[mod] engine: Anna's Archive - fetch search arguments from search form

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

3 files changed, 183 insertions, 4 deletions

diff --git a/searx/data/engine_traits.json b/searx/data/engine_traits.json
index aef8bae0b..072c9a5c4 100644
--- a/searx/data/engine_traits.json
+++ b/searx/data/engine_traits.json
@@ -1,4 +1,134 @@
 {
+  "annas archive": {
+    "all_locale": "",
+    "custom": {
+      "content": [
+        "",
+        "journal_article",
+        "book_any",
+        "book_fiction",
+        "book_unknown",
+        "book_nonfiction",
+        "book_comic",
+        "magazine",
+        "standards_document"
+      ],
+      "ext": [
+        "",
+        "pdf",
+        "epub",
+        "cbr",
+        "fb2",
+        "mobi",
+        "cbz",
+        "djvu",
+        "azw3",
+        "fb2.zip",
+        "txt",
+        "rar",
+        "zip",
+        "doc",
+        "lit",
+        "rtf",
+        "htm",
+        "html",
+        "lrf",
+        "mht",
+        "docx"
+      ],
+      "sort": [
+        "",
+        "newest",
+        "oldest",
+        "largest",
+        "smallest"
+      ]
+    },
+    "data_type": "traits_v1",
+    "languages": {
+      "af": "af",
+      "ar": "ar",
+      "az": "az",
+      "be": "be",
+      "bg": "bg",
+      "bn": "bn",
+      "bo": "bo",
+      "bs": "bs",
+      "ca": "ca",
+      "cs": "cs",
+      "da": "da",
+      "de": "de",
+      "el": "el",
+      "en": "en",
+      "eo": "eo",
+      "es": "es",
+      "et": "et",
+      "eu": "eu",
+      "fa": "fa",
+      "fi": "fi",
+      "fil": "tl",
+      "fr": "fr",
+      "gl": "gl",
+      "gu": "gu",
+      "he": "he",
+      "hi": "hi",
+      "hr": "hr",
+      "hu": "hu",
+      "hy": "hy",
+      "id": "id",
+      "is": "is",
+      "it": "it",
+      "ja": "ja",
+      "ka": "ka",
+      "kk": "kk",
+      "kn": "kn",
+      "ko": "ko",
+      "ku": "ku",
+      "ky": "ky",
+      "lo": "lo",
+      "lt": "lt",
+      "lv": "lv",
+      "mk": "mk",
+      "ml": "ml",
+      "mn": "mn",
+      "mr": "mr",
+      "ms": "ms",
+      "my": "my",
+      "nb": "nb",
+      "ne": "ne",
+      "nl": "nl",
+      "no": "no",
+      "pa": "pa",
+      "pl": "pl",
+      "ps": "ps",
+      "pt": "pt",
+      "ro": "ro",
+      "ru": "ru",
+      "sa": "sa",
+      "sd": "sd",
+      "si": "si",
+      "sk": "sk",
+      "sl": "sl",
+      "so": "so",
+      "sq": "sq",
+      "sr": "sr",
+      "sv": "sv",
+      "sw": "sw",
+      "ta": "ta",
+      "te": "te",
+      "tg": "tg",
+      "tr": "tr",
+      "tt": "tt",
+      "ug": "ug",
+      "uk": "uk",
+      "ur": "ur",
+      "uz": "uz",
+      "vi": "vi",
+      "yi": "yi",
+      "zh": "zh"
+    },
+    "regions": {}
+  },
   "arch linux wiki": {
     "all_locale": null,
     "custom": {
@@ -4127,4 +4257,4 @@
     },
     "regions": {}
   }
-}
\ No newline at end of file
+}
diff --git a/searx/engines/annas_archive.py b/searx/engines/annas_archive.py
index 1d5aa41ee..c845d67c6 100644
--- a/searx/engines/annas_archive.py
+++ b/searx/engines/annas_archive.py
@@ -7,7 +7,8 @@ from typing import List, Dict, Any, Optional
 from urllib.parse import quote
 from lxml import html
 
-from searx.utils import extract_text, eval_xpath
+from searx.utils import extract_text, eval_xpath, eval_xpath_list
+from searx.enginelib.traits import EngineTraits
 
 # about
 about: Dict[str, Any] = {
@@ -42,7 +43,6 @@ def request(query, params: Dict[str, Any]) -> Dict[str, Any]:
         lang = params["language"]
 
     params["url"] = search_url.format(search_query=quote(query), lang=lang)
-    print(params)
     return params
 
 
@@ -66,3 +66,52 @@ def response(resp) -> List[Dict[str, Optional[str]]]:
         results.append(result)
 
     return results
+
+
+def fetch_traits(engine_traits: EngineTraits):
+    """Fetch languages and other search arguments from Anna's search form."""
+    # pylint: disable=import-outside-toplevel
+
+    import babel
+    from searx.network import get  # see https://github.com/searxng/searxng/issues/762
+    from searx.locales import language_tag
+
+    engine_traits.all_locale = ''
+    engine_traits.custom['content'] = []
+    engine_traits.custom['ext'] = []
+    engine_traits.custom['sort'] = []
+
+    resp = get(base_url + '/search')
+    if not resp.ok:  # type: ignore
+        raise RuntimeError("Response from Anna's search page is not OK.")
+    dom = html.fromstring(resp.text)  # type: ignore
+
+    # supported language codes
+
+    lang_map = {}
+    for x in eval_xpath_list(dom, "//form//select[@name='lang']//option"):
+        eng_lang = x.get("value")
+        if eng_lang in ('', '_empty', 'nl-BE', 'und'):
+            continue
+        try:
+            locale = babel.Locale.parse(lang_map.get(eng_lang, eng_lang), sep='-')
+        except babel.UnknownLocaleError:
+            # silently ignore unknown languages
+            # print("ERROR: %s -> %s is unknown by babel" % (x.get("data-name"), eng_lang))
+            continue
+        sxng_lang = language_tag(locale)
+        conflict = engine_traits.languages.get(sxng_lang)
+        if conflict:
+            if conflict != eng_lang:
+                print("CONFLICT: babel %s --> %s, %s" % (sxng_lang, conflict, eng_lang))
+            continue
+        engine_traits.languages[sxng_lang] = eng_lang
+
+    for x in eval_xpath_list(dom, "//form//select[@name='content']//option"):
+        engine_traits.custom['content'].append(x.get("value"))
+
+    for x in eval_xpath_list(dom, "//form//select[@name='ext']//option"):
+        engine_traits.custom['ext'].append(x.get("value"))
+
+    for x in eval_xpath_list(dom, "//form//select[@name='sort']//option"):
+        engine_traits.custom['sort'].append(x.get("value"))
diff --git a/searx/settings.yml b/searx/settings.yml
index 8877fba54..e42373a82 100644
--- a/searx/settings.yml
+++ b/searx/settings.yml
@@ -297,7 +297,7 @@ engines:
     shortcut: 9g
     disabled: true
 
-  - name: anna's archive
+  - name: annas archive
     engine: annas_archive
     paging: false
     categories: files