[mod] lower memory footprint by lazy loading JSON data

This patch implements lazy loading of the JSON data.

Motivation: in most requests not all JSON data is needed, but loaded.  By
example these four JSON files:

- currencies.json ~550KB
- engine_descriptions.json ~1,3MB
- external_bangs.json ~1,3MB
- osm_keys_tags.json ~ 2,2MB

most often not used and consume a lot of memory and BTW they also extend the
time required to instantiate a walker.

Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>

searx/data/__init__.py

@@ -20,13 +20,20 @@ __all__ = [
 
 import json
 from pathlib import Path
+from searx import logger
 
 data_dir = Path(__file__).parent
+logger = logger.getChild('data')
 
-
-def _load(filename):
-    with open(data_dir / filename, encoding='utf-8') as f:
-        return json.load(f)
+CURRENCIES: dict
+USER_AGENTS: dict
+EXTERNAL_URLS: dict
+WIKIDATA_UNITS: dict
+EXTERNAL_BANGS: dict
+OSM_KEYS_TAGS: dict
+ENGINE_DESCRIPTIONS: dict
+ENGINE_TRAITS: dict
+LOCALES: dict
 
 
 def ahmia_blacklist_loader():
@@ -42,12 +49,27 @@ def ahmia_blacklist_loader():
         return f.read().split()
 
 
-CURRENCIES = _load('currencies.json')
-USER_AGENTS = _load('useragents.json')
-EXTERNAL_URLS = _load('external_urls.json')
-WIKIDATA_UNITS = _load('wikidata_units.json')
-EXTERNAL_BANGS = _load('external_bangs.json')
-OSM_KEYS_TAGS = _load('osm_keys_tags.json')
-ENGINE_DESCRIPTIONS = _load('engine_descriptions.json')
-ENGINE_TRAITS = _load('engine_traits.json')
-LOCALES = _load('locales.json')
+NAME_TO_JSON_FILE = {
+    'CURRENCIES': 'currencies.json',
+    'USER_AGENTS': 'useragents.json',
+    'EXTERNAL_URLS': 'external_urls.json',
+    'WIKIDATA_UNITS': 'wikidata_units.json',
+    'EXTERNAL_BANGS': 'external_bangs.json',
+    'OSM_KEYS_TAGS': 'osm_keys_tags.json',
+    'ENGINE_DESCRIPTIONS': 'engine_descriptions.json',
+    'ENGINE_TRAITS': 'engine_traits.json',
+    'LOCALES': 'locales.json',
+}
+
+
+def __getattr__(name: str):
+    # lazy load of JSON files ..
+    filename = NAME_TO_JSON_FILE.get(name)
+    if filename:
+        filename = data_dir / filename
+        logger.debug("init global %s from JSON file %s", name, filename)
+        with open(filename, encoding='utf-8') as f:
+            globals()[name] = json.load(f)
+            return globals()[name]
+    else:
+        raise AttributeError(f"module {__name__!r} has no attribute {name!r}")

searx/enginelib/traits.py

@@ -16,7 +16,7 @@ import types
 from typing import Dict, Literal, Iterable, Union, Callable, Optional, TYPE_CHECKING
 
 from searx import locales
-from searx.data import data_dir, ENGINE_TRAITS
+from searx import data
 
 if TYPE_CHECKING:
     from . import Engine
@@ -193,7 +193,7 @@ class EngineTraits:
 class EngineTraitsMap(Dict[str, EngineTraits]):
     """A python dictionary to map :class:`EngineTraits` by engine name."""
 
-    ENGINE_TRAITS_FILE = (data_dir / 'engine_traits.json').resolve()
+    ENGINE_TRAITS_FILE = (data.data_dir / 'engine_traits.json').resolve()
     """File with persistence of the :py:obj:`EngineTraitsMap`."""
 
     def save_data(self):
@@ -205,7 +205,7 @@ class EngineTraitsMap(Dict[str, EngineTraits]):
     def from_data(cls) -> 'EngineTraitsMap':
         """Instantiate :class:`EngineTraitsMap` object from :py:obj:`ENGINE_TRAITS`"""
         obj = cls()
-        for k, v in ENGINE_TRAITS.items():
+        for k, v in data.ENGINE_TRAITS.items():
             obj[k] = EngineTraits(**v)
         return obj
 

searx/engines/annas_archive.py

@@ -37,9 +37,9 @@ from typing import List, Dict, Any, Optional
 from urllib.parse import quote
 from lxml import html
 
