selectolax 0.3.29__cp310-cp310-win_amd64.whl → 0.3.31__cp310-cp310-win_amd64.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of selectolax might be problematic. Click here for more details.
- selectolax/__init__.py +1 -1
- selectolax/lexbor/attrs.pxi +27 -9
- selectolax/lexbor/node.pxi +9 -6
- selectolax/lexbor/selection.pxi +7 -0
- selectolax/lexbor.c +53277 -55311
- selectolax/lexbor.cp310-win_amd64.pyd +0 -0
- selectolax/lexbor.pxd +5 -6
- selectolax/lexbor.pyi +648 -61
- selectolax/lexbor.pyx +6 -0
- selectolax/modest/node.pxi +8 -6
- selectolax/parser.c +50957 -52325
- selectolax/parser.cp310-win_amd64.pyd +0 -0
- selectolax/parser.pyi +487 -43
- selectolax/parser.pyx +15 -4
- {selectolax-0.3.29.dist-info → selectolax-0.3.31.dist-info}/METADATA +10 -17
- selectolax-0.3.31.dist-info/RECORD +26 -0
- {selectolax-0.3.29.dist-info → selectolax-0.3.31.dist-info}/WHEEL +1 -1
- selectolax-0.3.29.dist-info/RECORD +0 -26
- {selectolax-0.3.29.dist-info → selectolax-0.3.31.dist-info}/licenses/LICENSE +0 -0
- {selectolax-0.3.29.dist-info → selectolax-0.3.31.dist-info}/top_level.txt +0 -0
selectolax/parser.pyx
CHANGED
|
@@ -124,7 +124,7 @@ cdef class HTMLParser:
|
|
|
124
124
|
status = myhtml_parse(self.html_tree, self._encoding, html, html_len)
|
|
125
125
|
|
|
126
126
|
if status != 0:
|
|
127
|
-
raise RuntimeError("Can't parse HTML
|
|
127
|
+
raise RuntimeError("Can't parse HTML (status code: %d)" % status)
|
|
128
128
|
|
|
129
129
|
assert self.html_tree.node_html != NULL
|
|
130
130
|
|
|
@@ -147,9 +147,13 @@ cdef class HTMLParser:
|
|
|
147
147
|
def root(self):
|
|
148
148
|
"""Returns root node."""
|
|
149
149
|
if self.html_tree and self.html_tree.node_html:
|
|
150
|
-
|
|
151
|
-
|
|
152
|
-
|
|
150
|
+
try:
|
|
151
|
+
node = Node()
|
|
152
|
+
node._init(self.html_tree.node_html, self)
|
|
153
|
+
return node
|
|
154
|
+
except Exception:
|
|
155
|
+
# If Node creation or initialization fails, return None
|
|
156
|
+
return None
|
|
153
157
|
return None
|
|
154
158
|
|
|
155
159
|
@property
|
|
@@ -185,6 +189,12 @@ cdef class HTMLParser:
|
|
|
185
189
|
name : str (e.g. div)
|
|
186
190
|
|
|
187
191
|
"""
|
|
192
|
+
# Validate tag name
|
|
193
|
+
if not name:
|
|
194
|
+
raise ValueError("Tag name cannot be empty")
|
|
195
|
+
if len(name) > 100: # Reasonable limit for tag names
|
|
196
|
+
raise ValueError("Tag name is too long")
|
|
197
|
+
|
|
188
198
|
cdef myhtml_collection_t* collection = NULL
|
|
189
199
|
pybyte_name = name.encode('UTF-8')
|
|
190
200
|
cdef mystatus_t status = 0;
|
|
@@ -428,6 +438,7 @@ cdef class HTMLParser:
|
|
|
428
438
|
if self.html_tree != NULL:
|
|
429
439
|
myhtml = self.html_tree.myhtml
|
|
430
440
|
myhtml_tree_destroy(self.html_tree)
|
|
441
|
+
self.html_tree = NULL # Prevent double-free
|
|
431
442
|
if myhtml != NULL:
|
|
432
443
|
myhtml_destroy(myhtml)
|
|
433
444
|
|
|
@@ -1,42 +1,35 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: selectolax
|
|
3
|
-
Version: 0.3.
