neon-mmd / websurfx

:rocket: An open source alternative to searx which provides a modern-looking :sparkles:, lightning-fast :zap:, privacy respecting :disguised_face:, secure :lock: meta search engine
https://github.com/neon-mmd/websurfx/tree/rolling/docs
GNU Affero General Public License v3.0
758 stars 97 forks source link

:bug: Unexpected random delays in self-hosted instances for `individual/non-production` use. #615

Closed neon-mmd closed 1 month ago

neon-mmd commented 1 month ago

Description

Work Expected From The Issue

Provide a fix for the unexpected random delays in server response in non-production self-hosted instances by fixing the code responsible for conditionally enabling/disabling random delays based on whether the config options debug or production_use were enabled or disabled.

The issue expects the following files to be changed/modified:

[!Note] All the files that are expected to be changed are located under the codebase (websurfx directory).

Reason Behind These Changes

The reason behind having these changes is to only enable random delays for production servers only (which is enabled to evade bot detection and IP banning from the upstream search engines by emulating random delays similar to human based inputs). By providing these changes it significantly improves user experience.

Sample Code

The changes that are required to be made in the search.rs file for this issue are shown below:

//! This module handles the search route of the search engine website.

use crate::{
    cache::cacher::SharedCache,
    config::parser::Config,
    handler::{file_path, FileType},
    models::{
        aggregation_models::SearchResults,
        engine_models::EngineHandler,
        server_models::{self, SearchParams},
    },
    results::aggregator::aggregate,
};
use actix_web::{get, http::header::ContentType, web, HttpRequest, HttpResponse};
use itertools::Itertools;
use regex::Regex;
use std::time::{SystemTime, UNIX_EPOCH};
use std::{borrow::Cow, time::Duration};
use tokio::{
    fs::File,
    io::{AsyncBufReadExt, BufReader},
    join,
};

/// Handles the route of search page of the `websurfx` meta search engine website and it takes
/// two search url parameters `q` and `page` where `page` parameter is optional.
///
/// # Example
///
/// ```bash
/// curl "http://127.0.0.1:8080/search?q=sweden&page=1"
/// ```
///
/// Or
///
/// ```bash
/// curl "http://127.0.0.1:8080/search?q=sweden"
/// ```
#[get("/search")]
pub async fn search(
    req: HttpRequest,
    config: web::Data<&'static Config>,
    cache: web::Data<&'static SharedCache>,
) -> Result<HttpResponse, Box<dyn std::error::Error>> {
    let params = web::Query::<SearchParams>::from_query(req.query_string())?;
    match &params.q {
        Some(query) => {
            if query.trim().is_empty() {
                return Ok(HttpResponse::TemporaryRedirect()
                    .insert_header(("location", "/"))
                    .finish());
            }

            let cookie = req.cookie("appCookie");

            // Get search settings using the user's cookie or from the server's config
            let mut search_settings: server_models::Cookie<'_> = cookie
                .and_then(|cookie_value| serde_json::from_str(cookie_value.value()).ok())
                .unwrap_or_else(|| {
                    server_models::Cookie::build(
                        &config.style,
                        config
                            .upstream_search_engines
                            .iter()
                            .filter_map(|(engine, enabled)| {
                                enabled.then_some(Cow::Borrowed(engine.as_str()))
                            })
                            .collect(),
                        config.safe_search,
                    )
                });

            search_settings.safe_search_level = get_safesearch_level(
                params.safesearch,
                search_settings.safe_search_level,
                config.safe_search,
            );

            // Closure wrapping the results function capturing local references
            let get_results = |page| results(&config, &cache, query, page, &search_settings);

            // .max(1) makes sure that the page >= 0.
            let page = params.page.unwrap_or(1).max(1) - 1;
            let previous_page = page.saturating_sub(1);
            let next_page = page + 1;

            // Add a random delay before making the request.
-            if config.aggregator.random_delay || !config.debug {
+            if config.aggregator.random_delay || config.debug {
                let nanos = SystemTime::now().duration_since(UNIX_EPOCH)?.subsec_nanos() as f32;
                let delay = ((nanos / 1_0000_0000 as f32).floor() as u64) + 1;
                tokio::time::sleep(Duration::from_secs(delay)).await;
            }

            let results: (SearchResults, String, bool);
            if page != previous_page {
                let (previous_results, current_results, next_results) = join!(
                    get_results(previous_page),
                    get_results(page),
                    get_results(next_page)
                );

                results = current_results?;

                let (results_list, cache_keys): (Vec<SearchResults>, Vec<String>) =
                    [previous_results?, results.clone(), next_results?]
                        .into_iter()
                        .filter_map(|(result, cache_key, flag)| flag.then_some((result, cache_key)))
                        .multiunzip();

                tokio::spawn(async move { cache.cache_results(&results_list, &cache_keys).await });
            } else {
                let (current_results, next_results) =
                    join!(get_results(page), get_results(page + 1));

                results = current_results?;

                let (results_list, cache_keys): (Vec<SearchResults>, Vec<String>) =
                    [results.clone(), next_results?]
                        .into_iter()
                        .filter_map(|(result, cache_key, flag)| flag.then_some((result, cache_key)))
                        .multiunzip();

                tokio::spawn(async move { cache.cache_results(&results_list, &cache_keys).await });
            }

