From 5b621393e2ced88c159094e49734ef981e063cd3 Mon Sep 17 00:00:00 2001
From: Paul Masurel <paul.masurel@gmail.com>
Date: Thu, 7 Apr 2016 22:35:59 +0900
Subject: [PATCH] take in account doc_freq for the docid buffer.

---
 src/core/merger.rs |  2 +-
 src/core/reader.rs | 13 +++++++------
 src/core/writer.rs |  3 +--
 src/lib.rs         |  2 ++
 4 files changed, 11 insertions(+), 9 deletions(-)

diff --git a/src/core/merger.rs b/src/core/merger.rs
index ebcb40497..96655f7fd 100644
--- a/src/core/merger.rs
+++ b/src/core/merger.rs
@@ -95,7 +95,7 @@ impl<'a> PostingsMerger<'a> {
         {
             let offset = self.doc_offsets[heap_item.segment_ord];
             let reader = &self.readers[heap_item.segment_ord];
-            for doc_id in reader.read_postings(heap_item.term_info.postings_offset) {
+            for doc_id in reader.read_postings(&heap_item.term_info) {
                 self.doc_ids.push(offset + doc_id);
             }
         }
diff --git a/src/core/reader.rs b/src/core/reader.rs
index 7d77e8c87..fa553d64b 100644
--- a/src/core/reader.rs
+++ b/src/core/reader.rs
@@ -42,10 +42,10 @@ impl SegmentPostings {
         }
     }
 
-    pub fn from_data(data: &[u8]) -> SegmentPostings {
+    pub fn from_data(doc_freq: DocId, data: &[u8]) -> SegmentPostings {
         let mut cursor = Cursor::new(data);
         let data: Vec<u32> = Vec::deserialize(&mut cursor).unwrap();
-        let mut doc_ids: Vec<u32> = (0u32..data.len() as u32 + 1_000 as u32).collect();
+        let mut doc_ids: Vec<u32> = (0u32..doc_freq).collect();
         let decoder = Decoder::new();
         let num_doc_ids = decoder.decode_sorted(&data, &mut doc_ids);
         doc_ids.truncate(num_doc_ids);
@@ -153,9 +153,10 @@ impl SegmentReader {
         self.fast_fields_reader.get_field(u32_field)
     }
 
-    pub fn read_postings(&self, offset: u32) -> SegmentPostings {
-        let postings_data = &self.postings_data.as_slice()[(offset as usize)..];
-        SegmentPostings::from_data(&postings_data)
+    pub fn read_postings(&self, term_info: &TermInfo) -> SegmentPostings {
+        let offset = term_info.postings_offset as usize;
+        let postings_data = &self.postings_data.as_slice()[offset..];
+        SegmentPostings::from_data(term_info.doc_freq, &postings_data)
     }
 
     fn get_term<'a>(&'a self, term: &Term) -> Option<TermInfo> {
@@ -170,7 +171,7 @@ impl SegmentReader {
         for term in terms.iter() {
             match self.get_term(term) {
                 Some(term_info) => {
-                    let segment_posting = self.read_postings(term_info.postings_offset);
+                    let segment_posting = self.read_postings(&term_info);
                     segment_postings.push(segment_posting);
                 }
                 None => {
diff --git a/src/core/writer.rs b/src/core/writer.rs
index b291162e0..446dabf65 100644
--- a/src/core/writer.rs
+++ b/src/core/writer.rs
@@ -76,7 +76,6 @@ impl IndexWriter {
 					segment_writer.finalize().unwrap();
 					index_clone.sync(&segment).unwrap();
 					index_clone.publish_segment(&segment).unwrap();
-					// segment_writer.commit().unwrap();
 				}
 			})
 		}).collect();
@@ -107,7 +106,7 @@ impl IndexWriter {
 	}
 
     pub fn add_document(&mut self, doc: Document) -> io::Result<()> {
-		let arc_doc = ArcDoc::new(doc);
+        let arc_doc = ArcDoc::new(doc);
 		try!(
 			self.queue_input.send(arc_doc)
 				.map_err(|e| io::Error::new(ErrorKind::Other, e))
diff --git a/src/lib.rs b/src/lib.rs
index 0b1af23c9..848b43e27 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -30,6 +30,8 @@ extern crate num_cpus;
 
 mod core;
 
+pub use core::directory::Directory;
+pub use core::searcher::Searcher;
 pub use core::index::Index;
 pub use core::schema;
 pub use core::schema::Term;