+from searx import data
 from searx.utils import extract_text, eval_xpath, eval_xpath_list
 from searx.enginelib.traits import EngineTraits
-from searx.data import ENGINE_TRAITS
 
 # about
 about: Dict[str, Any] = {
@@ -86,7 +86,7 @@ aa_ext: str = ''
 
 def init(engine_settings=None):  # pylint: disable=unused-argument
     """Check of engine's settings."""
-    traits = EngineTraits(**ENGINE_TRAITS['annas archive'])
+    traits = EngineTraits(**data.ENGINE_TRAITS['annas archive'])
 
     if aa_content and aa_content not in traits.custom['content']:
         raise ValueError(f'invalid setting content: {aa_content}')

searx/engines/duckduckgo.py

@@ -12,6 +12,7 @@ import babel
 import lxml.html
 
 from searx import (
+    data,
     locales,
     redislib,
     external_bang,
@@ -230,7 +231,7 @@ def quote_ddg_bangs(query):
     for val in re.split(r'(\s+)', query):
         if not val.strip():
             continue
-        if val.startswith('!') and external_bang.get_node(external_bang.EXTERNAL_BANGS, val[1:]):
+        if val.startswith('!') and external_bang.get_node(data.EXTERNAL_BANGS, val[1:]):
             val = f"'{val}'"
         query_parts.append(val)
     return ' '.join(query_parts)

searx/engines/duckduckgo_definitions.py

@@ -18,7 +18,7 @@ from typing import TYPE_CHECKING
 from urllib.parse import urlencode, urlparse, urljoin
 from lxml import html
 
-from searx.data import WIKIDATA_UNITS
+from searx import data
 from searx.utils import extract_text, html_to_text, get_string_replaces_function
 from searx.external_urls import get_external_url, get_earth_coordinates_url, area_to_osm_zoom
 
@@ -238,7 +238,7 @@ def unit_to_str(unit):
     for prefix in WIKIDATA_PREFIX:
         if unit.startswith(prefix):
             wikidata_entity = unit[len(prefix) :]
-            real_unit = WIKIDATA_UNITS.get(wikidata_entity)
+            real_unit = data.WIKIDATA_UNITS.get(wikidata_entity)
             if real_unit is None:
                 return unit
             return real_unit['symbol']

searx/engines/openstreetmap.py

@@ -10,7 +10,7 @@ from functools import partial
 
 from flask_babel import gettext
 
-from searx.data import OSM_KEYS_TAGS, CURRENCIES
+from searx import data as searx_data
 from searx.utils import searx_useragent
 from searx.external_urls import get_external_url
 from searx.engines.wikidata import send_wikidata_query, sparql_string_escape, get_thumbnail
@@ -435,7 +435,7 @@ def get_label(labels, lang):
 def get_tag_label(tag_category, tag_name, lang):
     """Get tag label from OSM_KEYS_TAGS"""
     tag_name = '' if tag_name is None else tag_name
-    tag_labels = OSM_KEYS_TAGS['tags'].get(tag_category, {}).get(tag_name, {})
+    tag_labels = searx_data.OSM_KEYS_TAGS['tags'].get(tag_category, {}).get(tag_name, {})
     return get_label(tag_labels, lang)
 
 
@@ -449,12 +449,12 @@ def get_key_label(key_name, lang):
         # https://taginfo.openstreetmap.org/keys/currency#values
         currency = key_name.split(':')
         if len(currency) > 1:
-            o = CURRENCIES['iso4217'].get(currency[1])
+            o = searx_data.CURRENCIES['iso4217'].get(currency[1])
             if o:
                 return get_label(o, lang).lower()
             return currency[1]
 
-    labels = OSM_KEYS_TAGS['keys']
+    labels = searx_data.OSM_KEYS_TAGS['keys']
     for k in key_name.split(':') + ['*']:
         labels = labels.get(k)
         if labels is None:

searx/engines/wikidata.py

@@ -13,7 +13,7 @@ from json import loads
 from dateutil.parser import isoparse
 from babel.dates import format_datetime, format_date, format_time, get_datetime_format
 
