feat(preprocessor): add file-local conditional switches

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

1 files changed, 92 insertions, 0 deletions

diff --git a/lib/Lexer/Preprocessor.cc b/lib/Lexer/Preprocessor.cc
new file mode 100644
index 0000000..235dd4b
--- /dev/null
+++ b/lib/Lexer/Preprocessor.cc
@@ -0,0 +1,92 @@
+#include "cps/Preprocessor.h"
+#include "cps/Diagnostics.h"
+#include <string_view>
+#include <unordered_set>
+
+using namespace cps;
+
+namespace {
+bool nameStart(char C) {
+    return (C >= 'A' && C <= 'Z') || (C >= 'a' && C <= 'z') || C == '_';
+}
+
+bool nameContinue(char C) {
+    return nameStart(C) || (C >= '0' && C <= '9');
+}
+
+std::string_view trim(std::string_view Text) {
+    const auto First = Text.find_first_not_of(" \t\r\v\f");
+    if (First == std::string_view::npos) return {};
+    return Text.substr(First, Text.find_last_not_of(" \t\r\v\f") - First + 1);
+}
+}
+
+PreprocessedSource cps::preprocess(std::istream &Input) {
+    PreprocessedSource Result;
+    std::unordered_set<std::string> Switches;
+    struct Conditional { bool ParentActive; int Line; };
+    std::vector<Conditional> Conditions;
+    bool Active = true;
+    int LineNumber = 0;
+    auto Error = [&](const char *Message) {
+        sourceError(LineNumber, "%s", Message);
+        Result.HadError = true;
+    };
+    std::string Line;
+    while (std::getline(Input, Line)) {
+        ++LineNumber;
+        Result.LibraryLines.push_back(Switches.count("__LIB") != 0);
+        auto Code = trim(Line);
+        size_t WordEnd = 0;
+        while (WordEnd < Code.size() && nameContinue(Code[WordEnd])) ++WordEnd;
+        auto Word = Code.substr(0, WordEnd);
+        bool Define = Word == "PP_DEFINE", Undef = Word == "PP_UNDEF";
+        bool Ifdef = Word == "PP_IFDEF", Ifndef = Word == "PP_IFNDEF";
+        bool Endif = Word == "PP_ENDIF";
+        bool Directive = Define || Undef || Ifdef || Ifndef || Endif;
+        bool Mask = Directive || !Active;
+        if (Directive) {
+            auto Tail = Code.substr(WordEnd);
+            Tail = trim(Tail.substr(0, Tail.find("//")));
+            if (Endif) {
+                if (!Tail.empty()) Error("PP_ENDIF does not take arguments");
+                if (Conditions.empty()) Error("PP_ENDIF without matching PP_IFDEF or PP_IFNDEF");
+                else {
+                    Active = Conditions.back().ParentActive;
+                    Conditions.pop_back();
+                }
+            } else if (Ifdef || Ifndef || Active) {
+                size_t NameEnd = 0;
+                if (!Tail.empty() && nameStart(Tail[0])) {
+                    NameEnd = 1;
+                    while (NameEnd < Tail.size() && nameContinue(Tail[NameEnd])) ++NameEnd;
+                }
+                bool Valid = NameEnd != 0 && NameEnd == Tail.size();
+                if (!Valid) {
+                    sourceError(LineNumber, "%.*s requires exactly one switch name",
+                                static_cast<int>(Word.size()), Word.data());
+                    Result.HadError = true;
+                }
+                std::string Name(Tail.substr(0, NameEnd));
+                if (Ifdef || Ifndef) {
+                    Conditions.push_back({Active, LineNumber});
+                    bool Defined = Switches.count(Name) != 0;
+                    Active = Active && Valid && (Ifdef ? Defined : !Defined);
+                } else if (Valid) {
+                    if (Define) Switches.insert(Name);
+                    else Switches.erase(Name);
+                }
+            }
+        }
+        if (Mask)
+            for (char &C : Line)
+                if (C != '\r') C = ' ';
+        Result.Text += Line;
+        if (!Input.eof()) Result.Text += '\n';
+    }
+    for (const auto &Condition : Conditions) {
+        sourceError(Condition.Line, "Unterminated preprocessor conditional: expected PP_ENDIF");
+        Result.HadError = true;
+    }
+    return Result;
+}