5787

6 files changed, 262 insertions, 18 deletions

diff --git a/archive/2.vm/build2 b/archive/2.vm/build2
index b07e8a63..c6d5ef76 100755
--- a/archive/2.vm/build2
+++ b/archive/2.vm/build2
@@ -115,8 +115,8 @@ older_than mu.cc $LAYERS ../../enumerate/enumerate ../../tangle/tangle && {
   ../../tangle/tangle $LAYERS  > mu.cc
 }
 
-older_than ../../cleave/cleave ../../cleave/cleave.cc && {
-  $CXX $CXXFLAGS ../../cleave/cleave.cc -o ../../cleave/cleave
+older_than cleave/cleave cleave/cleave.cc && {
+  $CXX $CXXFLAGS cleave/cleave.cc -o cleave/cleave
   rm -rf .build
 }
 
@@ -132,8 +132,8 @@ grep -h "^\s*void test_" mu.cc  |sed 's/^\s*void \(.*\)() {.*/"\1",/'  |update .
 mkdir -p .build/termbox
 update_cp termbox/termbox.h .build/termbox
 
-older_than mu_bin mu.cc *_list ../../cleave/cleave termbox/* && {
-  ../../cleave/cleave mu.cc .build
+older_than mu_bin mu.cc *_list cleave/cleave termbox/* && {
+  cleave/cleave mu.cc .build
   noisy_cd .build
     # create the list of global variable declarations from the corresponding definitions
     grep ';' global_definitions_list  |sed 's/[=(].*/;/'  |sed 's/^[^\/# ]/extern &/'  |sed 's/^extern extern /extern /'  |update global_declarations_list
diff --git a/archive/2.vm/build3 b/archive/2.vm/build3
index 42e8ffd9..7bacfb62 100755
--- a/archive/2.vm/build3
+++ b/archive/2.vm/build3
@@ -140,8 +140,8 @@ older_than mu.cc $LAYERS ../../enumerate/enumerate ../../tangle/tangle && {
   ../../tangle/tangle $LAYERS  > mu.cc
 }
 
-older_than ../../cleave/cleave ../../cleave/cleave.cc && {
-  $CXX $CXXFLAGS ../../cleave/cleave.cc -o ../../cleave/cleave
+older_than cleave/cleave cleave/cleave.cc && {
+  $CXX $CXXFLAGS cleave/cleave.cc -o cleave/cleave
   rm -rf .build
 }
 
@@ -157,8 +157,8 @@ grep -h "^\s*void test_" mu.cc  |sed 's/^\s*void \(.*\)() {.*/"\1",/'  |update .
 mkdir -p .build/termbox
 update_cp termbox/termbox.h .build/termbox
 
-older_than mu_bin mu.cc *_list ../../cleave/cleave termbox/* && {
-  ../../cleave/cleave mu.cc .build
+older_than mu_bin mu.cc *_list cleave/cleave termbox/* && {
+  cleave/cleave mu.cc .build
   noisy_cd .build
     # create the list of global variable declarations from the corresponding definitions
     grep ';' global_definitions_list  |sed 's/[=(].*/;/'  |sed 's/^[^\/# ]/extern &/'  |sed 's/^extern extern /extern /'  |update global_declarations_list
diff --git a/archive/2.vm/build4 b/archive/2.vm/build4
index a1483089..08752831 100755
--- a/archive/2.vm/build4
+++ b/archive/2.vm/build4
@@ -187,19 +187,19 @@ mv_if_exists mu.cc $TMP
 ) &
 
 TMP=`mktemp`
-mv_if_exists ../../cleave/cleave $TMP
+mv_if_exists cleave/cleave $TMP
 (
-  wait_for_all ../../cleave/cleave.cc
-  older_than $TMP ../../cleave/cleave.cc && {
+  wait_for_all cleave/cleave.cc
+  older_than $TMP cleave/cleave.cc && {
     echo "building cleave"
-    $CXX $CXXFLAGS ../../cleave/cleave.cc -o $TMP  ||  quit
+    $CXX $CXXFLAGS cleave/cleave.cc -o $TMP  ||  quit
     rm -rf .build
     echo "done building cleave"
   }
-  mv $TMP ../../cleave/cleave
+  mv $TMP cleave/cleave
 ) &
 