-from searx.data import WIKIDATA_UNITS
+from searx import data
 from searx.network import post, get
 from searx.utils import searx_useragent, get_string_replaces_function
 from searx.external_urls import get_external_url, get_earth_coordinates_url, area_to_osm_zoom
@@ -762,7 +762,7 @@ def debug_explain_wikidata_query(query, method='GET'):
 
 def init(engine_settings=None):  # pylint: disable=unused-argument
     # WIKIDATA_PROPERTIES : add unit symbols
-    WIKIDATA_PROPERTIES.update(WIKIDATA_UNITS)
+    WIKIDATA_PROPERTIES.update(data.WIKIDATA_UNITS)
 
     # WIKIDATA_PROPERTIES : add property labels
     wikidata_property_names = []

searx/engines/zlibrary.py

@@ -40,9 +40,9 @@ from urllib.parse import quote
 from lxml import html
 from flask_babel import gettext
 
+from searx import data
 from searx.utils import extract_text, eval_xpath, eval_xpath_list
 from searx.enginelib.traits import EngineTraits
-from searx.data import ENGINE_TRAITS
 
 if TYPE_CHECKING:
     import httpx
@@ -80,7 +80,7 @@ zlib_ext: str = ""
 
 def init(engine_settings=None) -> None:  # pylint: disable=unused-argument
     """Check of engine's settings."""
-    traits: EngineTraits = EngineTraits(**ENGINE_TRAITS["z-library"])
+    traits: EngineTraits = EngineTraits(**data.ENGINE_TRAITS["z-library"])
 
     if zlib_ext and zlib_ext not in traits.custom["ext"]:
         raise ValueError(f"invalid setting ext: {zlib_ext}")

searx/external_bang.py

@@ -2,7 +2,7 @@
 # pylint: disable=missing-module-docstring
 
 from urllib.parse import quote_plus, urlparse
-from searx.data import EXTERNAL_BANGS
+from searx import data
 
 LEAF_KEY = chr(16)
 
@@ -56,7 +56,7 @@ def resolve_bang_definition(bang_definition, query):
 
 def get_bang_definition_and_autocomplete(bang, external_bangs_db=None):  # pylint: disable=invalid-name
     if external_bangs_db is None:
-        external_bangs_db = EXTERNAL_BANGS
+        external_bangs_db = data.EXTERNAL_BANGS
 
     bang_definition, bang_ac_list = get_bang_definition_and_ac(external_bangs_db, bang)
 
@@ -90,7 +90,7 @@ def get_bang_url(search_query, external_bangs_db=None):
     ret_val = None
 
     if external_bangs_db is None:
-        external_bangs_db = EXTERNAL_BANGS
+        external_bangs_db = data.EXTERNAL_BANGS
 
     if search_query.external_bang:
         bang_definition, _ = get_bang_definition_and_ac(external_bangs_db, search_query.external_bang)

searx/external_urls.py

@@ -3,7 +3,7 @@
 
 import math
 
-from searx.data import EXTERNAL_URLS
+from searx import data
 
 
 IMDB_PREFIX_TO_URL_ID = {
@@ -43,7 +43,7 @@ def get_external_url(url_id, item_id, alternative="default"):
         elif url_id == 'wikimedia_image':
             item_id = get_wikimedia_image_id(item_id)
 
-    url_description = EXTERNAL_URLS.get(url_id)
+    url_description = data.EXTERNAL_URLS.get(url_id)
     if url_description:
         url_template = url_description["urls"].get(alternative)
         if url_template is not None:

searx/plugins/ahmia_filter.py

@@ -2,7 +2,7 @@
 # pylint: disable=missing-module-docstring
 
 from hashlib import md5
-from searx.data import ahmia_blacklist_loader
+from searx import data
 
 name = "Ahmia blacklist"
 description = "Filter out onion results that appear in Ahmia's blacklist. (See https://ahmia.fi/blacklist)"
@@ -24,5 +24,5 @@ def init(_app, settings):
     if not settings['outgoing']['using_tor_proxy']:
         # disable the plugin
         return False
-    ahmia_blacklist = ahmia_blacklist_loader()
+    ahmia_blacklist = data.ahmia_blacklist_loader()
     return True

searx/plugins/unit_converter.py

@@ -4,7 +4,7 @@
 
 from flask_babel import gettext
 
-from searx.data import WIKIDATA_UNITS
+from searx import data
 
 name = "Unit converter plugin"
 description = gettext("Convert between units")
@@ -38,7 +38,7 @@ def _parse_text_and_convert(search, splitted_query):
     from_unit = None
     to_unit = None
 
