gvsig-scripting / org.gvsig.scripting / trunk / org.gvsig.scripting / org.gvsig.scripting.app / org.gvsig.scripting.app.mainplugin / src / main / resources-plugin / scripting / lib / pynliner / soupselect.py @ 475
History | View | Annotate | Download (8.33 KB)
| 1 |
"""
|
|---|---|
| 2 |
# Included with pynliner since it isn't on PyPI #
|
| 3 |
|
| 4 |
soupselect.py
|
| 5 |
|
| 6 |
CSS selector support for BeautifulSoup.
|
| 7 |
|
| 8 |
soup = BeautifulSoup('<html>...')
|
| 9 |
select(soup, 'div')
|
| 10 |
- returns a list of div elements
|
| 11 |
|
| 12 |
select(soup, 'div#main ul a')
|
| 13 |
- returns a list of links inside a ul inside div#main
|
| 14 |
|
| 15 |
patched to support multiple class selectors here http://code.google.com/p/soupselect/issues/detail?id=4#c0
|
| 16 |
"""
|
| 17 |
import re |
| 18 |
import BeautifulSoup |
| 19 |
|
| 20 |
attribute_regex = re.compile('\[(?P<attribute>\w+)(?P<operator>[=~\|\^\$\*]?)=?["\']?(?P<value>[^\]"]*)["\']?\]')
|
| 21 |
pseudo_class_regex = re.compile(ur':(([^:.#(*\[]|\([^)]+\))+)')
|
| 22 |
|
| 23 |
def get_attribute_checker(operator, attribute, value=''): |
| 24 |
"""
|
| 25 |
Takes an operator, attribute and optional value; returns a function that
|
| 26 |
will return True for elements that match that combination.
|
| 27 |
"""
|
| 28 |
return {
|
| 29 |
'=': lambda el: el.get(attribute) == value, |
| 30 |
# attribute includes value as one of a set of space separated tokens
|
| 31 |
'~': lambda el: value in el.get(attribute, '').split(), |
| 32 |
# attribute starts with value
|
| 33 |
'^': lambda el: el.get(attribute, '').startswith(value), |
| 34 |
# attribute ends with value
|
| 35 |
'$': lambda el: el.get(attribute, '').endswith(value), |
| 36 |
# attribute contains value
|
| 37 |
'*': lambda el: value in el.get(attribute, ''), |
| 38 |
# attribute is either exactly value or starts with value-
|
| 39 |
'|': lambda el: el.get(attribute, '') == value \ |
| 40 |
or el.get(attribute, '').startswith('%s-' % value), |
| 41 |
}.get(operator, lambda el: el.has_key(attribute))
|
| 42 |
|
| 43 |
def is_white_space(el): |
| 44 |
if isinstance(el, BeautifulSoup.NavigableString) and str(el).strip() == '': |
| 45 |
return True |
| 46 |
if isinstance(el, BeautifulSoup.Comment): |
| 47 |
return True |
| 48 |
return False |
| 49 |
|
| 50 |
def is_last_content_node(el): |
| 51 |
result = False
|
| 52 |
if el is None: |
| 53 |
result = True
|
| 54 |
elif is_white_space(el):
|
| 55 |
result = is_last_content_node(el.nextSibling) |
| 56 |
return result
|
| 57 |
|
| 58 |
def is_first_content_node(el): |
| 59 |
result = False
|
| 60 |
if el is None: |
| 61 |
result = True
|
| 62 |
if is_white_space(el):
|
| 63 |
result = is_first_content_node(el.previousSibling) |
| 64 |
return result
|
| 65 |
|
| 66 |
def get_pseudo_class_checker(psuedo_class): |
| 67 |
"""
|
| 68 |
Takes a psuedo_class, like "first-child" or "last-child"
|
| 69 |
and returns a function that will check if the element satisfies
|
| 70 |
that psuedo class
|
| 71 |
"""
|
| 72 |
return {
|
| 73 |
'first-child': lambda el: is_first_content_node(getattr(el, 'previousSibling', None)), |
| 74 |
'last-child': lambda el: is_last_content_node(getattr(el, 'nextSibling', None)) |
| 75 |
}.get(psuedo_class, lambda el: False) |
| 76 |
|
| 77 |
def get_checker(functions): |
| 78 |
def checker(el): |
| 79 |
for func in functions: |
| 80 |
if not func(el): |
| 81 |
return False |
| 82 |
return el
|
| 83 |
return checker
|
| 84 |
|
| 85 |
|
| 86 |
def select(soup, selector): |
| 87 |
"""
|
| 88 |
soup should be a BeautifulSoup instance; selector is a CSS selector
|
| 89 |
specifying the elements you want to retrieve.
