Skip to content

fix: correct chunk overlap calculation to use configured overlap #25

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
echobt wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: correct chunk overlap calculation to use configured overlap #25

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
58 changes: 48 additions & 10 deletions src/core/indexer.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -351,11 +351,30 @@ impl Indexer {
end_line: (line_idx - 1) as i32,
});

let overlap_start = if line_idx > 0 {
line_idx.saturating_sub(self.chunk_overlap / 40)
} else {
0
};
// Calculate overlap based on characters, not arbitrary "magic number 40" division.
// We want to keep enough lines from the end of the current chunk to meet the overlap requirement.
let mut overlap_char_count = 0;
let mut overlap_lines_count = 0;
let mut overlap_start = line_idx;

// Walk backwards from current line to find where overlap should start
// Ensure we don't go back past the start of the current chunk to avoid infinite loops or weirdness,
// although logic suggests we just want *some* overlap from the immediately preceding content.
// Actually, we can go back as far as needed within the lines we have processed.
for i in (0..line_idx).rev() {
let line_len = lines[i].len() + 1; // +1 for newline
if overlap_char_count + line_len > self.chunk_overlap {
break;
}
overlap_char_count += line_len;
overlap_lines_count += 1;
overlap_start = i;
}

// If we couldn't find any lines (e.g. one huge line), just overlap the last line if possible
if overlap_lines_count == 0 && line_idx > 0 {
overlap_start = line_idx - 1;
}

current_chunk = lines[overlap_start..line_idx].join("\n");
if !current_chunk.is_empty() {
Expand Down Expand Up @@ -727,11 +746,30 @@ impl ServerIndexer {
end_line: (line_idx - 1) as i32,
});

let overlap_start = if line_idx > 0 {
line_idx.saturating_sub(self.chunk_overlap / 40)
} else {
0
};
// Calculate overlap based on characters, not arbitrary "magic number 40" division.
// We want to keep enough lines from the end of the current chunk to meet the overlap requirement.
let mut overlap_char_count = 0;
let mut overlap_lines_count = 0;
let mut overlap_start = line_idx;

// Walk backwards from current line to find where overlap should start
// Ensure we don't go back past the start of the current chunk to avoid infinite loops or weirdness,
// although logic suggests we just want *some* overlap from the immediately preceding content.
// Actually, we can go back as far as needed within the lines we have processed.
for i in (0..line_idx).rev() {
let line_len = lines[i].len() + 1; // +1 for newline
if overlap_char_count + line_len > self.chunk_overlap {
break;
}
overlap_char_count += line_len;
overlap_lines_count += 1;
overlap_start = i;
}

// If we couldn't find any lines (e.g. one huge line), just overlap the last line if possible
if overlap_lines_count == 0 && line_idx > 0 {
overlap_start = line_idx - 1;
}

current_chunk = lines[overlap_start..line_idx].join("\n");
if !current_chunk.is_empty() {
Expand Down
26 changes: 21 additions & 5 deletions src/watcher.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -415,11 +415,27 @@ impl FileWatcher {
end_line: (line_idx - 1) as i32,
});

let overlap_start = if line_idx > 0 {
line_idx.saturating_sub(self.config.chunk_overlap / 40)
} else {
0
};
// Calculate overlap based on characters, not arbitrary "magic number 40" division.
// We want to keep enough lines from the end of the current chunk to meet the overlap requirement.
let mut overlap_char_count = 0;
let mut overlap_lines_count = 0;
let mut overlap_start = line_idx;

// Walk backwards from current line to find where overlap should start
for i in (0..line_idx).rev() {
let line_len = lines[i].len() + 1; // +1 for newline
if overlap_char_count + line_len > self.config.chunk_overlap {
break;
}
overlap_char_count += line_len;
overlap_lines_count += 1;
overlap_start = i;
}

// If we couldn't find any lines (e.g. one huge line), just overlap the last line if possible
if overlap_lines_count == 0 && line_idx > 0 {
overlap_start = line_idx - 1;
}

current_chunk = lines[overlap_start..line_idx].join("\n");
if !current_chunk.is_empty() {
Expand Down
Loading