Translation phase 3: preprocessing tokens (WIP)

3 files changed, 61 insertions, 14 deletions

diff --git a/cpp.cpp b/cpp.cpp
index 54dd0d2..2f65a19 100644
--- a/cpp.cpp
+++ b/cpp.cpp
@@ -1,16 +1,13 @@
 #include "cpp.h"
 
+#include "bnf.h"
+#include "cppbnf.h"
+#include "grammer.h"
+#include "minicc.h"
+
 #include <gtest/gtest.h>
 #include <gmock/gmock.h>
 
-void CPP::PreprocessorTokensToTokens(std::vector<Token>& tokens)
-{
- for (auto& i : tokens) {
-  if (i.type == "preprocessing-op-or-punc")
-   i.type = i.value;
- }
-}
-
 // Phase 1: Map physical character set to basic source character set
 void CPP::source_charset_map()
 {
@@ -23,10 +20,37 @@ void CPP::backslash_escape()
  // TODO
 }
 
+namespace {
+
+std::vector<Token> sourceToCharTokens(const std::string& code)
+{
+ std::vector<Token> result;
+
+ Location location{1, 1};
+
+ for (char c: code) {
+  if (c == '\n') {
+   location.column = 1;
+   location.line++;
+  } else if (std::isprint(c)) {
+   location.column++;
+  }
+
+  result.emplace_back(Token{std::string(1, c), std::string(1, c), location});
+ }
+ return result;
+}
+
+}
+
 // Phase 3: Parse preprocessing tokens
-void CPP::preprocessing_tokenize()
+void CPP::preprocessing_tokenize(const std::string& s)
 {
- // TODO
+ auto charTokens {sourceToCharTokens(s)};
+
+ auto bnf{SubBNF(GetCppBNFLex(), "preprocessing-token")};
+ Gram::Compiler compiler(bnf, "preprocessing-token");
+ auto Tree = compiler.compile(charTokens);
 }
 
 // Phase 4: Preprocessing
@@ -50,24 +74,35 @@ void CPP::concatenate_strings()
 // Phase 7: Create tokens from preprocessing tokens
 void CPP::tokens_from_pptokens()
 {
+ // TODO
+}
+
+void CPP::PreprocessorTokensToTokens(std::vector<Token>& tokens)
+{
+ for (auto& i : tokens) {
+  if (i.type == "preprocessing-op-or-punc")
+   i.type = i.value;
+ }
 }
 
 // Phase 8: Instantiate objects
 void CPP::instantiate()
 {
+ // TODO
 }
 
 // Phase 9: Link libraries
 void CPP::link()
 {
+ // TODO
 }
 
 // phases of translation, according to standard
-void CPP::translate()
+void CPP::translate(const std::string& code)
 {
  source_charset_map();
  backslash_escape();
- preprocessing_tokenize();
+ preprocessing_tokenize(code);
  preprocess();
  execution_charset_map();
  concatenate_strings();
@@ -76,6 +111,12 @@ void CPP::translate()
  link();
 }
 
+TEST(Cpp, preprocessing_tokenize) {
+ CPP::preprocessing_tokenize("int main() { return 1; }");
+}
+
+#if 0
 TEST(Cpp, translate) {
  CPP::translate();
 }
+#endif
diff --git a/cpp.h b/cpp.h
index 7bb62c2..7388e94 100644
--- a/cpp.h
+++ b/cpp.h
@@ -11,7 +11,7 @@ void PreprocessorTokensToTokens(std::vector<Token>& tokens);
 // phases of translation, according to standard
 void source_charset_map(); // phase 1
 void backslash_escape(); // phase 2
-void preprocessing_tokenize(); // phase 3
+void preprocessing_tokenize(const std::string& s); // phase 3
 void preprocess(); // phase 4
 void execution_charset_map(); // phase 5
 void concatenate_strings(); // phase 6
@@ -20,6 +20,6 @@ void instantiate(); // phase 8
 void link(); // phase 9
 
 // all phases of translation
-void translate();
+void translate(const std::string& code);
 
 }
diff --git a/cppbnf.h b/cppbnf.h
new file mode 100644
index 0000000..384e6cf
--- /dev/null
+++ b/cppbnf.h
@@ -0,0 +1,6 @@
+#pragma once
+
+#include "bnf.h"
+
+BNF GetCppBNFLex();
+BNF GetCppBNFGram();