Archives/Athou_commafeed
1
0
Fork 1
mirror of https://github.com/Athou/commafeed.git synced 2026-09-23 12:35:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

Trie is thread-safe, no need to recreate one every time

Browse Source
This commit is contained in:
Athou
2026-05-22 20:29:36 +02:00
parent 88e24787e8
commit 5755ff762b
1 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
commafeed-server/src/main/java/com/commafeed/backend/feed/parser/FeedCleaner.java
View File

@@ -15,6 +15,8 @@ public class FeedCleaner {
private static final Pattern DOCTYPE_PATTERN = Pattern.compile("<!DOCTYPE[^>]*>", Pattern.CASE_INSENSITIVE); private static final Pattern DOCTYPE_PATTERN = Pattern.compile("<!DOCTYPE[^>]*>", Pattern.CASE_INSENSITIVE);
private final Trie trie = Trie.builder().ignoreOverlaps().addKeywords(HtmlEntities.HTML_ENTITIES).build();
public String clean(String xml) { public String clean(String xml) {
xml = removeCharactersBeforeFirstXmlTag(xml); xml = removeCharactersBeforeFirstXmlTag(xml);
xml = removeInvalidXmlCharacters(xml); xml = removeInvalidXmlCharacters(xml);
@@ -52,7 +54,7 @@ public class FeedCleaner {
// Create a buffer sufficiently large that re-allocations are minimized. // Create a buffer sufficiently large that re-allocations are minimized.
StringBuilder sb = new StringBuilder(source.length() << 1); StringBuilder sb = new StringBuilder(source.length() << 1);
Collection<Emit> emits = Trie.builder().ignoreOverlaps().addKeywords(HtmlEntities.HTML_ENTITIES).build().parseText(source); Collection<Emit> emits = trie.parseText(source);
int prevIndex = 0; int prevIndex = 0;
for (Emit emit : emits) { for (Emit emit : emits) {