Duplicate logic for formatting human readable file sizes

[Bnyro]

@return42 added a humanize_bytes function in e76ab1a.

However I figured out when working on a PR recently, that we already had a similar logic before https://github.com/searxng/searxng/blob/master/searx/templates/simple/result_templates/files.html#L38 used in the files.html template - it however covers less cases that the humanize_bytes function.

We should probably replace the logic in files.html with using humanize_bytes.

[return42]

Yeah, good point .. we should normalize all the implementation to ..

searxng/searx/utils.py

Lines 355 to 364 in e021441

	def humanize_bytes(size, precision=2):
	"""Determine the *human readable* value of bytes on 1024 base (1KB=1024B)."""
	s = ['B ', 'KB', 'MB', 'GB', 'TB']
	x = len(s)
	p = 0
	while size > 1024 and p < x:
	p += 1
	size = size / 1024.0
	return "%.*f %s" % (precision, size, s[p])

You already mentioned the template ..

searxng/searx/templates/simple/result_templates/files.html

Lines 38 to 45 in e021441

	{%- if result.size %}<tr><td>{{ _('Filesize') }}</td><td>
	{%- if result.size < 1024 %}{{ result.size }} {{ _('Bytes') -}}
	{%- elif result.size < 1024*1024 %}{{ '{0:0.2f}'.format(result.size/1024) }} {{ _('kiB') -}}
	{%- elif result.size < 1024*1024*1024 %}{{ '{0:0.2f}'.format(result.size/1024/1024) }} {{ _('MiB') -}}
	{%- elif result.size < 1024*1024*1024*1024 %}{{ '{0:0.2f}'.format(result.size/1024/1024/1024) }} {{ _('GiB') -}}
	{%- else %}{{ '{0:0.2f}'.format(result.size/1024/1024/1024/1024) }} {{ _('TiB') }}{% endif -%}
	</td></tr>
	{%- endif -%}

Haven't look deep, but it seems to me the torrent engines first do calculate from a human readable format to the number of bytes (see utils.get_torrent_size(..)) and in the file template of the torrent engine the number of bytes is calculated back to a human readable format .. may this can be also be simplified a little bit(?).

./searx/engines/btdigg.py:66:        filesize = get_torrent_size(filesize, filesize_multiplier)
./searx/engines/tokyotoshokan.py:78:                    params['filesize'] = get_torrent_size(groups[0], groups[1])
./searx/engines/1337x.py:45:        filesize = get_torrent_size(filesize, filesize_multiplier)
./searx/engines/bt4g.py:117:                'filesize': get_torrent_size(filesizeParsed[0], filesizeParsed[1]),
./searx/engines/piratebay.py:90:            filesize = get_torrent_size(result["size"], "B")
./searx/engines/digbt.py\08:from searx.utils import extract_text, get_torrent_size
./searx/engines/digbt.py:48:        filesize = get_torrent_size(files_data[FILESIZE], files_data[FILESIZE_MULTIPLIER])
./searx/engines/solidtorrents.py:17:    get_torrent_size,
./searx/engines/solidtorrents.py:66:            'filesize': get_torrent_size(*extract_text(stats[1]).split()),
./searx/engines/nyaa.py:106:            filesize = get_torrent_size(*filesize_info.split())
./searx/engines/kickass.py:57:        result['filesize'] = get_torrent_size(*extract_text(eval_xpath(tag, './/td[contains(@class, "nobr")]')).split())

searxng/searx/utils.py

Lines 332 to 352 in e021441

	def get_torrent_size(filesize: str, filesize_multiplier: str) -> Optional[int]:
	"""
	Args:
	* filesize (str): size
	* filesize_multiplier (str): TB, GB, .... TiB, GiB...
	Returns:
	* int: number of bytes
	Example:
	>>> get_torrent_size('5', 'GB')
	5368709120
	>>> get_torrent_size('3.14', 'MiB')
	3140000
	"""
	try:
	multiplier = _STORAGE_UNIT_VALUE.get(filesize_multiplier, 1)
	return int(float(filesize) * multiplier)
	except ValueError:
	return None