            Ok(HttpResponse::Ok().content_type(ContentType::html()).body(
                crate::templates::views::search::search(
                    &config.style.colorscheme,
                    &config.style.theme,
                    &config.style.animation,
                    query,
                    &results.0,
                )
                .0,
            ))
        }
        None => Ok(HttpResponse::TemporaryRedirect()
            .insert_header(("location", "/"))
            .finish()),
    }
}

/// Fetches the results for a query and page. It First checks the redis cache, if that
/// fails it gets proper results by requesting from the upstream search engines.
///
/// # Arguments
///
/// * `url` - It takes the url of the current page that requested the search results for a
/// particular search query.
/// * `config` - It takes a parsed config struct.
/// * `query` - It takes the page number as u32 value.
/// * `req` - It takes the `HttpRequest` struct as a value.
///
/// # Error
///
/// It returns the `SearchResults` struct if the search results could be successfully fetched from
/// the cache or from the upstream search engines otherwise it returns an appropriate error.
async fn results(
    config: &'static Config,
    cache: &'static SharedCache,
    query: &str,
    page: u32,
    search_settings: &server_models::Cookie<'_>,
) -> Result<(SearchResults, String, bool), Box<dyn std::error::Error>> {
    // eagerly parse cookie value to evaluate safe search level
    let safe_search_level = search_settings.safe_search_level;

    let cache_key = format!(
        "http://{}:{}/search?q={}&page={}&safesearch={}&engines={}",
        config.binding_ip,
        config.port,
        query,
        page,
        safe_search_level,
        search_settings.engines.join(",")
    );

    // fetch the cached results json.
    let cached_results = cache.cached_results(&cache_key).await;
    // check if fetched cache results was indeed fetched or it was an error and if so
    // handle the data accordingly.
    match cached_results {
        Ok(results) => Ok((results, cache_key, false)),
        Err(_) => {
            if safe_search_level == 4 {
                let mut results: SearchResults = SearchResults::default();

                let flag: bool =
                    !is_match_from_filter_list(file_path(FileType::BlockList)?, query).await?;
                // Return early when query contains disallowed words,
                if flag {
                    results.set_disallowed();
                    cache
                        .cache_results(&[results.clone()], &[cache_key.clone()])
                        .await?;
                    results.set_safe_search_level(safe_search_level);
                    return Ok((results, cache_key, true));
                }
            }

            // check if the cookie value is empty or not if it is empty then use the
            // default selected upstream search engines from the config file otherwise
            // parse the non-empty cookie and grab the user selected engines from the
            // UI and use that.
            let mut results: SearchResults = match search_settings.engines.is_empty() {
                false => {
                    aggregate(
                        query,
                        page,
                        config,
                        &search_settings
                            .engines
                            .iter()
                            .filter_map(|engine| EngineHandler::new(engine).ok())
                            .collect::<Vec<EngineHandler>>(),
                        safe_search_level,
                    )
                    .await?
                }
                true => {
                    let mut search_results = SearchResults::default();
                    search_results.set_no_engines_selected();
                    search_results
                }
            };
            let (engine_errors_info, results_empty_check, no_engines_selected) = (
                results.engine_errors_info().is_empty(),
                results.results().is_empty(),
                results.no_engines_selected(),
            );
            results.set_filtered(engine_errors_info & results_empty_check & !no_engines_selected);
            cache
                .cache_results(&[results.clone()], &[cache_key.clone()])
                .await?;
            results.set_safe_search_level(safe_search_level);
            Ok((results, cache_key, true))
        }
    }
}

/// A helper function which checks whether the search query contains any keywords which should be
/// disallowed/allowed based on the regex based rules present in the blocklist and allowlist files.
///
/// # Arguments
///
/// * `file_path` - It takes the file path of the list as the argument.
/// * `query` - It takes the search query to be checked against the list as an argument.
///
/// # Error
///
/// Returns a bool indicating whether the results were found in the list or not on success
/// otherwise returns a standard error type on a failure.
async fn is_match_from_filter_list(
    file_path: &str,
    query: &str,
) -> Result<bool, 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)?;
        if re.is_match(query) {
            return Ok(true);
        }
    }

