOAD_DIR}/rapidjson/include)
+
+# include generated files in project environment
+include_directories(${antlr4cpp_include_dirs_c1_recognizer})
+include_directories(include)
+
+# add generated grammar to c1_recognizer binary target
+add_library(c1recognizer
+  src/error_listener.cpp
+  src/error_reporter.cpp
+  src/syntax_tree.cpp
+  src/syntax_tree_builder.cpp
+  src/recognizer.cpp
+  ${antlr4cpp_src_files_c1_recognizer})
+add_dependencies(c1recognizer antlr4cpp antlr4cpp_generation_c1_recognizer)
+target_link_libraries(c1recognizer antlr4-runtime)
+
+include_directories(${rapidjson_include_dirs})
+add_executable(c1r_test test/main.cpp)
+add_dependencies(c1r_test c1recognizer rapidjson)
+target_link_libraries(c1r_test c1recognizer)
+
+install(
+  TARGETS c1recognizer
+  RUNTIME DESTINATION bin
+  LIBRARY DESTINATION lib
+  ARCHIVE DESTINATION lib)
+
+install(DIRECTORY include/c1recognizer DESTINATION include)
+
+install(DIRECTORY ${ANTLR4CPP_LIBS}/ DESTINATION lib)
+
+configure_file(c1recognizer-config.cmake.in c1recognizer-config.cmake @ONLY)
+
+install(FILES ${CMAKE_CURRENT_BINARY_DIR}/c1recognizer-config.cmake DESTINATION cmake)

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/c1recognizer/README.md

@@ -0,0 +1,83 @@
+# C1 Language Recognizer (Reference Version)
+
+A C1 language recognizer library built on ANTLR v4, which constructs an abstract syntax tree on given source input.
+
+## Build Requirement
+
+* CMake 3.5 or later
+* Any GCC compatible C++ compiler, C++ 11 must be supported
+* Any java runtime environment (JRE), preferring version for Java 8
+* pkg-config
+* uuid-dev
+
+If you have the patience to configure those correctly on Windows, using Windows as the expirement platform is okay.
+Otherwise please choose one of the easy-to-use Linux distributions. (Note: Windows Subsystem for Linux will work fine)
+
+For example, installing those in Ubuntu with `apt`:
+
+```bash
+sudo apt install cmake g++ openjdk-8-jre-headless pkg-config uuid-dev
+```
+
+Also, it would take a while to build it for the first time. This is because it needs to build the dependency antlr4cpp.
+And parallel building won't work on it; it seems that CMake is the one to be blamed.
+
+Besides these, you're also expected to download an ANTLR complete jar file. This is expected throughout this project.
+You may find it from [ANTLR v4 Official Site](http://www.antlr.org/).
+
+## Building
+
+For first time building:
+```bash
+mkdir build
+cd build
+cmake -DCMAKE_BUILD_TYPE=Debug -DANTLR4CPP_JAR_LOCATION=/path/to/your/antlr.jar ..
+make
+cmake ..
+make -j
+```
+
+After that, each time you have any modification, simply do
+```bash
+make -j
+```
+
+To avoid UI unresponding, you may want to replace `-j` option with `-jN`, where `N` is parallel task number you expecting.
+
+## ANTLRv4 Usage
+
+For the building work, I've setup the whole CMake project that will work easily as described above.
+
+And for the most special, magic and wonderful part of ANTLR, visualization, follow these:
+
+* Rail-road or ADT graphs: Using Visual Studio Code will be the best choice. Have ANTLR plugin installed, you'll gonna
+  find those options when you right-click on the editor.
+* Testing grammars (including parse tree visualization):
+  > If you'd like to see GUI rather than console-printed parse tree, you'll need to first install JDK on your host
+  > system (Windows or Linux, DO NOT use WSL here if you don't have sufficient knowledge on configuring X11
+  > environment, which is needed for the GUI to show up). However if you don't want the GUI version visualized parse
+  > tree, just ignore this tip.
+  
+  First of all, you need to ensure `.` and ANTLR v4 JAR file exists in your CLASSPATH.
+  * In PowerShell, it is `$env:CLASSPATH`.
+  * In Linux-like shells, it is `$CLASSPATH`.
+
+  Then, use ANTLR to generate `.java` source files from grammar files.
+  ```sh
+  java org.antlr.v4.Tool *.g4
+  ```
+
+  After that, compile generated sources.
+  ```sh
+  javac *.java
+  ```
+
+  Then, use the `grun` tool comes with antlr to test your grammars.
+  ```sh
+  # Testing lexer
+  java org.antlr.v4.gui.TestRig C1Lexer tokens -tokens ../../test/test_cases/simple.c1
+  # Testing lexer + parser, GUI version parse tree
+  java org.antlr.v4.gui.TestRig C1 compilationUnit -gui ../../test/test_cases/simple.c1
+  # Testing lexer + parser, console printed version parse tree
+  java org.antlr.v4.gui.TestRig C1 compilationUnit -tree ../../test/test_cases/simple.c1
+  ```

大三上/编译原理和技术(H)/ppt/2017fall-judgement-testcases/c1recognizer/c1recognizer-config.cmake.in

@@ -0,0 +1,4 @@
+
+set(C1RECOGNIZER_INCLUDE_DIRS "@CMAKE_INSTALL_PREFIX@/include")
+set(C1RECOGNIZER_LIBRARY_DIRS "@CMAKE_INSTALL_PREFIX@/lib")
+set(C1RECOGNIZER_LIBS c1recognizer antlr4-runtime)