|
|
3
|
+
Version: 0.3.31
|
|
4
4
|
Summary: Fast HTML5 parser with CSS selectors.
|
|
5
5
|
Home-page: https://github.com/rushter/selectolax
|
|
6
6
|
Author: Artem Golubin
|
|
7
|
-
Author-email: me@rushter.com
|
|
8
|
-
License: MIT
|
|
9
|
-
Project-URL:
|
|
10
|
-
|
|
7
|
+
Author-email: Artem Golubin <me@rushter.com>
|
|
8
|
+
License: MIT
|
|
9
|
+
Project-URL: Repository, https://github.com/rushter/selectolax
|
|
10
|
+
Project-URL: Documentation, https://selectolax.readthedocs.io/en/latest/parser.html
|
|
11
|
+
Project-URL: Changelog, https://github.com/rushter/selectolax/blob/main/CHANGES.rst
|
|
12
|
+
Keywords: selectolax,html,parser,css,fast
|
|
11
13
|
Classifier: Development Status :: 5 - Production/Stable
|
|
12
14
|
Classifier: Topic :: Text Processing :: Markup :: HTML
|
|
13
15
|
Classifier: Topic :: Internet
|
|
14
16
|
Classifier: Topic :: Internet :: WWW/HTTP
|
|
15
17
|
Classifier: Intended Audience :: Developers
|
|
16
|
-
Classifier: License :: OSI Approved :: MIT License
|
|
17
18
|
Classifier: Natural Language :: English
|
|
18
19
|
Classifier: Programming Language :: Python :: 3
|
|
19
|
-
Classifier: Programming Language :: Python :: 3.7
|
|
20
|
-
Classifier: Programming Language :: Python :: 3.8
|
|
21
20
|
Classifier: Programming Language :: Python :: 3.9
|
|
22
21
|
Classifier: Programming Language :: Python :: 3.10
|
|
23
22
|
Classifier: Programming Language :: Python :: 3.11
|
|
24
23
|
Classifier: Programming Language :: Python :: 3.12
|
|
25
24
|
Classifier: Programming Language :: Python :: 3.13
|
|
25
|
+
Requires-Python: >=3.9
|
|
26
|
+
Description-Content-Type: text/x-rst
|
|
26
27
|
License-File: LICENSE
|
|
27
28
|
Provides-Extra: cython
|
|
28
|
-
Requires-Dist: Cython
|
|
29
|
+
Requires-Dist: Cython; extra == "cython"
|
|
29
30
|
Dynamic: author
|
|
30
|
-
Dynamic: author-email
|
|
31
|
-
Dynamic: classifier
|
|
32
|
-
Dynamic: description
|
|
33
31
|
Dynamic: home-page
|
|
34
|
-
Dynamic: keywords
|
|
35
|
-
Dynamic: license
|
|
36
32
|
Dynamic: license-file
|
|
37
|
-
Dynamic: project-url
|
|
38
|
-
Dynamic: provides-extra
|
|
39
|
-
Dynamic: summary
|
|
40
33
|
|
|
41
34
|
.. image:: docs/logo.png
|
|
42
35
|
:alt: selectolax logo
|
|
@@ -0,0 +1,26 @@
|
|
|
1
|
+
selectolax/__init__.py,sha256=FqcMB0_lRBSC5GjhhahFsX1piTpOGEj2Mr7MkrzFh1U,185
|
|
2
|
+
