asp/keep
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

refactor: extract binary detection and word line count logic into helper methods

Browse Source 
Co-authored-by: aider (openai/andrew/openrouter/qwen/qwen3-coder) <aider@aider.chat>
This commit is contained in:
Andrew Phillips
2025-08-26 19:44:01 -03:00
parent 0ad8f3ccfa
commit e9b9532160
1 changed files with 68 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files

130
src/meta_plugin/text.rs
View File

@@ -111,27 +111,10 @@ impl MetaPlugin for TextMetaPlugin {
self.is_finalized = finalized;
}
fn update(&mut self, data: &[u8]) -> MetaPluginResponse {
// If already finalized, don't process more data
if self.is_finalized {
return MetaPluginResponse {
metadata: Vec::new(),
is_finalized: true,
};
}
/// Helper method to perform binary detection and return appropriate metadata
/// Returns (metadata, should_finalize) tuple
fn perform_binary_detection(&mut self, buffer: &[u8]) -> (Vec<crate::meta_plugin::MetaData>, bool) {
let mut metadata = Vec::new();
// If we haven't determined if content is binary yet, build buffer and check
if self.is_binary_content.is_none() {
if let Some(buffer) = &mut self.buffer {
// Add data to our buffer up to max_buffer_size
let remaining_capacity = self.max_buffer_size.saturating_sub(buffer.len());
let bytes_to_take = std::cmp::min(data.len(), remaining_capacity);
buffer.extend_from_slice(&data[..bytes_to_take]);
// If we have enough data to make a binary determination, do it now
if buffer.len() >= std::cmp::min(1024, self.max_buffer_size) {
let is_binary_result = is_binary(buffer);
self.is_binary_content = Some(is_binary_result);
@@ -156,8 +139,64 @@ impl MetaPlugin for TextMetaPlugin {
metadata.push(meta_data);
}
(metadata, is_binary_result)
}
/// Helper method to output word and line counts
fn output_word_line_counts(&mut self) -> Vec<crate::meta_plugin::MetaData> {
let mut metadata = Vec::new();
// Process any remaining data in utf8_buffer
if !self.utf8_buffer.is_empty() {
self.count_text_stats(&[]);
}
// Output word and line counts
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"text_word_count",
self.word_count.to_string(),
self.base.outputs()
) {
metadata.push(meta_data);
}
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"text_line_count",
self.line_count.to_string(),
self.base.outputs()
) {
metadata.push(meta_data);
}
metadata
}
fn update(&mut self, data: &[u8]) -> MetaPluginResponse {
// If already finalized, don't process more data
if self.is_finalized {
return MetaPluginResponse {
metadata: Vec::new(),
is_finalized: true,
};
}
let mut metadata = Vec::new();
// If we haven't determined if content is binary yet, build buffer and check
if self.is_binary_content.is_none() {
if let Some(buffer) = &mut self.buffer {
// Add data to our buffer up to max_buffer_size
let remaining_capacity = self.max_buffer_size.saturating_sub(buffer.len());
let bytes_to_take = std::cmp::min(data.len(), remaining_capacity);
buffer.extend_from_slice(&data[..bytes_to_take]);
// If we have enough data to make a binary determination, do it now
if buffer.len() >= std::cmp::min(1024, self.max_buffer_size) {
let (binary_metadata, is_binary) = self.perform_binary_detection(buffer);
metadata.extend(binary_metadata);
// If it's binary, we're done with this plugin
if is_binary_result {
if is_binary {
self.buffer = None; // Drop the buffer
self.is_finalized = true;
return MetaPluginResponse {
@@ -213,31 +252,11 @@ impl MetaPlugin for TextMetaPlugin {
if self.is_binary_content.is_none() {
if let Some(buffer) = &self.buffer {
if !buffer.is_empty() {
let is_binary_result = is_binary(buffer);
self.is_binary_content = Some(is_binary_result);
// Output text and binary status
let text_value = if is_binary_result { "false".to_string() } else { "true".to_string() };
let binary_value = if is_binary_result { "true".to_string() } else { "false".to_string() };
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"text",
text_value,
self.base.outputs()
) {
metadata.push(meta_data);
}
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"binary",
binary_value,
self.base.outputs()
) {
metadata.push(meta_data);
}
let (binary_metadata, is_binary) = self.perform_binary_detection(buffer);
metadata.extend(binary_metadata);
// If it's binary, we're done
if is_binary_result {
if is_binary {
self.buffer = None; // Drop the buffer
self.is_finalized = true;
return MetaPluginResponse {
@@ -251,27 +270,8 @@ impl MetaPlugin for TextMetaPlugin {
// If content is text, output word and line counts
if self.is_binary_content == Some(false) {
// Process any remaining data in utf8_buffer
if !self.utf8_buffer.is_empty() {
self.count_text_stats(&[]);
}
// Output word and line counts
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"text_word_count",
self.word_count.to_string(),
self.base.outputs()
) {
metadata.push(meta_data);
}
if let Some(meta_data) = crate::meta_plugin::process_metadata_outputs(
"text_line_count",
self.line_count.to_string(),
self.base.outputs()
) {
metadata.push(meta_data);
}
let word_line_metadata = self.output_word_line_counts();
metadata.extend(word_line_metadata);
}
// Drop the buffer since we're done with it