From 811ddf2226bd10f6b6e7894be2d0a6dcf083aeaf Mon Sep 17 00:00:00 2001
From: Paul Masurel <paul.masurel@gmail.com>
Date: Wed, 8 Aug 2018 11:15:20 +0900
Subject: [PATCH] Closes #364 (#365)

* Closes #364

* Trying to raise the recursion limit

* Better unit test and bug fix on token offsets
---
 src/lib.rs                          |  1 +
 src/tokenizer/token_stream_chain.rs | 34 +++++++++++++++++++++++++++--
 2 files changed, 33 insertions(+), 2 deletions(-)

diff --git a/src/lib.rs b/src/lib.rs
index d3b1ff771..c01226c55 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -7,6 +7,7 @@
 #![allow(new_without_default)]
 #![allow(decimal_literal_representation)]
 #![warn(missing_docs)]
+#![recursion_limit="80"]
 
 //! # `tantivy`
 //!
diff --git a/src/tokenizer/token_stream_chain.rs b/src/tokenizer/token_stream_chain.rs
index 652ac078a..01d631e2b 100644
--- a/src/tokenizer/token_stream_chain.rs
+++ b/src/tokenizer/token_stream_chain.rs
@@ -1,5 +1,7 @@
 use tokenizer::{Token, TokenStream};
 
+const POSITION_GAP: usize = 2;
+
 pub(crate) struct TokenStreamChain<TTokenStream: TokenStream> {
     offsets: Vec<usize>,
     token_streams: Vec<TTokenStream>,
@@ -37,14 +39,14 @@ where
                 let token = token_stream.token();
                 let offset_offset = self.offsets[self.stream_idx];
                 self.token.offset_from = token.offset_from + offset_offset;
-                self.token.offset_from = token.offset_from + offset_offset;
+                self.token.offset_to = token.offset_to + offset_offset;
                 self.token.position = token.position + self.position_shift;
                 self.token.text.clear();
                 self.token.text.push_str(token.text.as_str());
                 return true;
             } else {
                 self.stream_idx += 1;
-                self.position_shift = self.token.position + 2;
+                self.position_shift = self.token.position.wrapping_add(POSITION_GAP);
             }
         }
         false
@@ -66,3 +68,31 @@ where
         &mut self.token
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::POSITION_GAP;
+    use super::TokenStreamChain;
+    use super::super::{Tokenizer, TokenStream, SimpleTokenizer};
+
+    #[test]
+    fn test_chain_first_emits_no_tokens() {
+        let token_streams = vec![SimpleTokenizer.token_stream(""), SimpleTokenizer.token_stream("hello world")];
+        let mut token_chain = TokenStreamChain::new(vec![0, 0], token_streams);
+
+        assert!(token_chain.advance());
+        assert_eq!(token_chain.token().text, "hello");
+        assert_eq!(token_chain.token().offset_from, 0);
+        assert_eq!(token_chain.token().offset_to, 5);
+        assert_eq!(token_chain.token().position, POSITION_GAP - 1);
+
+        assert!(token_chain.advance());
+        assert_eq!(token_chain.token().text, "world");
+        assert_eq!(token_chain.token().offset_from, 6);
+        assert_eq!(token_chain.token().offset_to, 11);
+        assert_eq!(token_chain.token().position, POSITION_GAP);
+        
+        assert!(!token_chain.advance());
+    }
+
+}
\ No newline at end of file