oai-swift/oAI/Services/WebSearchService.swift

//
//  WebSearchService.swift
//  oAI
//
//  DuckDuckGo web search for non-OpenRouter providers
//
// SPDX-License-Identifier: PolyForm-Noncommercial-1.0.0
// Copyright (C) 2026 Rune Olsen
//
// This file is part of oAI.
//
// oAI is licensed under the PolyForm Noncommercial License 1.0.0.
// You may use, study, modify, and share it for any noncommercial
// purpose. Commercial use — including selling oAI or any part of
// it, standalone or bundled into another product or service —
// requires a separate commercial license from the copyright holder.
//
// See the LICENSE file or
// <https://polyformproject.org/licenses/noncommercial/1.0.0> for
// the full license text. For commercial licensing, contact Rune
// Olsen via <https://oai.pm>.


import Foundation
import os

struct SearchResult: Sendable {
    let title: String
    let url: String
    let snippet: String
}

final class WebSearchService: Sendable {
    nonisolated static let shared = WebSearchService()

    private let session: URLSession

    nonisolated private init() {
        let config = URLSessionConfiguration.default
        config.timeoutIntervalForRequest = 10
        session = URLSession(configuration: config)
    }

    /// Search DuckDuckGo HTML interface (no API key needed)
    nonisolated func search(query: String, maxResults: Int = 5) async -> [SearchResult] {
        Log.search.info("Web search: \(query)")
        guard let encoded = query.addingPercentEncoding(withAllowedCharacters: .urlQueryAllowed),
              let url = URL(string: "https://html.duckduckgo.com/html/?q=\(encoded)")
        else { return [] }

        var request = URLRequest(url: url)
        request.httpMethod = "GET"
        request.setValue(
            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36",
            forHTTPHeaderField: "User-Agent"
        )

        do {
            let (data, _) = try await session.data(for: request)
            guard let html = String(data: data, encoding: .utf8) else { return [] }
            return parseResults(from: html, maxResults: maxResults)
        } catch {
            Log.search.error("Web search failed: \(error.localizedDescription)")
            return []
        }
    }

    /// Format search results as markdown for prompt injection
    nonisolated func formatResults(_ results: [SearchResult], maxLength: Int = 2000) -> String {
        if results.isEmpty { return "No search results found." }

        var formatted = "**Web Search Results:**\n\n"

        for (i, result) in results.enumerated() {
            var entry = "\(i + 1). **\(result.title)**\n"
            entry += "   URL: \(result.url)\n"
            if !result.snippet.isEmpty {
                entry += "   \(result.snippet)\n"
            }
            entry += "\n"

            if formatted.count + entry.count > maxLength {
                formatted += "... (\(results.count - i) more results truncated)\n"
                break
            }

            formatted += entry
        }

        return formatted.trimmingCharacters(in: .whitespacesAndNewlines)
    }

    // MARK: - HTML Parsing

    private nonisolated func parseResults(from html: String, maxResults: Int) -> [SearchResult] {
        var results: [SearchResult] = []

        // Match result blocks: <div class="result results_links ...">
        let blockPattern = #"<div class="result results_links.*?(?=<div class="result results_links|<div id="links")"#
        guard let blockRegex = try? NSRegularExpression(pattern: blockPattern, options: .dotMatchesLineSeparators) else {
            return []
        }

        let range = NSRange(html.startIndex..., in: html)
        let blocks = blockRegex.matches(in: html, range: range)

        for match in blocks.prefix(maxResults) {
            guard let blockRange = Range(match.range, in: html) else { continue }
            let block = String(html[blockRange])

            // Extract title and URL from <a class="result__a" href="...">Title</a>
            let titlePattern = #"<a[^>]*class="result__a"[^>]*href="([^"]+)"[^>]*>([^<]+)</a>"#
            guard let titleRegex = try? NSRegularExpression(pattern: titlePattern),
                  let titleMatch = titleRegex.firstMatch(in: block, range: NSRange(block.startIndex..., in: block)),
                  let urlRange = Range(titleMatch.range(at: 1), in: block),
                  let titleRange = Range(titleMatch.range(at: 2), in: block)
            else { continue }

            var resultURL = String(block[urlRange])
            let title = decodeHTMLEntities(String(block[titleRange]).trimmingCharacters(in: .whitespaces))

            // Extract snippet from <a class="result__snippet" ...>text</a>
            let snippetPattern = #"<a[^>]*class="result__snippet"[^>]*>([^<]+)</a>"#
            var snippet = ""
            if let snippetRegex = try? NSRegularExpression(pattern: snippetPattern),
               let snippetMatch = snippetRegex.firstMatch(in: block, range: NSRange(block.startIndex..., in: block)),
               let snippetRange = Range(snippetMatch.range(at: 1), in: block) {
                snippet = decodeHTMLEntities(String(block[snippetRange]).trimmingCharacters(in: .whitespaces))
            }

            // Decode DDG redirect URL
            if resultURL.contains("uddg=") {
                let uddgPattern = #"uddg=([^&]+)"#
                if let uddgRegex = try? NSRegularExpression(pattern: uddgPattern),
                   let uddgMatch = uddgRegex.firstMatch(in: resultURL, range: NSRange(resultURL.startIndex..., in: resultURL)),
                   let uddgRange = Range(uddgMatch.range(at: 1), in: resultURL) {
                    resultURL = String(resultURL[uddgRange]).removingPercentEncoding ?? resultURL
                }
            }

            results.append(SearchResult(title: title, url: resultURL, snippet: snippet))
        }

        return results
    }

    private nonisolated func decodeHTMLEntities(_ string: String) -> String {
        var result = string
        let entities: [(String, String)] = [
            ("&amp;", "&"), ("&lt;", "<"), ("&gt;", ">"),
            ("&quot;", "\""), ("&#39;", "'"), ("&apos;", "'"),
            ("&#x27;", "'"), ("&#x2F;", "/"), ("&nbsp;", " "),
        ]
        for (entity, char) in entities {
            result = result.replacingOccurrences(of: entity, with: char)
        }
        return result
    }
}