mirrors/zed
1
0
Fork 0
mirror of https://github.com/zed-industries/zed.git synced 2025-01-09 02:44:49 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 10

expanded embeddable context to accomodate for struct context and file paths

Browse source
This commit is contained in:
KCaverly 2023-07-11 15:58:33 -04:00
parent dd0dbdc5bd
commit 02f523094b
1 changed files with 31 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

46
crates/vector_store/src/parsing.rs
View file

@ -20,6 +20,9 @@ pub struct ParsedFile {
pub documents: Vec<Document>,
}
const CODE_CONTEXT_TEMPLATE: &str =
"The below code snippet is from file '<path>'\n\n```<language>\n<item>\n```";
pub struct CodeContextRetriever {
pub parser: Parser,
pub cursor: QueryCursor,
@ -58,27 +61,40 @@ impl CodeContextRetriever {
tree.root_node(),
content.as_bytes(),
) {
let mut item_range: Option<Range<usize>> = None;
let mut name_range: Option<Range<usize>> = None;
let mut name: Vec<&str> = vec![];
let mut item: Option<&str> = None;
let mut offset: Option<usize> = None;
for capture in mat.captures {
if capture.index == embedding_config.item_capture_ix {
item_range = Some(capture.node.byte_range());
offset = Some(capture.node.byte_range().start);
item = content.get(capture.node.byte_range());
} else if capture.index == embedding_config.name_capture_ix {
name_range = Some(capture.node.byte_range());
if let Some(name_content) = content.get(capture.node.byte_range()) {
name.push(name_content);
}
}
if let Some(context_capture_ix) = embedding_config.context_capture_ix {
if capture.index == context_capture_ix {
if let Some(context) = content.get(capture.node.byte_range()) {
name.push(context);
}
}
}
}
if let Some((item_range, name_range)) = item_range.zip(name_range) {
if let Some((item, name)) =
content.get(item_range.clone()).zip(content.get(name_range))
{
context_spans.push(item.to_string());
documents.push(Document {
name: name.to_string(),
offset: item_range.start,
embedding: Vec::new(),
});
}
if item.is_some() && offset.is_some() && name.len() > 0 {
let context_span = CODE_CONTEXT_TEMPLATE
.replace("<path>", pending_file.relative_path.to_str().unwrap())
.replace("<language>", &pending_file.language.name().to_lowercase())
.replace("<item>", item.unwrap());
context_spans.push(context_span);
documents.push(Document {
name: name.join(" "),
offset: offset.unwrap(),
embedding: Vec::new(),
})
}
}