-    for unit in WIKIDATA_UNITS.values():
+    for unit in data.WIKIDATA_UNITS.values():
         if unit['symbol'] == from_unit_key:
             from_unit = unit
 

searx/search/processors/online_currency.py

@@ -6,7 +6,7 @@
 import unicodedata
 import re
 
-from searx.data import CURRENCIES
+from searx import data
 from .online import OnlineProcessor
 
 parser_re = re.compile('.*?(\\d+(?:\\.\\d+)?) ([^.0-9]+) (?:in|to) ([^.0-9]+)', re.I)
@@ -20,14 +20,14 @@ def normalize_name(name):
 
 def name_to_iso4217(name):
     name = normalize_name(name)
-    currency = CURRENCIES['names'].get(name, [name])
+    currency = data.CURRENCIES['names'].get(name, [name])
     if isinstance(currency, str):
         return currency
     return currency[0]
 
 
 def iso4217_to_name(iso4217, language):
-    return CURRENCIES['iso4217'].get(iso4217, {}).get(language, iso4217)
+    return data.CURRENCIES['iso4217'].get(iso4217, {}).get(language, iso4217)
 
 
 class OnlineCurrencyProcessor(OnlineProcessor):

searx/utils.py

@@ -21,7 +21,7 @@ from lxml import html
 from lxml.etree import ElementBase, XPath, XPathError, XPathSyntaxError
 
 from searx import settings
-from searx.data import USER_AGENTS, data_dir
+from searx import data as searx_data
 from searx.version import VERSION_TAG
 from searx.sxng_locales import sxng_locales
 from searx.exceptions import SearxXPathSyntaxException, SearxEngineXPathException
@@ -81,7 +81,9 @@ def gen_useragent(os_string: Optional[str] = None) -> str:
 
     See searx/data/useragents.json
     """
-    return USER_AGENTS['ua'].format(os=os_string or choice(USER_AGENTS['os']), version=choice(USER_AGENTS['versions']))
+    return searx_data.USER_AGENTS['ua'].format(
+        os=os_string or choice(searx_data.USER_AGENTS['os']), version=choice(searx_data.USER_AGENTS['versions'])
+    )
 
 
 class _HTMLTextExtractorException(Exception):
@@ -600,7 +602,7 @@ def _get_fasttext_model() -> "fasttext.FastText._FastText":  # type: ignore
 
         # Monkey patch: prevent fasttext from showing a (useless) warning when loading a model.
         fasttext.FastText.eprint = lambda x: None
-        _FASTTEXT_MODEL = fasttext.load_model(str(data_dir / 'lid.176.ftz'))
+        _FASTTEXT_MODEL = fasttext.load_model(str(searx_data.data_dir / 'lid.176.ftz'))
     return _FASTTEXT_MODEL
 
 

searx/webapp.py

@@ -58,7 +58,7 @@ from searx import infopage
 from searx import limiter
 from searx.botdetection import link_token
 
-from searx.data import ENGINE_DESCRIPTIONS
+from searx import data
 from searx.results import Timing
 from searx.settings_defaults import OUTPUT_FORMATS
 from searx.settings_loader import get_default_settings_path
@@ -1102,14 +1102,14 @@ def image_proxy():
 @app.route('/engine_descriptions.json', methods=['GET'])
 def engine_descriptions():
     locale = get_locale().split('_')[0]
-    result = ENGINE_DESCRIPTIONS['en'].copy()
+    result = data.ENGINE_DESCRIPTIONS['en'].copy()
     if locale != 'en':
-        for engine, description in ENGINE_DESCRIPTIONS.get(locale, {}).items():
+        for engine, description in data.ENGINE_DESCRIPTIONS.get(locale, {}).items():
             result[engine] = description
     for engine, description in result.items():
         if len(description) == 2 and description[1] == 'ref':
             ref_engine, ref_lang = description[0].split(':')
-            description = ENGINE_DESCRIPTIONS[ref_lang][ref_engine]
+            description = data.ENGINE_DESCRIPTIONS[ref_lang][ref_engine]
         if isinstance(description, str):
             description = [description, 'wikipedia']
         result[engine] = description