-wait_for_all mu.cc ../../cleave/cleave  # cleave/cleave just for the .build cleanup
+wait_for_all mu.cc cleave/cleave  # cleave/cleave just for the .build cleanup
 mkdir -p .build
 # auto-generate function declarations, so I can define them in any order
 # functions start out unindented, have all args on the same line, and end in ') {'
@@ -215,10 +215,10 @@ update_cp termbox/termbox.h .build/termbox
 TMP=`mktemp`
 mv_if_exists mu_bin $TMP
 (
-  wait_for_all mu.cc ../../cleave/cleave termbox/*.c termbox/*.h termbox/*.inl
-  older_than $TMP mu.cc *_list ../../cleave/cleave termbox/* && {
+  wait_for_all mu.cc cleave/cleave termbox/*.c termbox/*.h termbox/*.inl
+  older_than $TMP mu.cc *_list cleave/cleave termbox/* && {
     echo "building mu_bin"
-    ../../cleave/cleave mu.cc .build  ||  quit
+    cleave/cleave mu.cc .build  ||  quit
     cd .build
       # create the list of global variable declarations from the corresponding definitions
       grep ';' global_definitions_list  |sed 's/[=(].*/;/'  |sed 's/^[^\/# ]/extern &/'  |sed 's/^extern extern /extern /'  |update global_declarations_list
diff --git a/archive/2.vm/clean b/archive/2.vm/clean
index 24a0300f..e3678430 100755
--- a/archive/2.vm/clean
+++ b/archive/2.vm/clean
@@ -6,4 +6,4 @@ rm -rf mu.cc core.mu mu_bin* *_list .build
 rm -rf termbox/*.o termbox/libtermbox.a
 rm -rf .until .quit
 test $# -gt 0 && exit 0  # convenience: 'clean top-level' to leave subsidiary tools alone
-rm -rf ../../enumerate/enumerate ../../tangle/tangle ../../tangle/*_list ../../cleave/cleave ../../*/*.dSYM
+rm -rf ../../enumerate/enumerate ../../tangle/tangle ../../tangle/*_list cleave/cleave cleave/cleave.dSYM ../../*/*.dSYM
diff --git a/archive/2.vm/cleave/Readme b/archive/2.vm/cleave/Readme
new file mode 100644
index 00000000..038b6c98
--- /dev/null
+++ b/archive/2.vm/cleave/Readme
@@ -0,0 +1 @@
+Tool to construct compilation units out of Mu's layers.
diff --git a/archive/2.vm/cleave/cleave.cc b/archive/2.vm/cleave/cleave.cc
new file mode 100644
index 00000000..a3637e81
--- /dev/null
+++ b/archive/2.vm/cleave/cleave.cc
@@ -0,0 +1,243 @@
+// Read a single-file C++ program having a very specific structure, and split
+// it up into multiple separate compilation units to reduce the work needed to
+// rebuild after a small change. Write each compilation unit only if it has
+// changed from what was on disk before.
+//
+// This tool is tightly coupled with the build system for this project. The
+// makefile already auto-generates various things; we only do here what
+// standard unix tools can't easily do.
+//
+// Usage:
+//  cleave [input C++ file] [existing output directory]
+//
+// The input C++ file has the following structure:
+//   [#includes]
+//   [type definitions]
+//   // Globals
+//   [global variable definitions]
+//   // End Globals
+//   [function definitions]
+//
+// Afterwards, the output directory contains:
+//   header -- everything before the '// Globals' delimiter
+//   global_definitions_list -- everything between '// Globals' and '// End Globals' delimiters
+//   [.cc files partitioning function definitions]
+//
+// Each output function definition file contains:
+//   #include "header"
+//   #include "global_declarations_list"
+//   [function definitions]
+//
+// To preserve the original layer-based line numbers in error messages and the
+// debugger, we'll chunk the files only at boundaries where we encounter a
+// '#line ' directive (generated by the previous tangle/ stage) between
+// functions.
+//
+// One exception: the first file emitted #includes "global_definitions_list" instead
+// of "global_declarations_list"
+
+// Tune this parameter to balance time for initial vs incremental build.
+//
+//   decrease value -> faster initial build
+//   increase value -> faster incremental build
+int Num_compilation_units = 3;
+
+#include<assert.h>
+#include<cstdlib>
+#include<cstring>
+
+#include<vector>
+using std::vector;
+#include<list>
+using std::list;
+#include<utility>
+using std::pair;
+
+#include<string>
+using std::string;
+
+#include<iostream>
+using std::istream;
+using std::ostream;
+using std::cin;
+using std::cout;
+using std::cerr;
+
+#include<sstream>
+using std::istringstream;
+using std::ostringstream;
+
+#include<fstream>
+using std::ifstream;
+using std::ofstream;
+
+#include <locale>
+using std::isspace;  // unicode-aware
+
+string trim(const string& s) {
+  string::const_iterator first = s.begin();
+  while (first != s.end() && isspace(*first))
+    ++first;
+  if (first == s.end()) return "";
+
+  string::const_iterator last = --s.end();
+  while (last != s.begin() && isspace(*last))
+    --last;
+  ++last;
+  return string(first, last);
+}
+
+bool starts_with(const string& s, const string& pat) {
+  string::const_iterator a=s.begin(), b=pat.begin();
