Full sentences mode!

4 files changed, 18 insertions, 5 deletions

@@ -43,7 +43,7 @@ int main(int argc, char** args)
   std::cout << "Generating..." << std::endl;
   for (;;)
   {
-    std::string doc = stats->randomSentence(rand() % 45 + 5);
+    std::string doc = stats->randomSentence(140);
     std::string hi = doc;
     hi.resize(140);
     
@@ -51,7 +51,7 @@ int main(int argc, char** args)
   std::cout << "Generating..." << std::endl;
   for (;;)
   {
-    std::string doc = stats->randomSentence(rand() % 35 + 15);
+    std::string doc = stats->randomSentence(140);
     std::string hi = doc;
     hi.resize(140);
 
@@ -457,14 +457,14 @@ void printKgram(kgram k)
 }
 
 // runs in O(n log t) time where n is the input number of sentences and t is the number of tokens in the input corpus
-std::string kgramstats::randomSentence(int n)
+std::string kgramstats::randomSentence(int max)
 {
   std::string result;
   kgram cur(1, wildcardQuery);
   int cuts = 0;
   std::stack<parentype> open_delimiters;
         
-  for (int i=0; i<n; i++)
+  for (;;)
   {
     if (cur.size() == maxK)
     {
@@ -611,6 +611,19 @@ std::string kgramstats::randomSentence(int n)
     {
       break;
     }
+    
+    // Went over the limit, so reset
+    if (result.length() > max)
+    {
+      result = "";
+      cur = kgram(1, wildcardQuery);
+      cuts = 0;
+      
+      while (!open_delimiters.empty())
+      {
+        open_delimiters.pop();
+      }
+    }
   }
   
   // Remove the trailing space
@@ -97,7 +97,7 @@ class kgramstats
 {
 public:
         kgramstats(std::string corpus, int maxK);
-        std::string randomSentence(int n);
+        std::string randomSentence(int max);
         
 private:
         struct token_data