Merge 40ae65d into 52cd8cf

LodestoneHQ · Mar 6, 2020 · 96ea9c0 · 96ea9c0
2 parents 52cd8cf + 40ae65d
commit 96ea9c0
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/pkg/processor/document/document.go b/pkg/processor/document/document.go
@@ -189,9 +189,8 @@ func (dp *DocumentProcessor) parseDocument(bucketName string, bucketPath string,
 		return model.Document{}, err
 	}
 	//trim whitespace/newline characters
-	dp.logger.Debugf("docContent: %s", docContent)
-	dp.logger.Debugf("trim Content: %s", strings.TrimSpace(docContent))
-	//docContent = strings.TrimSpace(docContent)
+	docContent = strings.TrimSpace(docContent)
+	dp.logger.Debugf("docContent: '%s'", docContent)
 
 	metaFile, err := os.Open(localFilePath)
 	if err != nil {

diff --git a/pkg/processor/document/tika_client.go b/pkg/processor/document/tika_client.go
@@ -13,6 +13,9 @@ func (mrt TikaRoundTripper) RoundTrip(r *http.Request) (*http.Response, error) {
 	if r.URL.Path == "/meta" {
 		r.Header.Add("Accept", "application/json")
 	}
+	if r.URL.Path == "/tika" {
+		r.Header.Add("Accept", "text/plain")
+	}
 
 	return mrt.r.RoundTrip(r)
 }