From bf542ec12f1c7ca2f275f90d0b8d0a120a07a461 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jes=C3=BAs=20Espino?= <jespinog@gmail.com>
Date: Wed, 21 Apr 2021 10:29:52 +0200
Subject: [PATCH] Fix performance problem on document extraction (#17470)

---
 services/docextractor/plain.go | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/services/docextractor/plain.go b/services/docextractor/plain.go
index cac6d06e1a..12393e2767 100644
--- a/services/docextractor/plain.go
+++ b/services/docextractor/plain.go
@@ -20,10 +20,15 @@ func (pe *plainExtractor) Extract(filename string, r io.ReadSeeker) (string, err
 	// This detects any visible character plus any whitespace
 	validRanges := append(unicode.GraphicRanges, unicode.White_Space)
 
-	text, _ := ioutil.ReadAll(r)
+	runes := make([]byte, 1028)
+	_, err := r.Read(runes)
+	if err != nil {
+		return "", err
+	}
+
 	count := 0
 	for {
-		c, size := utf8.DecodeRune(text[count:])
+		c, size := utf8.DecodeRune(runes[count:])
 		if !unicode.In(c, validRanges...) {
 			return "", nil
 		}
@@ -36,5 +41,6 @@ func (pe *plainExtractor) Extract(filename string, r io.ReadSeeker) (string, err
 		}
 	}
 
+	text, _ := ioutil.ReadAll(r)
 	return string(text), nil
 }