morrigan/crabbysearch
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

remove bad if not even unnecessary relevance calculation and its dependencies

Browse source
This commit is contained in:
Milim 2024-08-15 09:09:41 +02:00
parent 215a928eb4
commit 05b6dfda86
No known key found for this signature in database
4 changed files with 3 additions and 268 deletions

86
Cargo.lock generated
View file

@ -235,12 +235,6 @@ version = "1.0.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f26201604c87b1e01bd3d98f8d5d9a8fcbb815e8cedb41ffccbeb4bf593a35fe"
[[package]]
name = "adler32"
version = "1.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "aae1277d39aeec15cb388266ecc24b11c80469deae6067e17a1a7aa9e5c1f234"
[[package]]
name = "ahash"
version = "0.8.11"
@ -298,12 +292,6 @@ dependencies = [
"tokio 1.39.2",
]
[[package]]
name = "async-once-cell"
version = "0.5.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9338790e78aa95a416786ec8389546c4b6a1dfc3dc36071ed9518a9413a542eb"
[[package]]
name = "async-trait"
version = "0.1.81"
@ -1493,16 +1481,6 @@ dependencies = [
"winapi-build",
]
[[package]]
name = "keyword_extraction"
version = "1.4.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f0efa28e79b3a5f72586318c07c24477a169c688e5065fde647c71b3952a2d42"
dependencies = [
"regex",
"unicode-segmentation",
]
[[package]]
name = "language-tags"
version = "0.3.2"
@ -1521,26 +1499,6 @@ version = "0.2.155"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "97b3888a4aecf77e811145cadf6eef5901f4782c53886191b2f693f24761847c"
[[package]]
name = "libflate"
version = "1.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5ff4ae71b685bbad2f2f391fe74f6b7659a34871c08b210fdc039e43bee07d18"
dependencies = [
"adler32",
"crc32fast",
"libflate_lz77",
]
[[package]]
name = "libflate_lz77"
version = "1.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a52d3a8bfc85f250440e4424db7d857e241a3aebbbe301f3eb606ab15c39acbf"
dependencies = [
"rle-decode-fast",
]
[[package]]
name = "linux-raw-sys"
version = "0.4.14"
@ -2567,12 +2525,6 @@ dependencies = [
"windows-sys 0.52.0",
]
[[package]]
name = "rle-decode-fast"
version = "1.0.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3582f63211428f83597b51b2ddb88e2a91a9d52d12831f9d08f5e624e8977422"
[[package]]
name = "rustc-demangle"
version = "0.1.24"
@ -2952,15 +2904,6 @@ version = "1.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a8f112729512f8e442d81f95a8a7ddf2b7c6b8a1a6f509a95864142b30cab2d3"
[[package]]
name = "stop-words"
version = "0.8.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8500024d809de02ecbf998472b7bed3c4fca380df2be68917f6a473bdb28ddcc"
dependencies = [
"serde_json",
]
[[package]]
name = "string"
version = "0.2.1"
@ -3117,25 +3060,6 @@ dependencies = [
"utf-8",
]
[[package]]
name = "thesaurus"
version = "0.5.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3e33ea271e53da683cd3439c04ff3b734f3d6052ea33a65ec9e0fa89a4f96369"
dependencies = [
"lazy_static",
"thesaurus-moby",
]
[[package]]
name = "thesaurus-moby"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "28f7806d5dbe7d9b627e332f88269a014a6a1d40ec411d4ea66cb702aabce4cf"
dependencies = [
"libflate",
]
[[package]]
name = "thiserror"
version = "1.0.63"
@ -3526,12 +3450,6 @@ dependencies = [
"tinyvec",
]
[[package]]
name = "unicode-segmentation"
version = "1.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d4c87d22b6e3f4a18d4d40ef354e97c90fcb14dd91d7dc0aa9d8a1172ebf7202"
[[package]]
name = "unicode-xid"
version = "0.1.0"
@ -3741,7 +3659,6 @@ dependencies = [
"actix-files",
"actix-web",
"async-compression",
"async-once-cell",
"async-trait",
"cfg-if 1.0.0",
"env_logger",
@ -3749,7 +3666,6 @@ dependencies = [
"fake-useragent",
"figment",
"futures 0.3.30",
"keyword_extraction",
"log",
"maud",
"mini-moka",
@ -3758,9 +3674,7 @@ dependencies = [
"scraper",
"serde",
"serde_json",
"stop-words",
"tempfile",
"thesaurus",
"tokio 1.39.2",
]

