fix tokenize

base/str.cc:
    Fix tokenize so that it doesn't behave incorrectly when there
    are empty strings.
test/tokentest.cc:
    Clean up the test function so it's easier to see what's going on

--HG--
extra : convert_revision : c7a3db7bc516d3575b1cc4ab7afbd0f1fbe1ec6f

diff --git a/base/str.cc b/base/str.cc
index 3a11bb17deb8e684d54535fccf991254a5aa86b4..15f44dad27a6260d112586b90780b81f08278bf8 100644 (file)
--- a/base/str.cc
+++ b/base/str.cc
@@ -26,11 +26,10 @@
  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-#include <iostream>
-
-#include <string.h>
 #include <ctype.h>
 
+#include <cstring>
+#include <iostream>
 #include <string>
 #include <vector>
 
@@ -75,15 +74,17 @@ tokenize(vector<string>& v, const string &s, char token, bool ignore)
     string::size_type first = 0;
     string::size_type last = s.find_first_of(token);
 
-    if (ignore) {
-        if (last == first) {
-            while (last == first)
-                last = s.find_first_of(token, ++first);
+    if (s.empty())
+        return;
 
-            if (last == string::npos) {
-                v.push_back(s);
-                return;
-            }
+    if (ignore && last == first) {
+        while (last == first)
+            last = s.find_first_of(token, ++first);
+
+        if (last == string::npos) {
+            if (first != s.size())
+                v.push_back(s.substr(first));
+            return;
         }
     }
 

diff --git a/test/tokentest.cc b/test/tokentest.cc
index cd2182141606acd5ee488919470a0b1c950325d7..7f27d58fe8431cfa3ed044d284e4997dfc127776 100644 (file)
--- a/test/tokentest.cc
+++ b/test/tokentest.cc
@@ -26,7 +26,7 @@
  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-#include <iostream.h>
+#include <iostream>
 #include <string>
 #include <vector>
 
@@ -35,35 +35,47 @@
 int
 main(int argc, char *argv[])
 {
-  if (argc != 3) {
-    cout << "Usage: " << argv[0] << " <string> <token>\n";
-    exit(1);
-  }
+    using namespace std;
 
-  int i;
-  string test = argv[1];
-  vector<string> tokens1;
-  vector<string> tokens2;
-  char token = argv[2][0];
+    if (argc != 3) {
+        cout << "Usage: " << argv[0] << " <string> <token>\n";
+        exit(1);
+    }
 
-  cout << "string = \"" << test << "\", token = \'" << token << "\'\n";
-  cout << "testing without ignore\n";
-  tokenize(tokens1, test, token, false);
+    int i;
+    string test = argv[1];
+    vector<string> tokens1;
+    vector<string> tokens2;
+    char token = argv[2][0];
 
-  if (tokens1.size()) {
-      for (i = 0; i < tokens1.size() - 1; i++)
-          cout << tokens1[i] << "(" << tokens1[i].size() << "), ";
-      cout << tokens1[i]  << "(" << tokens1[i].size() << ")\n";
-  }
+    cout << "string = \"" << test << "\", token = \'" << token << "\'\n";
+    cout << "testing without ignore\n";
+    tokenize(tokens1, test, token, false);
 
-  cout << "testing with ignore\n";
-  tokenize(tokens2, test, token, true);
+    if (tokens1.size()) {
+        int size = tokens1.size();
+        cout << "size = " << size << "\n";
+        for (i = 0; i < size; i++) {
+            cout << "'" << tokens1[i] << "' (" << tokens1[i].size()
+                 << ")" << ((i == size - 1) ? "\n" : ", ");
+        }
+    } else {
+        cout << "no tokens" << endl;
+    }
 
-  if (tokens2.size()) {
-      for (i = 0; i < tokens2.size() - 1; i++)
-          cout << tokens2[i] << "(" << tokens2[i].size() << "), ";
-      cout << tokens2[i] << "(" << tokens2[i].size() << ")\n";
-  }
+    cout << "testing with ignore\n";
+    tokenize(tokens2, test, token, true);
 
-  return 0;
+    if (tokens2.size()) {
+        int size = tokens2.size();
+        cout << "size = " << size << "\n";
+        for (i = 0; i < size; i++) {
+            cout << "'" << tokens2[i] << "' (" << tokens2[i].size()
+                 << ")" << ((i == size - 1) ? "\n" : ", ");
+        }
+    } else {
+        cout << "no tokens" << endl;
+    }
+
+    return 0;
 }