Merge pull request litecoin-project#14 from laanwj/2015_11_escape_plan

Escape all control characters
Electronic-Gulden-Foundation · Dec 2, 2015 · 5e7985a · 5e7985a
2 parents 46098ee + 7482163
commit 5e7985a
Show file tree

Hide file tree

Showing 6 changed files with 58 additions and 35 deletions.
diff --git a/Makefile.am b/Makefile.am
@@ -82,6 +82,7 @@ TEST_FILES = \
 	$(TEST_DATA_DIR)/fail9.json \
 	$(TEST_DATA_DIR)/pass1.json \
 	$(TEST_DATA_DIR)/pass2.json \
-	$(TEST_DATA_DIR)/pass3.json
+	$(TEST_DATA_DIR)/pass3.json \
+	$(TEST_DATA_DIR)/round1.json
 
 EXTRA_DIST=$(TEST_FILES) $(GEN_SRCS)
diff --git a/gen/gen.cpp b/gen/gen.cpp
@@ -15,17 +15,25 @@
 using namespace std;
 
 static bool initEscapes;
-static const char *escapes[256];
+static std::string escapes[256];
 
 static void initJsonEscape()
 {
+    // Escape all lower control characters (some get overridden with smaller sequences below)
+    for (int ch=0x00; ch<0x20; ++ch) {
+        char tmpbuf[20];
+        snprintf(tmpbuf, sizeof(tmpbuf), "\\u%04x", ch);
+        escapes[ch] = std::string(tmpbuf);
+    }
+
     escapes[(int)'"'] = "\\\"";
     escapes[(int)'\\'] = "\\\\";
     escapes[(int)'\b'] = "\\b";
     escapes[(int)'\f'] = "\\f";
     escapes[(int)'\n'] = "\\n";
     escapes[(int)'\r'] = "\\r";
     escapes[(int)'\t'] = "\\t";
+    escapes[(int)'\x7f'] = "\\u007f"; // U+007F DELETE
 
     initEscapes = true;
 }
@@ -38,13 +46,13 @@ static void outputEscape()
 		"static const char *escapes[256] = {\n");
 
 	for (unsigned int i = 0; i < 256; i++) {
-		if (!escapes[i]) {
+		if (escapes[i].empty()) {
 			printf("\tNULL,\n");
 		} else {
 			printf("\t\"");
 
 			unsigned int si;
-			for (si = 0; si < strlen(escapes[i]); si++) {
+			for (si = 0; si < escapes[i].size(); si++) {
 				char ch = escapes[i][si];
 				switch (ch) {
 				case '"':

diff --git a/lib/univalue_escapes.h b/lib/univalue_escapes.h
@@ -2,38 +2,38 @@
 #ifndef BITCOIN_UNIVALUE_UNIVALUE_ESCAPES_H
 #define BITCOIN_UNIVALUE_UNIVALUE_ESCAPES_H
 static const char *escapes[256] = {
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
+	"\\u0000",
+	"\\u0001",
+	"\\u0002",
+	"\\u0003",
+	"\\u0004",
+	"\\u0005",
+	"\\u0006",
+	"\\u0007",
 	"\\b",
 	"\\t",
 	"\\n",
-	NULL,
+	"\\u000b",
 	"\\f",
 	"\\r",
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
-	NULL,
+	"\\u000e",
+	"\\u000f",
+	"\\u0010",
+	"\\u0011",
+	"\\u0012",
+	"\\u0013",
+	"\\u0014",
+	"\\u0015",
+	"\\u0016",
+	"\\u0017",
+	"\\u0018",
+	"\\u0019",
+	"\\u001a",
+	"\\u001b",
+	"\\u001c",
+	"\\u001d",
+	"\\u001e",
+	"\\u001f",
 	NULL,
 	NULL,
 	"\\\"",
@@ -129,7 +129,7 @@ static const char *escapes[256] = {
 	NULL,
 	NULL,
 	NULL,
-	NULL,
+	"\\u007f",
 	NULL,
 	NULL,
 	NULL,

diff --git a/lib/univalue_write.cpp b/lib/univalue_write.cpp
@@ -24,10 +24,10 @@ static string json_escape(const string& inS)
         if (escStr)
             outS += escStr;
 
-        else if (isprint(ch))
+        else if (ch < 0x80)
             outS += ch;
 
-        else {
+        else { // TODO handle UTF-8 properly
             char tmpesc[16];
             sprintf(tmpesc, "\\u%04x", ch);
             outS += tmpesc;

diff --git a/test/round1.json b/test/round1.json
@@ -0,0 +1 @@
+["\u0000\u0001\u0002\u0003\u0004\u0005\u0006\u0007\b\t\n\u000b\f\r\u000e\u000f\u0010\u0011\u0012\u0013\u0014\u0015\u0016\u0017\u0018\u0019\u001a\u001b\u001c\u001d\u001e\u001f !\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\u007f"]
diff --git a/test/unitester.cpp b/test/unitester.cpp
@@ -23,12 +23,19 @@ static bool test_failed = false;
 
 #define d_assert(expr) { if (!(expr)) { test_failed = true; fprintf(stderr, "%s failed\n", filename.c_str()); } }
 
+static std::string rtrim(std::string s)
+{
+    s.erase(s.find_last_not_of(" \n\r\t")+1);
+    return s;
+}
+
 static void runtest(string filename, const string& jdata)
 {
         string prefix = filename.substr(0, 4);
 
-        bool wantPass = (prefix == "pass");
+        bool wantPass = (prefix == "pass") || (prefix == "roun");
         bool wantFail = (prefix == "fail");
+        bool wantRoundTrip = (prefix == "roun");
         assert(wantPass || wantFail);
 
         UniValue val;
@@ -39,6 +46,11 @@ static void runtest(string filename, const string& jdata)
         } else {
             d_assert(testResult == false);
         }
+
+        if (wantRoundTrip) {
+            std::string odata = val.write(0, 0);
+            assert(odata == rtrim(jdata));
+        }
 }
 
 static void runtest_file(const char *filename_)
@@ -106,6 +118,7 @@ static const char *filenames[] = {
         "pass1.json",
         "pass2.json",
         "pass3.json",
+        "round1.json",              // round-trip test
 };
 
 int main (int argc, char *argv[])
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		["\u0000\u0001\u0002\u0003\u0004\u0005\u0006\u0007\b\t\n\u000b\f\r\u000e\u000f\u0010\u0011\u0012\u0013\u0014\u0015\u0016\u0017\u0018\u0019\u001a\u001b\u001c\u001d\u001e\u001f !\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{\|}~\u007f"]