11
Cargo.toml
View file

@ -41,7 +41,6 @@ error-stack = { version = "0.4.0", default-features = false, features = [
async-trait = { version = "0.1.80", default-features = false }
regex = { version = "1.9.4", features = ["perf"], default-features = false }
futures = { version = "0.3.30", default-features = false, features = ["alloc"] }
async-once-cell = { version = "0.5.3", default-features = false }
mini-moka = { version = "0.10", default-features = false, features = [
"sync",
] }
@ -50,20 +49,12 @@ async-compression = { version = "0.4.11", default-features = false, features = [
"tokio",
], optional = true }
cfg-if = { version = "1.0.0", default-features = false, optional = true }
keyword_extraction = { version = "1.4.3", default-features = false, features = [
"tf_idf",
] }
figment = { version = "0.10", features = ["env"] }
stop-words = { version = "0.8.0", default-features = false, features = ["iso"] }
thesaurus = { version = "0.5.2", default-features = false, optional = true, features = [
"moby",
] }
[dev-dependencies]
tempfile = { version = "3.10.1", default-features = false }
[features]
use-synonyms-search = ["thesaurus/static"]
compress-cache-results = ["dep:async-compression", "dep:cfg-if"]
experimental-io-uring = ["actix-web/experimental-io-uring"]
use-non-static-synonyms-search = ["thesaurus"]

94
src/models/aggregation_models.rs
View file

@ -4,11 +4,6 @@
use super::engine_models::EngineError;
use serde::{Deserialize, Serialize};
#[cfg(any(
feature = "use-synonyms-search",
feature = "use-non-static-synonyms-search"
))]
use thesaurus::synonyms;
/// A named struct to store the raw scraped search results scraped search results from the
/// upstream search engines before aggregating it.It derives the Clone trait which is needed
/// to write idiomatic rust using `Iterators`.
@ -47,45 +42,6 @@ impl SearchResult {
engine: engine.iter().map(|name| name.to_string()).collect(),
}
}
/// calculates and update the relevance score of the current search.
/// # Arguments
///
/// * query - the query string used to obtain the results
///
///
pub fn calculate_relevance(&mut self, query: &str) {
use stop_words::{get, LANGUAGE};
// when language settings can change to any of the ones supported on this crate: https://docs.rs/crate/stop-words/0.8.0
let documents = [
self.title.clone(),
self.url.clone(),
self.description.clone(),
];
let stop_words = get(LANGUAGE::English);
let punctuation = [
".".to_owned(),
",".to_owned(),
":".to_owned(),
";".to_owned(),
"!".to_owned(),
"?".to_owned(),
"(".to_owned(),
")".to_owned(),
"[".to_owned(),
"]".to_owned(),
"{".to_owned(),
"}".to_owned(),
"\"".to_owned(),
"'".to_owned(),
"<".to_owned(),
">".to_owned(),
];
self.relevance_score = calculate_tf_idf(query, &documents, &stop_words, &punctuation);
}
/// A function which adds the engine name provided as a string into a vector of strings.
///
@ -228,53 +184,3 @@ impl SearchResults {
self.no_engines_selected = true;
}
}
/// Helper function to calculate the tf-idf for the search query.
/// <br> The approach is as [`as`](https://en.wikipedia.org/wiki/Tf%E2%80%93idf).
/// <br> Find a sample article about TF-IDF [`here`](https://medium.com/analytics-vidhya/tf-idf-term-frequency-technique-easiest-explanation-for-text-classification-in-nlp-with-code-8ca3912e58c3)
/// ### Arguments
/// * `query` - a user's search query
/// * `documents` - a list of text used for comparision (url, title, description)
/// * `stop_words` - A list of language specific stop words.
/// * `punctuation` - list of punctuation symbols.
/// ### Returns
/// * `score` - The average tf-idf score of the word tokens (and synonyms) in the query
fn calculate_tf_idf(
query: &str,
documents: &[String],
stop_words: &[String],
punctuation: &[String],
) -> f32 {
use keyword_extraction::{
tf_idf::{TfIdf, TfIdfParams},
tokenizer::Tokenizer,
};
let params = TfIdfParams::UnprocessedDocuments(documents, stop_words, Some(punctuation));
let tf_idf = TfIdf::new(params);
let tokener = Tokenizer::new(query, stop_words, Some(punctuation));
let query_tokens = tokener.split_into_words();
let mut search_tokens = vec![];
for token in query_tokens {
#[cfg(any(
feature = "use-synonyms-search",
feature = "use-non-static-synonyms-search"
))]
{
// find some synonyms and add them to the search (from wordnet or moby if feature is enabled)
let synonyms = synonyms(&token);
search_tokens.extend(synonyms)
}
search_tokens.push(token);
}
let mut total_score = 0.0f32;
for token in search_tokens.iter() {
total_score += tf_idf.get_score(token);
}
let result = total_score / (search_tokens.len() as f32);
f32::from(!result.is_nan()) * result
}

