searxng/searx/engines/yandex.py

# SPDX-License-Identifier: AGPL-3.0-or-later
"""Yandex (Web, images)"""

from datetime import datetime, timedelta
import typing as t
from json import JSONDecodeError, loads
from urllib.parse import urlencode
from lxml import html

from searx.search.processors.abstract import TimeRangeType
from searx.exceptions import SearxEngineCaptchaException, SearxEngineResponseException
from searx.result_types import EngineResults
from searx.utils import humanize_bytes, eval_xpath, eval_xpath_list, extract_text, html_to_text

if t.TYPE_CHECKING:
    from searx import logger  # logger is injected by searx.engines.set_loggers()
    from searx.extended_types import SXNG_Response
    from searx.search.processors import OnlineParams

# Engine metadata
about = {
    "website": "https://yandex.com/",
    "wikidata_id": "Q5281",
    "official_api_documentation": "?",
    "use_official_api": False,
    "require_api_key": False,
    "results": "HTML",
}

# Engine configuration
categories = []
time_range_support = True
paging = True
enable_http3 = True

search_type = ""

# Search URL
base_url_web = "https://yandex.com/search/site/"
base_url_images = "https://yandex.com/images/search"

# Supported languages
yandex_supported_langs = [
    "ru",  # Russian
    "en",  # English
    "be",  # Belarusian
    "fr",  # French
    "de",  # German
    "id",  # Indonesian
    "kk",  # Kazakh
    "tt",  # Tatar
    "tr",  # Turkish
    "uk",  # Ukrainian
]

results_xpath = '//li[contains(@class, "serp-item")]'
url_xpath = './/a[@class="b-serp-item__title-link"]/@href'
title_xpath = './/h3[@class="b-serp-item__title"]/a[@class="b-serp-item__title-link"]/span'
content_xpath = './/div[@class="b-serp-item__content"]//div[@class="b-serp-item__text"]'
date_xpath = './/span[@data-mtime]/@data-mtime'


def catch_bad_response(resp: "SXNG_Response") -> None:
    if resp.headers.get("x-yandex-captcha") == "captcha":
        raise SearxEngineCaptchaException()


time_range_to_days_map: dict[TimeRangeType, int] = {
    "day": 1,
    "week": 7,
    "month": 30,
    "year": 365,
}


def _time_range_to_start_date(time_range: TimeRangeType) -> datetime:
    diff = timedelta(days=time_range_to_days_map[time_range])
    return datetime.now() - diff


def request(query: str, params: "OnlineParams") -> None:
    args: dict[str, t.Any] = {
        "text": query,
    }
    if params['pageno'] > 1:
        args["p"] = params["pageno"] - 1

    if search_type == 'web':
        lang = params["searxng_locale"].split("-")[0]
        if lang in yandex_supported_langs:
            args["lang"] = lang

        if time_range := params["time_range"]:
            start_date = _time_range_to_start_date(time_range)
            end_date = datetime.now()
            args.update(
                {
                    "within": 777,  # must be set for time range search, meaning unclear
                    "from_day": start_date.day,
                    "from_month": start_date.month,
                    "from_year": start_date.year,
                    "to_day": end_date.day,
                    "to_month": end_date.month,
                    "to_year": end_date.year,
                }
            )

        args.update(
            {
                "tmpl_version": "releases",
                "web": "1",
                "frame": "1",
                "searchid": "3131712",
            }
        )
        params['url'] = f"{base_url_web}?{urlencode(args)}"
    elif search_type == 'images':
        params['url'] = f"{base_url_images}?{urlencode(args)}"


def _parse_json_results(dom: html.HtmlElement) -> dict[str, t.Any]:
    # attempt to parse using xpath - finding element with "data-state" attribute
    data_elements = dom.xpath("//*[@data-state]/@data-state")
    for json_data in data_elements:
        try:
            json_resp = loads(json_data)
            if json_resp.get("location") == "/images/search/":
                return json_resp
        except JSONDecodeError:
            logger.debug("failed parsing data-state json")

    raise SearxEngineResponseException("failed to parse JSON data")


def response(resp: "SXNG_Response") -> EngineResults:
    catch_bad_response(resp)
    results = EngineResults()
    dom = html.fromstring(resp.text)

    match search_type:
        case "web":
            for result in eval_xpath_list(dom, results_xpath):
                url = extract_text(eval_xpath(result, url_xpath))
                title = extract_text(eval_xpath(result, title_xpath))
                content = extract_text(eval_xpath(result, content_xpath))

                # published date is only shown if time range search is used
                publishedDate = None
                if publishedDateMillis := extract_text(eval_xpath(result, date_xpath)):
                    publishedDate = datetime.utcfromtimestamp(int(publishedDateMillis))
                results.add(
                    results.types.MainResult(
                        url=url, title=str(title), content=str(content), publishedDate=publishedDate
                    )
                )
        case "images":
            json_resp = _parse_json_results(dom)

            # build results from loaded json values
            for item_data in json_resp["initialState"]["serpList"]["items"]["entities"].values():
                viewerData: dict[str, t.Any] = item_data["viewerData"]
                snippet: dict[str, str] = viewerData.get("snippet", {})

                # return the image with largest dimensions
                image_sources = viewerData.get("dups", []) + viewerData.get("preview", [])
                image_source = max(image_sources, key=lambda x: x["h"] * x["w"])

                humanized_filesize = None
                if image_source.get("fileSizeInBytes"):
                    humanized_filesize = humanize_bytes(image_source["fileSizeInBytes"])

                results.add(
                    results.types.Image(
                        title=snippet.get("title"),  # type: ignore
                        content=html_to_text(snippet.get("text")),  # type: ignore
                        url=snippet.get("url"),
                        img_src=image_source["url"],
                        filesize=humanized_filesize,  # type: ignore
                        thumbnail_src=item_data["image"],
                        resolution=f"{image_source['w']} x {image_source['h']}",
                    )
                )

    return results