Merge pull request #1342 from Lekensteyn/depfile-fixes

Follow GCC/Clang behavior wrt depfiles

3 files changed, 157 insertions, 46 deletions

diff --git a/src/depfile_parser.cc b/src/depfile_parser.cc
index 405289f..6faeac6 100644
--- a/src/depfile_parser.cc
+++ b/src/depfile_parser.cc
@@ -30,9 +30,15 @@ DepfileParser::DepfileParser(DepfileParserOptions options)
 // How do you end a line with a backslash?  The netbsd Make docs suggest
 // reading the result of a shell command echoing a backslash!
 //
-// Rather than implement all of above, we do a simpler thing here:
-// Backslashes escape a set of characters (see "escapes" defined below),
-// otherwise they are passed through verbatim.
+// Rather than implement all of above, we follow what GCC/Clang produces:
+// Backslashes escape a space or hash sign.
+// When a space is preceded by 2N+1 backslashes, it is represents N backslashes
+// followed by space.
+// When a space is preceded by 2N backslashes, it represents 2N backslashes at
+// the end of a filename.
+// A hash sign is escaped by a single backslash. All other backslashes remain
+// unchanged.
+//
 // If anyone actually has depfiles that rely on the more complicated
 // behavior we can adjust this.
 bool DepfileParser::Parse(string* content, string* err) {
@@ -72,7 +78,7 @@ bool DepfileParser::Parse(string* content, string* err) {
         128, 128, 128, 128, 128, 128, 128, 128, 
         128, 128, 128, 128, 128, 128, 128, 128, 
         128, 128, 128, 128, 128, 128, 128, 128, 
-        128, 128, 128,   0,   0,   0,   0, 128, 
+        128, 128, 128, 128,   0, 128,   0, 128, 
           0, 128, 128, 128, 128, 128, 128, 128, 
         128, 128, 128, 128, 128, 128, 128, 128, 
         128, 128, 128, 128, 128, 128, 128, 128, 
@@ -111,7 +117,8 @@ bool DepfileParser::Parse(string* content, string* err) {
           if (yych <= '#') goto yy4;
           goto yy12;
         } else {
-          if (yych == '\\') goto yy13;
+          if (yych <= '?') goto yy4;
+          if (yych <= '\\') goto yy13;
           goto yy4;
         }
       }
@@ -143,6 +150,7 @@ yy9:
       if (yybm[0+yych] & 128) {
         goto yy9;
       }
+yy11:
       {
         // Got a span of plain text.
         int len = (int)(in - start);
@@ -158,24 +166,22 @@ yy12:
       goto yy5;
 yy13:
       yych = *(yymarker = ++in);
-      if (yych <= '"') {
-        if (yych <= '\f') {
+      if (yych <= 0x1F) {
+        if (yych <= '\n') {
           if (yych <= 0x00) goto yy5;
-          if (yych == '\n') goto yy18;
-          goto yy16;
+          if (yych <= '\t') goto yy16;
+          goto yy17;
         } else {
-          if (yych <= '\r') goto yy20;
-          if (yych == ' ') goto yy22;
+          if (yych == '\r') goto yy19;
           goto yy16;
         }
       } else {
-        if (yych <= 'Z') {
-          if (yych <= '#') goto yy22;
-          if (yych == '*') goto yy22;
-          goto yy16;
+        if (yych <= '#') {
+          if (yych <= ' ') goto yy21;
+          if (yych <= '"') goto yy16;
+          goto yy23;
         } else {
-          if (yych <= ']') goto yy22;
-          if (yych == '|') goto yy22;
+          if (yych == '\\') goto yy25;
           goto yy16;
         }
       }
@@ -188,30 +194,93 @@ yy14:
       }
 yy16:
       ++in;
-      {
-        // Let backslash before other characters through verbatim.
-        *out++ = '\\';
-        *out++ = yych;
-        continue;
-      }
-yy18:
+      goto yy11;
+yy17:
       ++in;
       {
         // A line continuation ends the current file name.
         break;
       }
-yy20:
+yy19:
       yych = *++in;
-      if (yych == '\n') goto yy18;
+      if (yych == '\n') goto yy17;
       in = yymarker;
       goto yy5;
-yy22:
+yy21:
       ++in;
       {
-        // De-escape backslashed character.
-        *out++ = yych;
+        // 2N+1 backslashes plus space -> N backslashes plus space.
+        int len = (int)(in - start);
+        int n = len / 2 - 1;
+        if (out < start)
+          memset(out, '\\', n);
+        out += n;
+        *out++ = ' ';
         continue;
       }
+yy23:
+      ++in;
+      {
+        // De-escape hash sign, but preserve other leading backslashes.
+        int len = (int)(in - start);
+        if (len > 2 && out < start)
+          memset(out, '\\', len - 2);
+        out += len - 2;
+        *out++ = '#';
+        continue;
+      }
+yy25:
+      yych = *++in;
+      if (yych <= 0x1F) {
+        if (yych <= '\n') {
+          if (yych <= 0x00) goto yy11;
+          if (yych <= '\t') goto yy16;
+          goto yy11;
+        } else {
+          if (yych == '\r') goto yy11;
+          goto yy16;
+        }
+      } else {
+        if (yych <= '#') {
+          if (yych <= ' ') goto yy26;
+          if (yych <= '"') goto yy16;
+          goto yy23;
+        } else {
+          if (yych == '\\') goto yy28;
+          goto yy16;
+        }
+      }
+yy26:
+      ++in;
+      {
+        // 2N backslashes plus space -> 2N backslashes, end of filename.
+        int len = (int)(in - start);
+        if (out < start)
+          memset(out, '\\', len - 1);
+        out += len - 1;
+        break;
+      }
+yy28:
+      yych = *++in;
+      if (yych <= 0x1F) {
+        if (yych <= '\n') {
+          if (yych <= 0x00) goto yy11;
+          if (yych <= '\t') goto yy16;
+          goto yy11;
+        } else {
+          if (yych == '\r') goto yy11;
+          goto yy16;
+        }
+      } else {
+        if (yych <= '#') {
+          if (yych <= ' ') goto yy21;
+          if (yych <= '"') goto yy16;
+          goto yy23;
+        } else {
+          if (yych == '\\') goto yy25;
+          goto yy16;
+        }
+      }
     }
 
