You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by GitBox <gi...@apache.org> on 2021/06/07 16:18:26 UTC
[GitHub] [arrow] pitrou commented on a change in pull request #10448: ARROW-12949: [C++] Add starts_with and ends_with
pitrou commented on a change in pull request #10448:
URL: https://github.com/apache/arrow/pull/10448#discussion_r646740864
##########
File path: cpp/src/arrow/compute/kernels/scalar_string.cc
##########
@@ -573,13 +609,69 @@ struct MatchSubstring<Type, PlainSubstringMatcher> {
}
};
+template <typename Type>
+struct MatchSubstring<Type, PlainStartsWithMatcher> {
+ static Status Exec(KernelContext* ctx, const ExecBatch& batch, Datum* out) {
+ auto options = MatchSubstringState::Get(ctx);
+ if (options.ignore_case) {
+#ifdef ARROW_WITH_RE2
+ MatchSubstringOptions converted_options = options;
+ converted_options.pattern = "^" + RE2::QuoteMeta(options.pattern);
+ ARROW_ASSIGN_OR_RAISE(auto matcher, RegexSubstringMatcher::Make(converted_options));
+ return MatchSubstringImpl<Type, RegexSubstringMatcher>::Exec(ctx, batch, out,
+ matcher.get());
+#else
+ return Status::NotImplemented("ignore_case requires RE2");
+#endif
+ }
+ ARROW_ASSIGN_OR_RAISE(auto matcher, PlainStartsWithMatcher::Make(options));
+ return MatchSubstringImpl<Type, PlainStartsWithMatcher>::Exec(ctx, batch, out,
+ matcher.get());
+ }
+};
+
+template <typename Type>
+struct MatchSubstring<Type, PlainEndsWithMatcher> {
+ static Status Exec(KernelContext* ctx, const ExecBatch& batch, Datum* out) {
+ auto options = MatchSubstringState::Get(ctx);
+ if (options.ignore_case) {
+#ifdef ARROW_WITH_RE2
+ MatchSubstringOptions converted_options = options;
+ converted_options.pattern = RE2::QuoteMeta(options.pattern) + "$";
+ ARROW_ASSIGN_OR_RAISE(auto matcher, RegexSubstringMatcher::Make(converted_options));
+ return MatchSubstringImpl<Type, RegexSubstringMatcher>::Exec(ctx, batch, out,
+ matcher.get());
+#else
+ return Status::NotImplemented("ignore_case requires RE2");
+#endif
+ }
+ ARROW_ASSIGN_OR_RAISE(auto matcher, PlainEndsWithMatcher::Make(options));
+ return MatchSubstringImpl<Type, PlainEndsWithMatcher>::Exec(ctx, batch, out,
+ matcher.get());
+ }
+};
+
const FunctionDoc match_substring_doc(
"Match strings against literal pattern",
("For each string in `strings`, emit true iff it contains a given pattern.\n"
"Null inputs emit null. The pattern must be given in MatchSubstringOptions. "
"If ignore_case is set, only simple case folding is performed."),
{"strings"}, "MatchSubstringOptions");
+const FunctionDoc starts_with_doc(
+ "Check if strings start with a pattern",
Review comment:
"literal pattern" perhaps (to disambiguate with regex)?
##########
File path: cpp/src/arrow/compute/kernels/scalar_string.cc
##########
@@ -636,13 +728,15 @@ std::string MakeLikeRegex(const MatchSubstringOptions& options) {
}
// A LIKE pattern matching this regex can be translated into a substring search.
-static RE2 kLikePatternIsSubstringMatch("%+([^%_]*)%+");
+static RE2 kLikePatternIsSubstringMatch(R"(%+([^%_]*[^\\%_])?%+)");
+// A LIKE pattern matching this regex can be translated into a prefix search.
+static RE2 kLikePatternIsStartsWith(R"(([^%_]*[^\\%_])?%+)");
+// A LIKE pattern matching this regex can be translated into a suffix search.
+static RE2 kLikePatternIsEndsWith(R"(%+([^%_]*))");
Review comment:
Out of curiosity, did you try to benchmark with/without those optimizations? `re2` might be doing similar under the hood...
##########
File path: cpp/src/arrow/compute/kernels/scalar_string.cc
##########
@@ -484,6 +484,42 @@ struct PlainSubstringMatcher {
bool Match(util::string_view current) const { return Find(current) >= 0; }
};
+struct PlainStartsWithMatcher {
+ const MatchSubstringOptions& options_;
+
+ explicit PlainStartsWithMatcher(const MatchSubstringOptions& options)
+ : options_(options) {}
+
+ static Result<std::unique_ptr<PlainStartsWithMatcher>> Make(
+ const MatchSubstringOptions& options) {
+ // Should be handled by partial template specialization below
+ DCHECK(!options.ignore_case);
+ return ::arrow::internal::make_unique<PlainStartsWithMatcher>(options);
+ }
+
+ bool Match(util::string_view current) const {
+ return current.starts_with(options_.pattern);
+ }
+};
+
+struct PlainEndsWithMatcher {
+ const MatchSubstringOptions& options_;
+
+ explicit PlainEndsWithMatcher(const MatchSubstringOptions& options)
+ : options_(options) {}
+
+ static Result<std::unique_ptr<PlainEndsWithMatcher>> Make(
+ const MatchSubstringOptions& options) {
+ // Should be handled by partial template specialization below
+ DCHECK(!options.ignore_case);
+ return ::arrow::internal::make_unique<PlainEndsWithMatcher>(options);
+ }
+
+ bool Match(util::string_view current) const {
+ return current.ends_with(options_.pattern);
Review comment:
Same here.
##########
File path: cpp/src/arrow/compute/kernels/scalar_string.cc
##########
@@ -484,6 +484,42 @@ struct PlainSubstringMatcher {
bool Match(util::string_view current) const { return Find(current) >= 0; }
};
+struct PlainStartsWithMatcher {
+ const MatchSubstringOptions& options_;
+
+ explicit PlainStartsWithMatcher(const MatchSubstringOptions& options)
+ : options_(options) {}
+
+ static Result<std::unique_ptr<PlainStartsWithMatcher>> Make(
+ const MatchSubstringOptions& options) {
+ // Should be handled by partial template specialization below
+ DCHECK(!options.ignore_case);
+ return ::arrow::internal::make_unique<PlainStartsWithMatcher>(options);
+ }
+
+ bool Match(util::string_view current) const {
+ return current.starts_with(options_.pattern);
Review comment:
`starts_with` is C++20, so should probably use the expanded version to ease transition to C++17 :-)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org