selectolax/base.pxi,sha256=zOj3BrCA71xd-mJFtkMIAglP4ZybfrHVoCoy6ljTBDQ,93
|
|
3
|
+
selectolax/lexbor.c,sha256=rvkQp6ZbkbXMmMBAZXNY8U8UBD41YbE60kwHji94CSk,2419967
|
|
4
|
+
selectolax/lexbor.cp310-win_amd64.pyd,sha256=_5a2bngCqD__f2nKQRDWIThZptFPUH2Vj3kIQL9uTOM,3142656
|
|
5
|
+
selectolax/lexbor.pxd,sha256=cG264E-tFNOFTy0k5bAqV_sZnz4G6a4a21WEhqTL-NI,21516
|
|
6
|
+
selectolax/lexbor.pyi,sha256=bDXv_EKVBKLGyweo7vpxg-yxNKa_mRo_K2hK9LMsxIQ,25200
|
|
7
|
+
selectolax/lexbor.pyx,sha256=-O-g03mLCQKc9F19eMvo3PyoLDtF09IIuFziXJAl6Ao,11520
|
|
8
|
+
selectolax/parser.c,sha256=xGQmCxrEmyQCi9I11d3py7cIcFumqhhVpdystsQ1dp0,2287084
|
|
9
|
+
selectolax/parser.cp310-win_amd64.pyd,sha256=dBbAr5rvyRB_cfR06bkNn0I9OBfBtSIF6CMP-Sl68oo,2108416
|
|
10
|
+
selectolax/parser.pxd,sha256=4pM_CcZlvJlaR8EMjZCnSmnCcJbwcYOldRTBEbfwm48,25145
|
|
11
|
+
selectolax/parser.pyi,sha256=5Czf63278MQC01IxY-CHzoDyTS1oHiYRD2OxEscyL1o,25584
|
|
12
|
+
selectolax/parser.pyx,sha256=bS2n70o_5OPJ6JuXTBAVUTc-XhxqC4DXzPE4H3-e5Ek,13987
|
|
13
|
+
selectolax/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
|
14
|
+
selectolax/utils.pxi,sha256=4rtdRcLWuemxN1qe7Eul5jvAmHZ65r7Gvf67_Wg8Bt4,3566
|
|
15
|
+
selectolax/lexbor/attrs.pxi,sha256=KvQaCo0jM3Bva6_xG3TGzkCVFjNQ4kgTxWO95gsGbkw,4007
|
|
16
|
+
selectolax/lexbor/node.pxi,sha256=BuJIYcTUucvHUP4w2wl90uBwjcqpv1vRyt4No2O_Quo,30921
|
|
17
|
+
selectolax/lexbor/selection.pxi,sha256=4I8cjYMjC2Gz7xLrYWrl7jbnwDoVmab-GvrJKiTALTI,7017
|
|
18
|
+
selectolax/lexbor/util.pxi,sha256=0I4ElWIwXxrZCfMmGCtyDU127oMsPCqC3IcUk4QmMAc,582
|
|
19
|
+
selectolax/modest/node.pxi,sha256=Da2b3cdmggCX736x0htGvac51SEeGCcY5l-LA5H4HNI,34376
|
|
20
|
+
selectolax/modest/selection.pxi,sha256=0elY7JwnpPVaw0QZE1T7A78s9FIph5uWIhwy4sEXGU8,6586
|
|
21
|
+
selectolax/modest/util.pxi,sha256=o2nPGGGtRlLqOCa7yPk94CfBzNlVr7ull7osFy6NRX4,570
|
|
22
|
+
selectolax-0.3.31.dist-info/licenses/LICENSE,sha256=A7Jb3WZcENcLfZRc7QPdm9zJdwfpIyPodPJu-kdMH6E,1087
|
|
23
|
+
selectolax-0.3.31.dist-info/METADATA,sha256=4WhsMNLtWvjn-th6KxIHAy7iEpBPvB-a3LFyqLxeJjk,6401
|
|
24
|
+
selectolax-0.3.31.dist-info/WHEEL,sha256=KUuBC6lxAbHCKilKua8R9W_TM71_-9Sg5uEP3uDWcoU,101
|
|
25
|
+
selectolax-0.3.31.dist-info/top_level.txt,sha256=e5MuEM2PrQzoDlWetkFli9uXSlxa_ktW5jJEihhaI1c,11
|
|
26
|
+
selectolax-0.3.31.dist-info/RECORD,,
|
|
@@ -1,26 +0,0 @@
|
|
|
1
|
-
selectolax/__init__.py,sha256=EJMTWQiycBDUJq3c6zs6gPVxNbcNCHfynjXd42bZWeA,185
