Merge pull request #35182 from vdimir/regex-replace-35117

Fix replaceRegexpAll
ClickHouse · Mar 11, 2022 · 38fa55f · 38fa55f
2 parents 585a9ed + e4e058d
commit 38fa55f
Show file tree

Hide file tree

Showing 3 changed files with 44 additions and 5 deletions.
diff --git a/src/Functions/ReplaceRegexpImpl.h b/src/Functions/ReplaceRegexpImpl.h
@@ -137,8 +137,14 @@ struct ReplaceRegexpImpl
 
                 if (replace_one)
                     can_finish_current_string = true;
-                else if (match.length() == 0)
-                    ++match_pos; /// Step one character to avoid infinite loop.
+
+                if (match.length() == 0)
+                {
+                    /// Step one character to avoid infinite loop
+                    ++match_pos;
+                    if (match_pos >= static_cast<size_t>(input.length()))
+                        can_finish_current_string = true;
+                }
             }
             else
                 can_finish_current_string = true;

diff --git a/tests/queries/0_stateless/02151_replace_regexp_all_empty_match_alternative.reference b/tests/queries/0_stateless/02151_replace_regexp_all_empty_match_alternative.reference
@@ -1,3 +1,18 @@
 1
 1
 1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
+1
diff --git a/tests/queries/0_stateless/02151_replace_regexp_all_empty_match_alternative.sql b/tests/queries/0_stateless/02151_replace_regexp_all_empty_match_alternative.sql
@@ -1,3 +1,21 @@
-select replaceRegexpAll(',,1,,', '^[,]*|[,]*$', '') x;
-select replaceRegexpAll(',,1', '^[,]*|[,]*$', '') x;
-select replaceRegexpAll('1,,', '^[,]*|[,]*$', '') x;
+SELECT replaceRegexpAll(',,1,,', '^[,]*|[,]*$', '');
+SELECT replaceRegexpAll(',,1', '^[,]*|[,]*$', '');
+SELECT replaceRegexpAll('1,,', '^[,]*|[,]*$', '');
+
+SELECT replaceRegexpAll(materialize(',,1,,'), '^[,]*|[,]*$', '');
+SELECT replaceRegexpAll(materialize(',,1'), '^[,]*|[,]*$', '');
+SELECT replaceRegexpAll(materialize('1,,'), '^[,]*|[,]*$', '');
+
+SELECT replaceRegexpAll('a', 'z*', '') == 'a';
+SELECT replaceRegexpAll('aa', 'z*', '') == 'aa';
+SELECT replaceRegexpAll('aaq', 'z*', '') == 'aaq';
+SELECT replaceRegexpAll('aazq', 'z*', '') == 'aaq';
+SELECT replaceRegexpAll('aazzq', 'z*', '') == 'aaq';
+SELECT replaceRegexpAll('aazzqa', 'z*', '') == 'aaqa';
+
+SELECT replaceRegexpAll(materialize('a'), 'z*', '') == 'a';
+SELECT replaceRegexpAll(materialize('aa'), 'z*', '') == 'aa';
+SELECT replaceRegexpAll(materialize('aaq'), 'z*', '') == 'aaq';
+SELECT replaceRegexpAll(materialize('aazq'), 'z*', '') == 'aaq';
+SELECT replaceRegexpAll(materialize('aazzq'), 'z*', '') == 'aaq';
+SELECT replaceRegexpAll(materialize('aazzqa'), 'z*', '') == 'aaqa';