From f253db0133a8ba680acad0ad7e1f8e4f64e10059 Mon Sep 17 00:00:00 2001 From: yanyiwu Date: Tue, 3 May 2016 21:24:40 +0800 Subject: [PATCH] use map/set instead of unordered_map/unordered_set to make result stable --- include/cppjieba/TextRankExtractor.hpp | 11 +++++++---- test/unittest/textrank_test.cpp | 14 +++++++++----- 2 files changed, 16 insertions(+), 9 deletions(-) diff --git a/include/cppjieba/TextRankExtractor.hpp b/include/cppjieba/TextRankExtractor.hpp index 948f3ad..a625695 100644 --- a/include/cppjieba/TextRankExtractor.hpp +++ b/include/cppjieba/TextRankExtractor.hpp @@ -12,15 +12,18 @@ namespace cppjieba { public: typedef struct _Word {string word;vector offsets;double weight;} Word; // struct Word private: - typedef std::unordered_map WordMap; + typedef std::map WordMap; class WordGraph{ private: typedef double Score; typedef string Node; - typedef std::unordered_set NodeSet; - typedef std::unordered_map Edges; - typedef std::unordered_map Graph; + typedef std::set NodeSet; + + typedef std::map Edges; + typedef std::map Graph; + //typedef std::unordered_map Edges; + //typedef std::unordered_map Graph; double d; Graph graph; diff --git a/test/unittest/textrank_test.cpp b/test/unittest/textrank_test.cpp index c4ae193..70dbc52 100644 --- a/test/unittest/textrank_test.cpp +++ b/test/unittest/textrank_test.cpp @@ -24,14 +24,16 @@ TEST(TextRankExtractorTest, Test1) { vector > words; Extractor.Extract(s, words, topN); res << words; - ASSERT_EQ(res, "[\"世界:1\", \"你好:0.514286\"]"); + //ASSERT_EQ(res, "[\"世界:1\", \"你好:0.514286\"]"); + ASSERT_EQ(res, "[\"\xE4\xB8\x96\xE7\x95\x8C:1\", \"\xE4\xBD\xA0\xE5\xA5\xBD:0.519787\"]"); } { vector words; Extractor.Extract(s, words, topN); res << words; - ASSERT_EQ(res, "[\"世界|[\"6\", \"12\"]|1\", \"你好|[\"0\"]|0.514286\"]"); + //ASSERT_EQ(res, "[\"世界|[\"6\", \"12\"]|1\", \"你好|[\"0\"]|0.514286\"]"); + ASSERT_EQ(res, "[\"\xE4\xB8\x96\xE7\x95\x8C|[\"6\", \"12\"]|1\", \"\xE4\xBD\xA0\xE5\xA5\xBD|[\"0\"]|0.519787\"]"); } } @@ -42,7 +44,7 @@ TEST(TextRankExtractorTest, Test1) { size_t topN = 5; Extractor.Extract(s, wordweights, topN); res << wordweights; - ASSERT_EQ(res, "[\"专业|[\"36\"]|1\", \"CEO|[\"94\"]|0.94764\", \"当上|[\"87\"]|0.79271\", \"手扶拖拉机|[\"21\"]|0.789347\", \"走上|[\"100\"]|0.768261\"]"); + ASSERT_EQ(res, "[\"\xE4\xB8\x93\xE4\xB8\x9A|[\"36\"]|1\", \"CEO|[\"94\"]|0.95375\", \"\xE6\x89\x8B\xE6\x89\xB6\xE6\x8B\x96\xE6\x8B\x89\xE6\x9C\xBA|[\"21\"]|0.801701\", \"\xE5\xBD\x93\xE4\xB8\x8A|[\"87\"]|0.798968\", \"\xE8\xB5\xB0\xE4\xB8\x8A|[\"100\"]|0.775505\"]"); // ASSERT_EQ(res, "[\"\xE4\xB8\x93\xE4\xB8\x9A|[\"36\"]|1\", \"CEO|[\"94\"]|0.953149\", \"\xE6\x89\x8B\xE6\x89\xB6\xE6\x8B\x96\xE6\x8B\x89\xE6\x9C\xBA|[\"21\"]|0.794203\", \"\xE5\xBD\x93\xE4\xB8\x8A|[\"87\"]|0.78716\", \"\xE8\xB5\xB0\xE4\xB8\x8A|[\"100\"]|0.767636\"]"); } @@ -53,7 +55,8 @@ TEST(TextRankExtractorTest, Test1) { size_t topN = 5; Extractor.Extract(s, wordweights, topN); res << wordweights; - ASSERT_EQ(res, "[\"iPhone6|[\"6\"]|1\", \"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|0.996126\"]"); + ASSERT_EQ(res, "[\"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|1\", \"iPhone6|[\"6\"]|0.996126\"]"); + //ASSERT_EQ(res, "[\"iPhone6|[\"6\"]|1\", \"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|0.996126\"]"); } } @@ -82,6 +85,7 @@ TEST(TextRankExtractorTest, Test2) { size_t topN = 5; Extractor.Extract(s, wordweights, topN); res << wordweights; - ASSERT_EQ(res, "[\"iPhone6|[\"6\"]|1\", \"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|0.996126\"]"); + //ASSERT_EQ(res, "[\"iPhone6|[\"6\"]|1\", \"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|0.996126\"]"); + ASSERT_EQ(res, "[\"\xE4\xB8\x80\xE9\x83\xA8|[\"0\"]|1\", \"iPhone6|[\"6\"]|0.996126\"]"); } }