Certain tags don't work when language is set to english. --clean-language doesn't produce any results

[stuffstu]

Using latest version, installed through pip. Windows 11

`python command.py nhentai -s "females only"

[17:16:42] banner: nHentai ver 0.5.3: あなたも変態。 いいね?

[17:16:42] main: Using mirror: https://nhentai.net

[17:16:42] main: Using viewer template "default"

[17:16:42] check_cookie: Login successfully! Your username: lol69xdyuri

[17:16:42] main: Using default language: eng

[17:16:42] search_parser: Searching doujinshis using keywords "females only language:eng" on page 1

[17:16:43] search_parser: No result in response in page 1

[17:16:43] main: All done.`

I've noticed this also occur in a few other tags like 'yaoi' and 'anal intercourse', works perfectly fine if I set the language to Japanese though.
Here's what happens when I run the --clean-language command:

`python command.py nhentai --clean-language
[17:16:08] banner: nHentai ver 0.5.3: あなたも変態。 いいね?
Usage:
nhentai --search [keyword] --download
NHENTAI=https://nhentai-mirror-url/ nhentai --id [ID ...]
nhentai --file [filename]

Environment Variable:
NHENTAI nhentai mirror url

Options:
-h, --help show this help message and exit
-D, --download download doujinshi (for search results)
-S, --show just show the doujinshi information
--id doujinshi ids set, e.g. 167680 167681 167682
-s KEYWORD, --search=KEYWORD
search doujinshi by keyword
-F, --favorites list or download your favorites
--page-all all search results
--page=PAGE, --page-range=PAGE
page number of search results. e.g. 1,2-5,14
--sorting=SORTING, --sort=SORTING
sorting of doujinshi (recent / popular /
popular-[today|week])
-o OUTPUT_DIR, --output=OUTPUT_DIR
output dir
-t THREADS, --threads=THREADS
thread count for downloading doujinshi
-T TIMEOUT, --timeout=TIMEOUT
timeout for downloading doujinshi
-d DELAY, --delay=DELAY
slow down between downloading every doujinshi
--proxy=PROXY store a proxy, for example: -p "http://127.0.0.1:1080"
-f FILE, --file=FILE read gallery IDs from file.
--format=NAME_FORMAT format the saved folder name
--dry-run Dry run, skip file download
--html generate a html viewer at current directory
--no-html don't generate HTML after downloading
--gen-main generate a main viewer contain all the doujin in the
folder
-C, --cbz generate Comic Book CBZ File
-P, --pdf generate PDF file
--rm-origin-dir remove downloaded doujinshi dir when generated CBZ or
PDF file
--meta generate a metadata file in doujinshi format
--regenerate-cbz regenerate the cbz file if exists
--cookie=COOKIE set cookie of nhentai to bypass Cloudflare captcha
--useragent=USERAGENT, --user-agent=USERAGENT
set useragent to bypass Cloudflare captcha
--language=LANGUAGE set default language to parse doujinshis
--clean-language set DEFAULT as language to parse doujinshis
--save-download-history
save downloaded doujinshis, whose will be skipped if
you re-download them
--clean-download-history
clean download history
--template=VIEWER_TEMPLATE
set viewer template
--legacy use legacy searching method`

update: was able to fix the language issue by setting language as blank. Still can't get a response from nh though for the tag "females only". Also doesn't work for the tags previously mentioned above.

python command.py -s "females only"
[19:15:26] banner: nHentai ver 0.5.3: あなたも変態。 いいね?
[19:15:26] main: Using mirror: https://nhentai.net
[19:15:26] main: Using viewer template "default"
[19:15:27] check_cookie: Login successfully! Your username: lol69xdyuri
[19:15:27] search_parser: Searching doujinshis using keywords "females only" on page 1
[19:15:28] search_parser: No result in response in page 1
[19:15:28] main: All done.

Furthermore, it only works when I specify "Japanese".

Starting to think that this is an api issue on nhentai's side. Any possible fixes?

Update 2: Managed to fix it by specifying 'tag:females only'. Idk why but it works.

Update 3: Using the legacy search parser gives full results for the tag I used, whereas the normal would have some page results missing.

[Spyridion]

I'd like to comment that I'm also noticing this problem as well. I noticed some artist searches are returning back 0 results despite seeing quite a few on the website. Ex: nhentai --search="artist:tamagou" --page-all gives {"result":[],"num_pages":0,"per_page":25} in the raw data received. Here is the url generated for the search: https://nhentai.net/api/galleries/search?query=artist%3Atamagou+language%3Aenglish