[feat] add spellcheck functionality to searxng, closes #3816

requirements.txt

@@ -16,3 +16,4 @@ redis==5.0.8
 markdown-it-py==3.0.0
 fasttext-predict==0.9.2.2
 pytomlpp==1.0.13; python_version < '3.11'
+pyspellchecker>=0.8.1

searx/templates/simple/preferences.html

@@ -176,6 +176,9 @@
     {% if 'safesearch' not in locked_preferences %}
       {%- include 'simple/preferences/safesearch.html' -%}
     {%- endif -%}
+    {% if 'spellcheck' not in locked_preferences %}
+      {%- include 'simple/preferences/spellcheck.html' -%}
+    {%- endif -%}
     {{- plugin_preferences('general') -}}
     {%- if 'doi_resolver' not in locked_preferences %}
       {%- include 'simple/preferences/doi_resolver.html' -%}

searx/templates/simple/preferences/spellcheck.html

@@ -0,0 +1,18 @@
+<fieldset>{{- '' -}}
+    <legend id="pref_spellcheck">{{- _('Spell Check') -}}</legend>{{- '' -}}
+    <div class="value">{{- '' -}}
+      <select name='spellcheck' aria-labelledby="pref_spellcheck">{{- '' -}}
+        <option value="1"
+                {%- if spellcheck == '1' %} selected="selected" {%- endif -%}>
+                {{- _('On') -}}
+        </option>{{- '' -}}
+        <option value="0"
+                {%- if spellcheck == '0' %} selected="selected" {%- endif -%}>
+                {{- _('Off') -}}
+        </option>{{- '' -}}
+      </select>{{- '' -}}
+    </div>{{- '' -}}
+    <div class="description">
+      {{- _('Spell check search queries') -}}
+    </div>{{- '' -}}
+  </fieldset>{{- '' -}}

searx/templates/simple/results.html

@@ -92,6 +92,25 @@
     </div>
     {% endif %}
 
+    {% if has_misspelled_words %}
+    <form method="{{ method or 'POST' }}" action="{{ url_for('search') }}">
+      <input type="hidden" name="q" value="{{ spellcheck_query }}" >
+      {% for category in selected_categories %}
+      <input type="hidden" name="category_{{ category }}" value="1" >
+      {% endfor %}
+      <input type="hidden" name="pageno" value="{{ pageno+1 }}" >
+      <input type="hidden" name="language" value="{{ current_language }}" >
+      <input type="hidden" name="time_range" value="{{ time_range }}" >
+      <input type="hidden" name="safesearch" value="{{ safesearch }}" >
+      <input type="hidden" name="theme" value="{{ theme }}" >
+      {% if timeout_limit %}<input type="hidden" name="timeout_limit" value="{{ timeout_limit|e }}" >{% endif %}
+      {{- engine_data_form(engine_data) -}}
+      <span style="color: #ff7769; font-size: 18px"> Did you mean:
+      <input style="background: none; border: none; color: #1a0dab; cursor: pointer; font-size: 20px" type="submit" role="link" value="{{ spellcheck_query }}">
+      </span>
+    </form>
+    {% endif %}
+
     <div id="urls" role="main">
     {% for result in results %}
         {% if result.open_group and not only_template %}<div class="template_group_{{ result['template']|replace('.html', '') }}">{% endif %}

searx/webapp.py

@@ -21,6 +21,7 @@ from typing import List, Dict, Iterable
 import urllib
 import urllib.parse
 from urllib.parse import urlencode, urlparse, unquote
+from spellchecker import SpellChecker
 
 import httpx
 
@@ -254,6 +255,17 @@ def code_highlighter(codelines, language=None):
     return html_code
 
 
+def spellcheck(query: str, user_language: str):
+    spell = SpellChecker(language=user_language)
+    spellcheck_list = query.split(' ')
+    spellcheck_list = [x for x in spellcheck_list if x != '']
+    has_misspelled_words = bool(spell.unknown(spellcheck_list))
+    for word, i in zip(spellcheck_list, range(0, len(spellcheck_list))):
+        if word != '':
+            spellcheck_list[i] = spell.correction(word) if spell.correction(word) is not None else ''
+    return ' '.join(spellcheck_list), has_misspelled_words
+
+
 def get_result_template(theme_name: str, template_name: str):
     themed_path = theme_name + '/result_templates/' + template_name
     if themed_path in result_templates:
@@ -760,6 +772,11 @@ def search():
             result_container.corrections,
         )
     )
+    spellcheck_query, has_misspelled_words = '', False
+    if request.preferences.get_value("spellcheck") == "1" and search.search_query.lang[0:2] in SpellChecker.languages():
+        spellcheck_query, has_misspelled_words = spellcheck(
+            request.form['q'], user_language=search.search_query.lang[0:2]
+        )
 
     # engine_timings: get engine response times sorted from slowest to fastest
     engine_timings = sorted(result_container.get_timings(), reverse=True, key=lambda e: e.total)
@@ -774,6 +791,8 @@ def search():
         'results.html',
         results = results,
         q=request.form['q'],
+        spellcheck_query = spellcheck_query,
+        has_misspelled_words = has_misspelled_words,
         selected_categories = search_query.categories,
         pageno = search_query.pageno,
         time_range = search_query.time_range or '',
@@ -1010,6 +1029,7 @@ def preferences():
         current_locale = request.preferences.get_value("locale"),
         image_proxy = image_proxy,
         engines_by_category = engines_by_category,
+        spellcheck = request.preferences.get_value("spellcheck"),
         stats = stats,
         max_rate95 = max_rate95,
         reliabilities = reliabilities,