From 90329e5453b5c0e93964c9ab6ada815f7977f441 Mon Sep 17 00:00:00 2001 From: Fangrui Song Date: Mon, 19 Mar 2018 00:39:22 -0700 Subject: [PATCH] Simplify ComputeGuessScore --- src/fuzzy_match.cc | 1 + src/project.cc | 36 +++++++----------------------------- 2 files changed, 8 insertions(+), 29 deletions(-) diff --git a/src/fuzzy_match.cc b/src/fuzzy_match.cc index c37f24cb..f91655ed 100644 --- a/src/fuzzy_match.cc +++ b/src/fuzzy_match.cc @@ -152,6 +152,7 @@ TEST_SUITE("fuzzy_match") { Ranks("ab", {"ab", "aoo_boo", "acb"}); Ranks("CC", {"CamelCase", "camelCase", "camelcase"}); Ranks("cC", {"camelCase", "CamelCase", "camelcase"}); + Ranks("c c", {"camel case", "camelCase", "CamelCase", "camelcase", "camel ace"}); Ranks("Da.Te", {"Data.Text", "Data.Text.Lazy", "Data.Aeson.Encoding.text"}); // prefix Ranks("is", {"isIEEE", "inSuf"}); diff --git a/src/project.cc b/src/project.cc index f8c21ccb..6411d1ae 100644 --- a/src/project.cc +++ b/src/project.cc @@ -496,37 +496,15 @@ std::vector LoadCompilationEntriesFromDirectory( // Computes a score based on how well |a| and |b| match. This is used for // argument guessing. int ComputeGuessScore(const std::string& a, const std::string& b) { - const int kMatchPrefixWeight = 100; - const int kMismatchDirectoryWeight = 100; - const int kMatchPostfixWeight = 1; - - int score = 0; - size_t i = 0; - - // Increase score based on matching prefix. - for (i = 0; i < a.size() && i < b.size(); ++i) { - if (a[i] != b[i]) - break; - score += kMatchPrefixWeight; - } + // Increase score based on common prefix and suffix. Prefixes are prioritized. + size_t i = std::mismatch(a.begin(), a.end(), b.begin()).first - a.begin(); + size_t j = std::mismatch(a.rbegin(), a.rend(), b.rbegin()).first - a.rbegin(); + int score = 10 * i + j; // Reduce score based on mismatched directory distance. - for (size_t j = i; j < a.size(); ++j) { - if (a[j] == '/') - score -= kMismatchDirectoryWeight; - } - for (size_t j = i; j < b.size(); ++j) { - if (b[j] == '/') - score -= kMismatchDirectoryWeight; - } - - // Increase score based on common ending. Don't increase as much as matching - // prefix or directory distance. - for (size_t offset = 1; offset <= a.size() && offset <= b.size(); ++offset) { - if (a[a.size() - offset] != b[b.size() - offset]) - break; - score += kMatchPostfixWeight; - } + if (i + j < std::min(a.size(), b.size())) + score -= 100 * (std::count(a.begin() + i, a.end() - j, '/') + + std::count(b.begin() + i, b.end() - j, '/')); return score; }