     }
diff --git a/src/depfile_parser.in.cc b/src/depfile_parser.in.cc
index f8c94b3..735a0c3 100644
--- a/src/depfile_parser.in.cc
+++ b/src/depfile_parser.in.cc
@@ -29,9 +29,15 @@ DepfileParser::DepfileParser(DepfileParserOptions options)
 // How do you end a line with a backslash?  The netbsd Make docs suggest
 // reading the result of a shell command echoing a backslash!
 //
-// Rather than implement all of above, we do a simpler thing here:
-// Backslashes escape a set of characters (see "escapes" defined below),
-// otherwise they are passed through verbatim.
+// Rather than implement all of above, we follow what GCC/Clang produces:
+// Backslashes escape a space or hash sign.
+// When a space is preceded by 2N+1 backslashes, it is represents N backslashes
+// followed by space.
+// When a space is preceded by 2N backslashes, it represents 2N backslashes at
+// the end of a filename.
+// A hash sign is escaped by a single backslash. All other backslashes remain
+// unchanged.
+//
 // If anyone actually has depfiles that rely on the more complicated
 // behavior we can adjust this.
 bool DepfileParser::Parse(string* content, string* err) {
@@ -68,12 +74,33 @@ bool DepfileParser::Parse(string* content, string* err) {
       re2c:indent:string = "  ";
 
       nul = "\000";
-      escape = [ \\#*[|\]];
       newline = '\r'?'\n';
 
