Merge pull request #763 from drbo/master

Allow paths with international characters in depfiles

3 files changed, 38 insertions, 31 deletions

diff --git a/src/depfile_parser.cc b/src/depfile_parser.cc
index 7268f31..7cee892 100644
--- a/src/depfile_parser.cc
+++ b/src/depfile_parser.cc
@@ -47,7 +47,7 @@ bool DepfileParser::Parse(string* content, string* err) {
       const char* start = in;
       
     {
-      char yych;
+      unsigned char yych;
       static const unsigned char yybm[] = {
           0,   0,   0,   0,   0,   0,   0,   0, 
           0,   0,   0,   0,   0,   0,   0,   0, 
@@ -65,22 +65,22 @@ 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,   0, 128, 128,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
-          0,   0,   0,   0,   0,   0,   0,   0, 
+        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, 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, 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, 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, 128, 128, 128, 128, 
+        128, 128, 128, 128, 128, 128, 128, 128, 
+        128, 128, 128, 128, 128, 128, 128, 128, 
       };
 
       yych = *in;
@@ -106,24 +106,28 @@ bool DepfileParser::Parse(string* content, string* err) {
           }
         }
       } else {
-        if (yych <= '^') {
-          if (yych <= 'Z') {
+        if (yych <= '_') {
+          if (yych <= '[') {
             if (yych <= '?') goto yy9;
-            goto yy5;
+            if (yych <= 'Z') goto yy5;
+            goto yy9;
           } else {
-            if (yych != '\\') goto yy9;
+            if (yych <= '\\') goto yy2;
+            if (yych <= '^') goto yy9;
+            goto yy5;
           }
         } else {
-          if (yych <= '{') {
-            if (yych == '`') goto yy9;
-            goto yy5;
-          } else {
-            if (yych <= '|') goto yy9;
-            if (yych <= '~') goto yy5;
+          if (yych <= '|') {
+            if (yych <= '`') goto yy9;
+            if (yych <= '{') goto yy5;
             goto yy9;
+          } else {
+            if (yych == 0x7F) goto yy9;
+            goto yy5;
           }
         }
       }
+yy2:
       ++in;
       if ((yych = *in) <= '"') {
         if (yych <= '\f') {
diff --git a/src/depfile_parser.in.cc b/src/depfile_parser.in.cc
index deaee5b..98c1621 100644
--- a/src/depfile_parser.in.cc
+++ b/src/depfile_parser.in.cc
@@ -45,7 +45,7 @@ bool DepfileParser::Parse(string* content, string* err) {
       // start: beginning of the current parsed span.
       const char* start = in;
       /*!re2c
-      re2c:define:YYCTYPE = "char";
+      re2c:define:YYCTYPE = "unsigned char";
       re2c:define:YYCURSOR = in;
       re2c:define:YYLIMIT = end;
 
@@ -73,7 +73,7 @@ bool DepfileParser::Parse(string* content, string* err) {
         *out++ = yych;
         continue;
       }
-      [a-zA-Z0-9+,/_:.~()}{@=!-]+ {
+      [a-zA-Z0-9+,/_:.~()}{@=!\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 fe9424a..ee798f8 100644
--- a/src/depfile_parser_test.cc
+++ b/src/depfile_parser_test.cc
@@ -121,18 +121,21 @@ TEST_F(DepfileParserTest, SpecialChars) {
   EXPECT_TRUE(Parse(
 "C:/Program\\ Files\\ (x86)/Microsoft\\ crtdefs.h: \n"
 " en@quot.header~ t+t-x!=1 \n"
-" openldap/slapd.d/cn=config/cn=schema/cn={0}core.ldif",
+" openldap/slapd.d/cn=config/cn=schema/cn={0}core.ldif\n"
+" Fu\303\244ball",
       &err));
   ASSERT_EQ("", err);
   EXPECT_EQ("C:/Program Files (x86)/Microsoft crtdefs.h",
             parser_.out_.AsString());
-  ASSERT_EQ(3u, parser_.ins_.size());
+  ASSERT_EQ(4u, parser_.ins_.size());
   EXPECT_EQ("en@quot.header~",
             parser_.ins_[0].AsString());
   EXPECT_EQ("t+t-x!=1",
             parser_.ins_[1].AsString());
   EXPECT_EQ("openldap/slapd.d/cn=config/cn=schema/cn={0}core.ldif",
             parser_.ins_[2].AsString());
+  EXPECT_EQ("Fu\303\244ball",
+            parser_.ins_[3].AsString());
 }
 
 TEST_F(DepfileParserTest, UnifyMultipleOutputs) {