|
| 90 |
"""
|
| 91 |
handle_token = True
|
| 92 |
current_context = [(soup, [])] |
| 93 |
operator = None
|
| 94 |
while selector:
|
| 95 |
if handle_token:
|
| 96 |
# Get the rightmost token
|
| 97 |
handle_token = False
|
| 98 |
match = re.search('([_0-9a-zA-Z-#.:*"\'\[\\]=]+)$', selector)
|
| 99 |
if not match: |
| 100 |
raise Exception("No match was found. We're done or something is broken") |
| 101 |
token = match.groups(1)[0] |
| 102 |
|
| 103 |
# remove this token from the selector
|
| 104 |
selector = selector.rsplit(token, 1)[0].rstrip() |
| 105 |
|
| 106 |
checker_functions = [] |
| 107 |
#
|
| 108 |
# Get attribute selectors from token
|
| 109 |
#
|
| 110 |
matches = attribute_regex.findall(token) |
| 111 |
for match in matches: |
| 112 |
checker_functions.append(get_attribute_checker(match[1], match[0], match[2])) |
| 113 |
|
| 114 |
#
|
| 115 |
# Get pseudo classes from token
|
| 116 |
#
|
| 117 |
for match in pseudo_class_regex.finditer(token): |
| 118 |
checker_functions.append(get_pseudo_class_checker(match.groups(1)[0])) |
| 119 |
|
| 120 |
checker = get_checker(checker_functions) |
| 121 |
#
|
| 122 |
# Get tag
|
| 123 |
#
|
| 124 |
tag = re.findall('^([a-zA-Z0-9]+)', token)
|
| 125 |
if len(tag) == 0: |
| 126 |
tag = True
|
| 127 |
elif len(tag) == 1: |
| 128 |
tag = tag[0]
|
| 129 |
else:
|
| 130 |
raise Exception("Multiple tags found (invalid CSS)") |
| 131 |
|
| 132 |
#
|
| 133 |
# Get ID
|
| 134 |
#
|
| 135 |
ids = re.findall('#([a-zA-Z0-9_-]+)', token)
|
| 136 |
if len(ids) > 1: |
| 137 |
raise Exception("Only single # OK") |
| 138 |
#
|
| 139 |
# Get classes
|
| 140 |
#
|
| 141 |
classes = re.findall('\.([a-zA-Z0-9_-]+)', token)
|
| 142 |
|
| 143 |
#
|
| 144 |
# Search contexts for matches
|
| 145 |
#
|
| 146 |
found = [] |
| 147 |
find_dict = {}
|
| 148 |
if ids:
|
| 149 |
find_dict['id'] = ids
|
| 150 |
if classes:
|
| 151 |
find_dict['class'] = lambda attr: attr and set(classes).issubset(attr.split()) |
| 152 |
if operator is None: |
| 153 |
# This is the first token: simply find all matches
|
| 154 |
for context in current_context: |
| 155 |
context_matches = [el for el in context[0].findAll(tag, find_dict) if checker(el)] |
| 156 |
for context_match in context_matches: |
| 157 |
found.append( |
| 158 |
(context_match, [context_match]), |
| 159 |
) |
| 160 |
elif operator == ' ': |
| 161 |
# for each context in current_context, ensure there
|
| 162 |
# exists an element somewhere above that element that
|
| 163 |
# matches the provided token
|
| 164 |
# ("descendant" selector)
|
| 165 |
for context in current_context: |
| 166 |
context_matches = [] |
| 167 |
for el in context[1]: |
| 168 |
if checker(el.findParent(tag, find_dict)):
|
| 169 |
context_matches.append(el) |
| 170 |
if context_matches:
|
| 171 |
found.append( |
| 172 |
(context[0], context_matches),
|
| 173 |
) |
| 174 |
elif operator == '>': |
| 175 |
# for each context in current_context,
|
| 176 |
# check if the parent satisfies the provided
|
| 177 |
# arguments.
|
| 178 |
for context in current_context: |
| 179 |
context_matches = [] |
| 180 |
for el in context[1]: |
| 181 |
if checker(el.findParent(tag, find_dict)) == el.parent:
|
| 182 |
context_matches.append(el.parent) |
| 183 |
if context_matches:
|
| 184 |
found.append( |
| 185 |
(context[0], context_matches),
|
| 186 |
) |
| 187 |
elif operator == '~': |
| 188 |
# for each context in current_context
|
| 189 |
# check
|
| 190 |
raise NotImplementedError("~ operator is not implemented. Sad face :(") |
| 191 |
elif operator == '+': |
| 192 |
# for each context in current_context
|
| 193 |
# check if the preceding sibling satisfies the
|
| 194 |
# provided arguments
|
| 195 |
for context in current_context: |
| 196 |
context_matches = [] |
| 197 |
for el in context[1]: |
| 198 |
if checker(el.findPreviousSibling(tag, find_dict)) == el.previousSibling:
|
| 199 |
context_matches.append(el.previousSibling) |
| 200 |
if context_matches:
|
| 201 |
found.append( |
| 202 |
(context[0], context_matches)
|
| 203 |
) |
| 204 |
current_context = found |
| 205 |
else:
|
| 206 |
# Get the next operator (whitespace, >, ~, +)
|
| 207 |
handle_token = True
|
| 208 |
operator = None
|
| 209 |
match = re.search('([>~+]+)$', selector)
|
| 210 |
if match:
|
| 211 |
operator = match.groups(1)[0] |
| 212 |
else:
|
| 213 |
operator = ' '
|
| 214 |
selector = selector.rsplit(operator, 1)[0].rstrip() |
| 215 |
return [entry[0] for entry in current_context] |
| 216 |
|
| 217 |
def monkeypatch(BeautifulSoupClass=None): |
| 218 |
"""
|
| 219 |
If you don't explicitly state the class to patch, defaults to the most
|
| 220 |
common import location for BeautifulSoup.
|
| 221 |
"""
|
| 222 |
if not BeautifulSoupClass: |
| 223 |
from BeautifulSoup import BeautifulSoup as BeautifulSoupClass |
| 224 |
BeautifulSoupClass.findSelect = select |
| 225 |
|
| 226 |
def unmonkeypatch(BeautifulSoupClass=None): |
| 227 |
if not BeautifulSoupClass: |
| 228 |
from BeautifulSoup import BeautifulSoup as BeautifulSoupClass |
| 229 |
delattr(BeautifulSoupClass, 'findSelect') |