    Ok(false)
}

/// A helper function to choose the safe search level value based on the URL parameters,
/// cookie value and config value.
///
/// # Argurments
///
/// * `safe_search_level_from_url` - Safe search level from the URL parameters.
/// * `cookie_safe_search_level` - Safe search level value from the cookie.
/// * `config_safe_search_level` - Safe search level value from the config file.
///
/// # Returns
///
/// Returns an appropriate safe search level value based on the safe search level values
/// from the URL parameters, cookie and the config file.
fn get_safesearch_level(
    safe_search_level_from_url: Option<u8>,
    cookie_safe_search_level: u8,
    config_safe_search_level: u8,
) -> u8 {
    (u8::from(safe_search_level_from_url.is_some())
        * ((u8::from(config_safe_search_level >= 3) * config_safe_search_level)
            + (u8::from(config_safe_search_level < 3) * safe_search_level_from_url.unwrap_or(0))))
        + (u8::from(safe_search_level_from_url.is_none())
            * ((u8::from(config_safe_search_level >= 3) * config_safe_search_level)
                + (u8::from(config_safe_search_level < 3) * cookie_safe_search_level)))
}

#[cfg(test)]
mod tests {
    use std::time::{SystemTime, UNIX_EPOCH};

    /// A helper function which creates a random mock safe search level value.
    ///
    /// # Returns
    ///
    /// Returns an optional u8 value.
    fn mock_safe_search_level_value() -> Option<u8> {
        let nanos = SystemTime::now()
            .duration_since(UNIX_EPOCH)
            .unwrap_or_default()
            .subsec_nanos() as f32;
        let delay = ((nanos / 1_0000_0000 as f32).floor() as i8) - 1;

        match delay {
            -1 => None,
            some_num => Some(if some_num > 4 { some_num - 4 } else { some_num } as u8),
        }
    }

    #[test]
    /// A test function to test whether the output of the branchless and branched code
    /// for the code to choose the appropriate safe search level is same or not.
    fn get_safesearch_level_branched_branchless_code_test() {
        // Get mock values for the safe search level values for URL parameters, cookie
        // and config.
        let safe_search_level_from_url = mock_safe_search_level_value();
        let cookie_safe_search_level = mock_safe_search_level_value().unwrap_or(0);
        let config_safe_search_level = mock_safe_search_level_value().unwrap_or(0);

        // Branched code
        let safe_search_level_value_from_branched_code = match safe_search_level_from_url {
            Some(safe_search_level_from_url_parsed) => {
                if config_safe_search_level >= 3 {
                    config_safe_search_level
                } else {
                    safe_search_level_from_url_parsed
                }
            }
            None => {
                if config_safe_search_level >= 3 {
                    config_safe_search_level
                } else {
                    cookie_safe_search_level
                }
            }
        };

        // branchless code
        let safe_search_level_value_from_branchless_code =
            (u8::from(safe_search_level_from_url.is_some())
                * ((u8::from(config_safe_search_level >= 3) * config_safe_search_level)
                    + (u8::from(config_safe_search_level < 3)
                        * safe_search_level_from_url.unwrap_or(0))))
                + (u8::from(safe_search_level_from_url.is_none())
                    * ((u8::from(config_safe_search_level >= 3) * config_safe_search_level)
                        + (u8::from(config_safe_search_level < 3) * cookie_safe_search_level)));

        assert_eq!(
            safe_search_level_value_from_branched_code,
            safe_search_level_value_from_branchless_code
        );
    }
}

[!Note]

  1. To get started contributing make sure to read the contributing.md file for the guidlines on how to contribute in this project
  2. To contribute first fork this project by following this video tutorial if you are not familliar with process and add your changes and make a pull request with the changes to this repository and if you are new to GitHub then follow this video tutorial to get started contributing :slightly_smiling_face: .

Screenshots

No response

Do you want to work on this issue?

None

Additional information

No response

github-actions[bot] commented 1 month ago

The issue has been unlocked and is now ready for dev. If you would like to work on this issue, you can comment to have it assigned to you. You can learn more in our contributing guide https://github.com/neon-mmd/websurfx/blob/rolling/CONTRIBUTING.md

KekmaTime commented 1 month ago

@neon-mmd ill work on this issue

neon-mmd commented 1 month ago

@neon-mmd ill work on this issue

Yes sure, we will assign this issue to you now. You may start working on it right away :rocket: :slightly_smiling_face: