def fix_image_url(url, repo_name):
'''Fixes a GitHub image urls.
Any links with `github.com` are invalid, because the return *html*
content. Image links would have `githubusercontent.com`. For example:
- This returns an html: https://github.com/Retrothopter/Niobium-Nanotech/blob/master/Preview.png
- This returns a png: https://githubusercontent.com/Retrothopter/Niobium-Nanotech/blob/master/Preview.png
Any links that are relative are also invalid. For example:
- preview.png
- sprites/preview.png
- /sprites/preview.png'''
# FIXME: this assumes `master` is always the branch we want, while in reality we need the
# `default_branch` of the repository, which could also for example be `main`
from urllib.parse import urlparse
from parsec import optional, string, regex, none_of, many, ParseError
glob = (
optional(string('/')) >> string(repo_name) >> string("/blob/master/")
>> many(none_of("?")).parsecmap(lambda x: "".join(x)))
o = urlparse(url)
if o.netloc == "raw.githubusercontent.com":
return url
try:
path = glob.parse(o.path)
except ParseError as e:
path = None
if o.netloc == "github.com" and path:
return f"https://raw.githubusercontent.com/{repo_name}/master/{path}"
if o.netloc == "":
return f"https://raw.githubusercontent.com/{repo_name}/master/{o.path}"
return url
def fix_image_url(url, repo_name):
'''Fixes a GitHub url, where the url should point to an image.
Any links with `github.com` are invalid, because they're html links, while
image links would have `githubusercontent.com`, for example:
- https://github.com/Retrothopter/Niobium-Nanotech/blob/master/Preview.png;
Any links that don't have a domain are relative and as such invalid, for example:
- preview.png;
- sprites/preview.png;
- /sprites/preview.png
This is also why a repo name is required.
'''
from urllib.parse import urlparse
from parsec import optional, string, regex, none_of, many, ParseError
glob = (
optional(string('/')) >> string(repo_name) >> string("/blob/master/")
>> many(none_of("?")).parsecmap(lambda x: "".join(x)))
o = urlparse(url)
if o.netloc == "raw.githubusercontent.com":
return url
try:
path = glob.parse(o.path)
except ParseError as e:
path = None
if o.netloc == "github.com" and path:
return f"https://raw.githubusercontent.com/{repo_name}/master/{path}"
if o.netloc == "":
return f"https://raw.githubusercontent.com/{repo_name}/master/{o.path}"
# print('[warning] non github url:', url)
return url
from parsec import string, none_of, sepBy, many
quoted_char = none_of('"') | string('""').result('"')
quoted = (string('"') >> many(quoted_char) <<
string('"')).parsecmap(lambda x: "".join(x))
cell = quoted | many(none_of(",\n")).parsecmap(lambda x: "".join(x))
cells = sepBy(cell, string(","))
table = sepBy(cells, string("\n"))
def quoted_string():
yield string('"')
chars = yield many(none_of('"'))
yield string('"')
return ''.join(chars)
def rest_line():
"Parse remainder of line."
chars = yield (many(none_of("\n\r")) << ignore
) # So that we still function as a lexeme.
return "".join(chars)
def rest_line():
"Parse remainder of line."
chars = yield (many(none_of("\n\r")) << ignore
) # So that we still function as a lexeme.
return "".join(chars)
skip_line = lexeme(rest_line).result('(Skipped.)')
name_only = regex(r"[_a-zA-Z0-9/\.()#-]+")
name = word(name_only)
symbol = lexeme(regex(r"[a-zA-Z_][^\s()\[\]]*"))
true = lexeme(string("True")).result(True)
false = lexeme(string("False")).result(False)
quoted_string = lexeme(regex(r'"[^"]*"'))
fail = one_of("")
skip_keyword = (skip_line >> many(none_of("[") >> skip_line)).result(
'(Skipped.)') # Skip over everything until the next keyword begins.
IBIS_num_suf = {
'T': 'e12',
'k': 'e3',
'n': 'e-9',
'G': 'e9',
'm': 'e-3',
'p': 'e-12',
'M': 'e6',
'u': 'e-6',
'f': 'e-15',
}
def parse_string():
"""Parse a string as any characters enclosed between two "s."""
yield parsec.string('"')
chars = yield parsec.many(parsec.none_of('"'))
yield parsec.string('"')
return wtypes.String(''.join(chars))