-      '\\' escape {
-        // De-escape backslashed character.
-        *out++ = yych;
+      '\\\\'* '\\ ' {
+        // 2N+1 backslashes plus space -> N backslashes plus space.
+        int len = (int)(in - start);
+        int n = len / 2 - 1;
+        if (out < start)
+          memset(out, '\\', n);
+        out += n;
+        *out++ = ' ';
+        continue;
+      }
+      '\\\\'+ ' ' {
+        // 2N backslashes plus space -> 2N backslashes, end of filename.
+        int len = (int)(in - start);
+        if (out < start)
+          memset(out, '\\', len - 1);
+        out += len - 1;
+        break;
+      }
+      '\\'+ '#' {
+        // De-escape hash sign, but preserve other leading backslashes.
+        int len = (int)(in - start);
+        if (len > 2 && out < start)
+          memset(out, '\\', len - 2);
+        out += len - 2;
+        *out++ = '#';
         continue;
       }
       '$$' {
@@ -81,13 +108,7 @@ bool DepfileParser::Parse(string* content, string* err) {
         *out++ = '$';
         continue;
       }
-      '\\' [^\000\r\n] {
-        // Let backslash before other characters through verbatim.
-        *out++ = '\\';
-        *out++ = yych;
-        continue;
-      }
-      [a-zA-Z0-9+,/_:.~()}{%@=!\x80-\xFF-]+ {
+      '\\'+ [^\000\r\n] | [a-zA-Z0-9+,/_:.~()}{%=@\x5B\x5D!\x80-\xFF-]+ {
         // Got a span of plain text.
         int len = (int)(in - start);
         // Need to shift it over if we're overwriting backslashes.
diff --git a/src/depfile_parser_test.cc b/src/depfile_parser_test.cc
index 52fe7cd..19224f3 100644
--- a/src/depfile_parser_test.cc
+++ b/src/depfile_parser_test.cc
@@ -101,15 +101,36 @@ TEST_F(DepfileParserTest, Spaces) {
             parser_.ins_[2].AsString());
 }
 
+TEST_F(DepfileParserTest, MultipleBackslashes) {
+  // Successive 2N+1 backslashes followed by space (' ') are replaced by N >= 0
+  // backslashes and the space. A single backslash before hash sign is removed.
+  // Other backslashes remain untouched (including 2N backslashes followed by
+  // space).
+  string err;
+  EXPECT_TRUE(Parse(
+"a\\ b\\#c.h: \\\\\\\\\\  \\\\\\\\ \\\\share\\info\\\\#1",
+      &err));
+  ASSERT_EQ("", err);
+  EXPECT_EQ("a b#c.h",
+            parser_.out_.AsString());
+  ASSERT_EQ(3u, parser_.ins_.size());
+  EXPECT_EQ("\\\\ ",
+            parser_.ins_[0].AsString());
+  EXPECT_EQ("\\\\\\\\",
+            parser_.ins_[1].AsString());
+  EXPECT_EQ("\\\\share\\info\\#1",
+            parser_.ins_[2].AsString());
+}
+
 TEST_F(DepfileParserTest, Escapes) {
   // Put backslashes before a variety of characters, see which ones make
   // it through.
   string err;
   EXPECT_TRUE(Parse(
-"\\!\\@\\#$$\\%\\^\\&\\\\:",
+"\\!\\@\\#$$\\%\\^\\&\\[\\]\\\\:",
       &err));
   ASSERT_EQ("", err);
-  EXPECT_EQ("\\!\\@#$\\%\\^\\&\\",
+  EXPECT_EQ("\\!\\@#$\\%\\^\\&\\[\\]\\\\",
             parser_.out_.AsString());
   ASSERT_EQ(0u, parser_.ins_.size());
 }
@@ -123,7 +144,7 @@ TEST_F(DepfileParserTest, SpecialChars) {
 " en@quot.header~ t+t-x!=1 \\\n"
 " openldap/slapd.d/cn=config/cn=schema/cn={0}core.ldif\\\n"
 " Fu\303\244ball\\\n"
-" a\\[1\\]b@2%c",
+" a[1]b@2%c",
       &err));
   ASSERT_EQ("", err);
   EXPECT_EQ("C:/Program Files (x86)/Microsoft crtdefs.h",