+  for (/*nada*/; a!=s.end() && b!=pat.end(); ++a, ++b)
+    if (*a != *b) return false;
+  return b == pat.end();
+}
+
+bool has_data(istream& in) {
+  return in && !in.eof();
+}
+
+void slurp(const string& filename, vector<string>& lines) {
+  lines.clear();
+  ifstream in(filename.c_str());
+  while (has_data(in)) {
+    string curr_line;
+    getline(in, curr_line);
+    lines.push_back(curr_line);
+  }
+}
+
+size_t slurp_some_functions(const vector<string>& in, size_t start, vector<string>& out, bool first) {
+  out.clear();
+  if (start >= in.size()) return start;
+  out.push_back("#include \"header\"");
+  if (first)
+    out.push_back("#include \"global_definitions_list\"");
+  else
+    out.push_back("#include \"global_declarations_list\"");
+  out.push_back("");
+  size_t curr = start;
+  while (true) {
+    if (curr >= in.size()) break;
+    if (out.size() >= in.size()/Num_compilation_units) break;
+    while (curr < in.size()) {
+      // read functions -- lines until unindented '}'
+      while (curr < in.size()) {
+        const string& line = in.at(curr);
+//?         cerr << curr << ": adding to function: " << line << '\n';
+        out.push_back(line);  ++curr;
+        if (!line.empty() && line.at(0) == '}') break;
+      }
+      // now look for a '#line' directive before the next non-comment non-empty
+      // line
+      while (curr < in.size()) {
+        const string& line = in.at(curr);
+        if (starts_with(line, "#line ")) goto try_return;
+        out.push_back(line);   ++curr;
+        if (trim(line).empty()) continue;
+        if (starts_with(trim(line), "//")) continue;
+        break;
+      }
+    }
+    try_return:;
+  }
+  return curr;
+}
+
+// compare contents of a file with a list of lines, ignoring #line directives
+// on both sides
+bool no_change(const vector<string>& lines, const string& output_filename) {
+  vector<string> old_lines;
+  slurp(output_filename, old_lines);
+  size_t l=0, o=0;
+  while (true) {
+    while (l < lines.size() &&
+            (lines.at(l).empty() || starts_with(lines.at(l), "#line "))) {
+      ++l;
+    }
+    while (o < old_lines.size() &&
+            (old_lines.at(o).empty() || starts_with(old_lines.at(o), "#line "))) {
+      ++o;
+    }
+    if (l >= lines.size() && o >= old_lines.size()) return true;  // no change
+    if (l >= lines.size() || o >= old_lines.size()) return false;  // contents changed
+//?     cerr << "comparing\n";
+//?     cerr << o << ": " << old_lines.at(o) << '\n';
+//?     cerr << l << ": " << lines.at(l) << '\n';
+    if (lines.at(l) != old_lines.at(o)) return false;  // contents changed
+    ++l;  ++o;
+  }
+  assert(false);
+}
+
+string next_output_filename(const string& output_directory) {
+  static int file_count = 0;
+  ostringstream out;
+  out << output_directory << "/mu_" << file_count << ".cc";
+  file_count++;
+  return out.str();
+}
+
+void emit_file(const vector<string>& lines, const string& output_filename) {
+  if (no_change(lines, output_filename)) return;
+  cerr << "  updating " << output_filename << '\n';
+  ofstream out(output_filename.c_str());
+  for (size_t i = 0; i < lines.size(); ++i)
+    out << lines.at(i) << '\n';
+}
+
+void emit_compilation_unit(const vector<string>& lines, const string& output_directory) {
+  string output_filename = next_output_filename(output_directory);
+  emit_file(lines, output_filename);
+}
+
+int main(int argc, const char* argv[]) {
+  if (argc != 3) {
+    cerr << "usage: cleave [input .cc file] [output directory]\n";
+    exit(0);
+  }
+
+  // read input
+  vector<string> lines;
+  slurp(argv[1], lines);
+
+  // write header until but excluding '// Global' delimiter
+  string output_directory = argv[2];
+  size_t line_num = 0;
+  {
+    vector<string> out;
+    while (line_num < lines.size()) {
+      const string& line = lines.at(line_num);
+      if (trim(line) == "// Globals") break;  // todo: #line directive for delimiters
+      out.push_back(line);
+      ++line_num;
+    }
+    emit_file(out, output_directory+"/header");
+  }
+
+  // write global_definitions_list (including delimiters)
+  {
+    vector<string> out;
+    while (line_num < lines.size()) {
+      const string& line = lines.at(line_num);
+      out.push_back(line);
+      ++line_num;
+      if (trim(line) == "// End Globals") break;
+    }
+    emit_file(out, output_directory+"/global_definitions_list");
+  }
+
+  // segment functions
+  // first one is special
+  if (line_num < lines.size()) {
+    vector<string> function;
+    line_num = slurp_some_functions(lines, line_num, function, true);
+    emit_compilation_unit(function, output_directory);
+  }
+  while (line_num < lines.size()) {
+    vector<string> function;
+    line_num = slurp_some_functions(lines, line_num, function, false);
+    emit_compilation_unit(function, output_directory);
+  }
+}