diff options
| author | sigoden <sigoden@gmail.com> | 2024-07-02 08:23:16 +0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-07-02 08:23:16 +0800 |
| commit | 2f2b13c891d19e29b18ff09f7f540d7fded15887 (patch) | |
| tree | 7a2c63a512519aa5a7872a73797dcd87cda09add /src/config | |
| parent | 7c6dac061b2a33aae334c08b725e02459c5e9ade (diff) | |
| download | aichat-2f2b13c891d19e29b18ff09f7f540d7fded15887.tar.gz | |
refactor: improve RAG (#675)
Diffstat (limited to 'src/config')
| -rw-r--r-- | src/config/mod.rs | 16 |
1 files changed, 10 insertions, 6 deletions
diff --git a/src/config/mod.rs b/src/config/mod.rs index 25891bd..db58b3d 100644 --- a/src/config/mod.rs +++ b/src/config/mod.rs @@ -1446,12 +1446,16 @@ impl Config { } fn setup_document_loaders(&mut self) { - [("pdf", "pdftotext $1 -"), ("docx", "pandoc --to plain $1")] - .into_iter() - .for_each(|(k, v)| { - let (k, v) = (k.to_string(), v.to_string()); - self.document_loaders.entry(k).or_insert(v); - }); + [ + ("pdf", "pdftotext $1 -"), + ("docx", "pandoc --to plain $1"), + (RECURSIVE_URL_LOADER, "rag-crawler $1 $2"), + ] + .into_iter() + .for_each(|(k, v)| { + let (k, v) = (k.to_string(), v.to_string()); + self.document_loaders.entry(k).or_insert(v); + }); } } |
