tldr always get data from url without using cache if the platform is empty.

[ramwin]

My system: ubuntu 20.04
My tldr version: latest.

In the code:

tldr-python-client/tldr.py

Lines 194 to 204 in daf1ebf

	def get_page(command, remote=None, platforms=None, languages=None):
	if platforms is None:
	platforms = get_platform_list()
	if languages is None:
	languages = get_language_list()
	for platform in platforms:
	for language in languages:
	if platform is None:
	continue
	try:
	return get_page_for_platform(command, platform, remote, language)

the tldr client will always try to find the command in linux platform first. So if a command is common (like, ls, zip), the tldr command will first try to find cache file only in <cache_dir>/linux/ls.md and since the ls.md is in the <cache_dir>/common/ls.md, not in the <cache_dir>/linux/ls.md, the cache failed.

Then, the client will try to get <remote_url>/linux/ls.md from remote server. and the remote server return 404.

After that, the client will use the platform common and the cache is found.

I suggest change code like this:

for platform in platforms:  # find cache first
    try:
        load_page_from_cache(command, platform, languate)
    except Exception:
        continue
for platform in platforms:  # if failed , get it from server
    try:
        load_page_from_server(command, platform, languate)
    except Exception:
        continue

[ramwin]

I created a pr to fix it.

[ZJUGuoShuai]

Yes, I just encountered the same issue.

In my area I couldn't connect to the URL directly without using an proxy. Consequently, the command tldr find won't work - it doesn't use the cache, the output is:

Error fetching from tldr: <urlopen error [Errno 111] Connection refused>

But when I use tldr -p common find it works :-(

[bl-ue]

Hi, I'll take a look at this tomorrow (UTC-0500 :)

[MasterOdin]

Done with #151