[LUCENE-4048] Move getLines out of ResourceLoader and require Charset

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Duplicate
• Fix Version/s: 4.0-BETA
• Component/s: modules/analysis
• Lucene Fields: New
• Assignee: Unassigned
• Reporter: Chris Male
• Created: 10/May/12 10:53
• Updated: 28/Aug/22 13:16
• Resolved: 08/Apr/20 21:05

Description

ResourceLoader.getLines() is only used by analysis factories. SolrResourceLoader's implementation does the job well and it's unlikely that another ResourceLoader implementation would handle it differently.

We should extract the getLines() method out to AbstractAnalysisFactory so it can be used by the factories. Additionally we shouldn't assume the files are encoded in UTF-8, instead we should allow a Charset to be specified.

This would take us one step closer to reducing the ResourceLoader interface just to what it says, a loader of resources.

Issue Links

• is duplicated by: LUCENE-4257 factor the getLines in ResourceLoader to WordListLoader