80
src/results/aggregator.rs
View file

@ -10,15 +10,9 @@ use crate::models::{
use error_stack::Report;
use futures::stream::FuturesUnordered;
use regex::Regex;
use reqwest::{Client, ClientBuilder};
use std::sync::Arc;
use tokio::{
fs::File,
io::{AsyncBufReadExt, BufReader},
task::JoinHandle,
time::Duration,
};
use tokio::{task::JoinHandle, time::Duration};
/// A constant for holding the prebuilt Client globally in the app.
static CLIENT: std::sync::OnceLock<Client> = std::sync::OnceLock::new();
@ -153,77 +147,7 @@ pub async fn aggregate(
};
}
let mut results: Vec<SearchResult> = result_map
.iter()
.map(|(_, value)| {
let mut copy = value.clone();
if !copy.url.contains("temu.com") {
copy.calculate_relevance(query.as_str())
}
copy
})
.collect();
sort_search_results(&mut results);
let results: Vec<SearchResult> = result_map.iter().map(|(_, value)| value.clone()).collect();
Ok(SearchResults::new(results, &engine_errors_info))
}
/// Filters a map of search results using a list of regex patterns.
///
/// # Arguments
///
/// * `map_to_be_filtered` - A mutable reference to a `Vec` of search results to filter, where the filtered results will be removed from.
/// * `resultant_map` - A mutable reference to a `Vec` to hold the filtered results.
/// * `file_path` - A `&str` representing the path to a file containing regex patterns to use for filtering.
///
/// # Errors
///
/// Returns an error if the file at `file_path` cannot be opened or read, or if a regex pattern is invalid.
pub async fn filter_with_lists(
map_to_be_filtered: &mut Vec<(String, SearchResult)>,
resultant_map: &mut Vec<(String, SearchResult)>,
file_path: &str,
) -> Result<(), Box<dyn std::error::Error>> {
let reader = BufReader::new(File::open(file_path).await?);
let mut lines = reader.lines();
while let Some(line) = lines.next_line().await? {
let re = Regex::new(line.trim())?;
let mut length = map_to_be_filtered.len();
let mut idx: usize = Default::default();
// Iterate over each search result in the map and check if it matches the regex pattern
while idx < length {
let ele = &map_to_be_filtered[idx];
let ele_inner = &ele.1;
match re.is_match(&ele.0.to_lowercase())
|| re.is_match(&ele_inner.title.to_lowercase())
|| re.is_match(&ele_inner.description.to_lowercase())
{
true => {
// If the search result matches the regex pattern, move it from the original map to the resultant map
resultant_map.push(map_to_be_filtered.swap_remove(idx));
length -= 1;
}
false => idx += 1,
};
}
}
Ok(())
}
/// Sorts SearchResults by relevance score.
/// <br> sort_unstable is used as its faster,stability is not an issue on our side.
/// For reasons why, check out [`this`](https://rust-lang.github.io/rfcs/1884-unstable-sort.html)
/// # Arguments
/// * `results` - A mutable slice or Vec of SearchResults
///
fn sort_search_results(results: &mut [SearchResult]) {
results.sort_unstable_by(|a, b| {
use std::cmp::Ordering;
b.relevance_score
.partial_cmp(&a.relevance_score)
.unwrap_or(Ordering::Less)
})
}