|
|
2
|
-
selectolax/base.pxi,sha256=zOj3BrCA71xd-mJFtkMIAglP4ZybfrHVoCoy6ljTBDQ,93
|
|
3
|
-
selectolax/lexbor.c,sha256=QBMO8PscLH9WEB4Lr9NL7gxmT4SPlq2pyOG4B4cK4ck,2368851
|
|
4
|
-
selectolax/lexbor.cp310-win_amd64.pyd,sha256=W_dKQCUbOX5qvsxLWTrdDxjiFHUDVH_hWcKl3kp122Y,3200512
|
|
5
|
-
selectolax/lexbor.pxd,sha256=1d9nvZd9rZl27gwPwVV5BlbR2LAi6jDK69Xm9Guz5Kk,21538
|
|
6
|
-
selectolax/lexbor.pyi,sha256=IeGh8NEcvybAHWxgh_-rHV9oxKjkdeWb9QUtgjBH12M,6782
|
|
7
|
-
selectolax/lexbor.pyx,sha256=6nvezEKHO5ffieso6jr0vYss8V0WkVup9NocJoVepuo,11353
|
|
8
|
-
selectolax/parser.c,sha256=9FirqHRzfMS3U67DKJk0YJ5x7HdEqOpmFAQnJskv_MM,2224906
|
|
9
|
-
selectolax/parser.cp310-win_amd64.pyd,sha256=Cd1bS6Uqb9gYg6_VYRs8AB2TsKjwzIcLWqS6CGiM5c8,2143744
|
|
10
|
-
selectolax/parser.pxd,sha256=4pM_CcZlvJlaR8EMjZCnSmnCcJbwcYOldRTBEbfwm48,25145
|
|
11
|
-
selectolax/parser.pyi,sha256=jFQa0_av9w9HT4XtbJioI86Le0N_Wvbc7BFbdPqDuHM,11598
|
|
12
|
-
selectolax/parser.pyx,sha256=CeZDNHF6G1RhUv2TfihZmCVMkGUwMhy6zrsffB1p7hE,13528
|
|
13
|
-
selectolax/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
|
14
|
-
selectolax/utils.pxi,sha256=4rtdRcLWuemxN1qe7Eul5jvAmHZ65r7Gvf67_Wg8Bt4,3566
|
|
15
|
-
selectolax/lexbor/attrs.pxi,sha256=r9DroDAkoxIvSMiDTRKpfYp503b7yUteDoYwglhQ0FM,3241
|
|
16
|
-
selectolax/lexbor/node.pxi,sha256=Kr5Oi3eQdxplTt8ltG1eWjFfHFrdcgsm5YYJnKFJDK8,30763
|
|
17
|
-
selectolax/lexbor/selection.pxi,sha256=DVtVnaCwzXPPkDqgW0vEtkQa1zWJ0c2Ud3KGSrvK5PM,6755
|
|
18
|
-
selectolax/lexbor/util.pxi,sha256=0I4ElWIwXxrZCfMmGCtyDU127oMsPCqC3IcUk4QmMAc,582
|
|
19
|
-
selectolax/modest/node.pxi,sha256=FP2_09dUKxK-z-j_A9hw98WwvBZFGvtifA2YYKbKpbU,34300
|
|
20
|
-
selectolax/modest/selection.pxi,sha256=0elY7JwnpPVaw0QZE1T7A78s9FIph5uWIhwy4sEXGU8,6586
|
|
21
|
-
selectolax/modest/util.pxi,sha256=o2nPGGGtRlLqOCa7yPk94CfBzNlVr7ull7osFy6NRX4,570
|
|
22
|
-
selectolax-0.3.29.dist-info/licenses/LICENSE,sha256=A7Jb3WZcENcLfZRc7QPdm9zJdwfpIyPodPJu-kdMH6E,1087
|
|
23
|
-
selectolax-0.3.29.dist-info/METADATA,sha256=wwLeEm4lfcwdZ58hApe0sxokMPTPsk4rnO16d2tPZBU,6471
|
|
24
|
-
selectolax-0.3.29.dist-info/WHEEL,sha256=KizrmDa55g-UYHe6LhGBVxVsOiLHH77kPP1Ia8vmad4,101
|
|
25
|
-
selectolax-0.3.29.dist-info/top_level.txt,sha256=e5MuEM2PrQzoDlWetkFli9uXSlxa_ktW5jJEihhaI1c,11
|
|
26
|
-
selectolax-0.3.29.dist-info/RECORD,,
|
|
File without changes
|
|
File without changes
|