From a6e6a369f40d8fb6a191fd7f74fc5efa8bfae2a0 Mon Sep 17 00:00:00 2001
From: Paul Baltescu <pauldb89@gmail.com>
Date: Wed, 27 Nov 2013 14:33:36 +0000
Subject: Unify sampling backoff strategy.

---
 extractor/suffix_array_sampler.h | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
 create mode 100644 extractor/suffix_array_sampler.h

(limited to 'extractor/suffix_array_sampler.h')

diff --git a/extractor/suffix_array_sampler.h b/extractor/suffix_array_sampler.h
new file mode 100644
index 00000000..bb3c2653
--- /dev/null
+++ b/extractor/suffix_array_sampler.h
@@ -0,0 +1,34 @@
+#ifndef _SUFFIX_ARRAY_SAMPLER_H_
+#define _SUFFIX_ARRAY_SAMPLER_H_
+
+#include "backoff_sampler.h"
+
+namespace extractor {
+
+class SuffixArray;
+
+class SuffixArrayRangeSampler : public BackoffSampler {
+ public:
+  SuffixArrayRangeSampler(shared_ptr<SuffixArray> suffix_array,
+                          int max_samples);
+
+  SuffixArrayRangeSampler();
+
+ private:
+  int GetNumSubpatterns(const PhraseLocation& location) const;
+
+  int GetRangeLow(const PhraseLocation& location) const;
+
+  int GetRangeHigh(const PhraseLocation& location) const;
+
+  int GetPosition(const PhraseLocation& location, int index) const;
+
+  void AppendMatching(vector<int>& samples, int index,
+                      const PhraseLocation& location) const;
+
+  shared_ptr<SuffixArray> source_suffix_array;
+};
+
+} // namespace extractor
+
+#endif
-- 
cgit v1.2.3