From 7ac529c23516efee522046daf2a650fd057a7348 Mon Sep 17 00:00:00 2001
From: Aliaksandr Valialkin <valyala@victoriametrics.com>
Date: Sat, 25 May 2024 22:59:13 +0200
Subject: [PATCH] lib/logstorage: work-in-progress

---
 docs/VictoriaLogs/LogsQL.md   | 29 ++++++++++++++++++++++++++++-
 lib/logstorage/pipe_unroll.go |  7 ++++++-
 lib/regexutil/regex_test.go   |  2 ++
 3 files changed, 36 insertions(+), 2 deletions(-)

diff --git a/docs/VictoriaLogs/LogsQL.md b/docs/VictoriaLogs/LogsQL.md
index 74cec09c1..ec36f58ab 100644
--- a/docs/VictoriaLogs/LogsQL.md
+++ b/docs/VictoriaLogs/LogsQL.md
@@ -254,6 +254,7 @@ The list of LogsQL filters:
 - [Word filter](#word-filter) - matches logs with the given [word](#word)
 - [Phrase filter](#phrase-filter) - matches logs with the given phrase
 - [Prefix filter](#prefix-filter) - matches logs with the given word prefix or phrase prefix
+- [Substring filter](#substring-filter) - matches logs with the given substring
 - [Empty value filter](#empty-value-filter) - matches logs without the given [log field](https://docs.victoriametrics.com/victorialogs/keyconcepts/#data-model)
 - [Any value filter](#any-value-filter) - matches logs with the given non-empty [log field](https://docs.victoriametrics.com/victorialogs/keyconcepts/#data-model)
 - [Exact filter](#exact-filter) - matches logs with the exact value
@@ -490,7 +491,7 @@ This query matches the following [log messages](https://docs.victoriametrics.com
 This query doesn't match the following log messages:
 
 - `Error: foobar`, since the `Error` [word](#word) starts with capital letter. Use `i(err*)` for this case. See [these docs](#case-insensitive-filter) for details.
-- `fooerror`, since the `fooerror` [word](#word) doesn't start with `err`. Use `~"err"` for this case. See [these docs](#regexp-filter) for details.
+- `fooerror`, since the `fooerror` [word](#word) doesn't start with `err`. Use `~"err"` for this case. See [these docs](#substring-filter) for details.
 
 Prefix filter can be applied to [phrases](#phrase-filter). For example, the following query matches
 [log messages](https://docs.victoriametrics.com/victorialogs/keyconcepts/#message-field) containing phrases with `unexpected fail` prefix:
@@ -549,6 +550,32 @@ See also:
 - [Logical filter](#logical-filter)
 
 
+### Substring filter
+
+If it is needed to find logs with some substring, then `~"substring"` filter can be used. For example, the following query matches log entries,
+which contain `ampl` text in the [`_msg` field](https://docs.victoriametrics.com/victorialogs/keyconcepts/#message-field):
+
+```logsql
+~"ampl"
+```
+
+It matches the following messages:
+
+- `Example message`
+- `This is a sample`
+
+It doesn't match `EXAMPLE message`, since `AMPL` substring here is in uppercase. Use `~"(?i)ampl"` filter instead. Note that case-insensitive filter
+may be much slower than case-sensitive one.
+
+Performance tip: prefer using [word filter](#word-filter) and [phrase filter](#phrase-filter), since substring filter may be quite slow.
+
+See also:
+
+- [Word filter](#word-filter)
+- [Phrase filter](#phrase-filter)
+- [Regexp filter](#regexp-filter)
+
+
 ### Empty value filter
 
 Sometimes it is needed to find log entries without the given [log field](https://docs.victoriametrics.com/victorialogs/keyconcepts/#data-model).
diff --git a/lib/logstorage/pipe_unroll.go b/lib/logstorage/pipe_unroll.go
index e1e4e3be9..180b5687b 100644
--- a/lib/logstorage/pipe_unroll.go
+++ b/lib/logstorage/pipe_unroll.go
@@ -74,9 +74,10 @@ func (pu *pipeUnroll) updateNeededFields(neededFields, unneededFields fieldsSet)
 	}
 }
 
-func (pu *pipeUnroll) newPipeProcessor(workersCount int, _ <-chan struct{}, _ func(), ppNext pipeProcessor) pipeProcessor {
+func (pu *pipeUnroll) newPipeProcessor(workersCount int, stopCh <-chan struct{}, _ func(), ppNext pipeProcessor) pipeProcessor {
 	return &pipeUnrollProcessor{
 		pu:     pu,
+		stopCh: stopCh,
 		ppNext: ppNext,
 
 		shards: make([]pipeUnrollProcessorShard, workersCount),
@@ -85,6 +86,7 @@ func (pu *pipeUnroll) newPipeProcessor(workersCount int, _ <-chan struct{}, _ fu
 
 type pipeUnrollProcessor struct {
 	pu     *pipeUnroll
+	stopCh <-chan struct{}
 	ppNext pipeProcessor
 
 	shards []pipeUnrollProcessorShard
@@ -139,6 +141,9 @@ func (pup *pipeUnrollProcessor) writeBlock(workerID uint, br *blockResult) {
 	fields := shard.fields
 	for rowIdx := range br.timestamps {
 		if bm.isSetBit(rowIdx) {
+			if needStop(pup.stopCh) {
+				return
+			}
 			shard.writeUnrolledFields(br, pu.fields, columnValues, rowIdx)
 		} else {
 			fields = fields[:0]
diff --git a/lib/regexutil/regex_test.go b/lib/regexutil/regex_test.go
index 2fedc2d81..a7208f962 100644
--- a/lib/regexutil/regex_test.go
+++ b/lib/regexutil/regex_test.go
@@ -165,6 +165,8 @@ func TestGetLiterals(t *testing.T) {
 	f("foo.*bar(a|b)baz.+", []string{"foo", "bar", "baz"})
 	f("(foo[ab](?:bar))", []string{"foo", "bar"})
 	f("foo|bar", nil)
+	f("(?i)foo", nil)
+	f("foo((?i)bar)baz", []string{"foo", "baz"})
 	f("((foo|bar)baz xxx(?:yzabc))", []string{"baz xxxyzabc"})
 	f("((foo|bar)baz xxx(?:yzabc)*)", []string{"baz xxx"})
 	f("((foo|bar)baz? xxx(?:yzabc)*)", []string{"ba", " xxx"})