warn-scraper 1.2.56__py3-none-any.whl → 1.2.57__py3-none-any.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.
Files changed (7) hide show
  1. warn/scrapers/hi.py +9 -6
  2. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/METADATA +1 -1
  3. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/RECORD +7 -7
  4. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/LICENSE +0 -0
  5. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/WHEEL +0 -0
  6. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/entry_points.txt +0 -0
  7. {warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/top_level.txt +0 -0
warn/scrapers/hi.py CHANGED
@@ -1,13 +1,14 @@
1
1
  import datetime
2
2
  import logging
3
3
  from pathlib import Path
4
+ from urllib.parse import quote
4
5
 
5
6
  from bs4 import BeautifulSoup
6
7
 
7
8
  from .. import utils
8
9
 
9
10
  __authors__ = ["Ash1R", "stucka"]
10
- __tags__ = ["html"]
11
+ __tags__ = ["html", "pdf"]
11
12
  __source__ = {
12
13
  "name": "Workforce Development Hawaii",
13
14
  "url": "https://labor.hawaii.gov/wdc/real-time-warn-updates/",
@@ -28,15 +29,17 @@ def scrape(
28
29
  cache_dir -- the Path where results can be cached (default WARN_CACHE_DIR)
29
30
  Returns: the Path where the file is written
30
31
  """
31
- firstpage = utils.get_url("https://labor.hawaii.gov/wdc/real-time-warn-updates/")
32
+ cacheprefix = "https://webcache.googleusercontent.com/search?q=cache%3A" # Use Google Cache, per #600
33
+
34
+ firstpage = utils.get_url(cacheprefix + quote("https://labor.hawaii.gov/wdc/real-time-warn-updates/"))
32
35
  soup = BeautifulSoup(firstpage.text, features="html5lib")
33
36
  pagesection = soup.select("div.primary-content")[0]
34
37
  subpageurls = []
35
38
  for atag in pagesection.find_all("a"):
36
39
  href = atag["href"]
37
40
  if href.endswith("/"):
38
- href = href[:-1]
39
- subpageurls.append(href)
41
+ href = href # [:-1]
42
+ subpageurls.append(cacheprefix + quote(href))
40
43
 
41
44
  headers = ["Company", "Date", "PDF url", "location", "jobs"]
42
45
  data = [headers]
@@ -85,8 +88,8 @@ def scrape(
85
88
  row.append(dates[i])
86
89
 
87
90
  row.append(url)
88
- row.append(None) # location
89
- row.append(None) # jobs
91
+ row.append(None) # location
92
+ row.append(None) # jobs
90
93
  data.append(row)
91
94
 
92
95
  output_csv = data_dir / "hi.csv"
{warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: warn-scraper
3
- Version: 1.2.56
3
+ Version: 1.2.57
4
4
  Summary: Command-line interface for downloading WARN Act notices of qualified plant closings and mass layoffs from state government websites
5
5
  Home-page: https://github.com/biglocalnews/warn-scraper
6
6
  Author: Big Local News
{warn_scraper-1.2.56.dist-info → warn_scraper-1.2.57.dist-info}/RECORD RENAMED
@@ -36,7 +36,7 @@ warn/scrapers/dc.py,sha256=kAWmERzEIOtGHla9tn8hK0NfP4B-aP4jknNGUiPw4C4,4493
36
36
  warn/scrapers/de.py,sha256=yMpCFAAlIO2f4DVUQpPKKLzm52_Zpn9IuPPFBNX1pjQ,1386
37
37
  warn/scrapers/fl.py,sha256=mHymxjwNGtYXRxAwjUSIG1qeSp4Y_zzr4XDxPz9LAfY,9560
38
38
  warn/scrapers/ga.py,sha256=EuqBrMlBojH6eXOHisNqJAQcsnb8FPHDwWomNopw9Ys,7285
39
- warn/scrapers/hi.py,sha256=ua32erw1Syc1SaAt28O0H_4p6WdUVVYwhg0EppyPWzI,3526
39
+ warn/scrapers/hi.py,sha256=IrwgUMNPqsHExiLZ8dFM25am7KTtVRrLDjIesNtJIsk,3736
40
40
  warn/scrapers/ia.py,sha256=zOncaA9M0d6paT4pB7UU_4D_yxUgeUiGRcnpKi9DsRA,1999
41
41
  warn/scrapers/id.py,sha256=rRkE9g9ZOL7JgTdIz46kyGOeetzSla3e1Xr6gJ1v_74,5443
42
42
  warn/scrapers/il.py,sha256=sygdvsNuB_Gvu3o_HidtpSP4FLz0szKb1zEHqGxVtlI,1563
@@ -66,9 +66,9 @@ warn/scrapers/va.py,sha256=DDuR4_2Jpaxg9nVmuM6PAR8v8xz3VgxTBG5sWJgz2q0,1582
66
66
  warn/scrapers/vt.py,sha256=d-bo4WK2hkrk4BhCCmLpEovcoZltlvdIUB6O0uaMx5A,1186
67
67
  warn/scrapers/wa.py,sha256=Il3RmJpKr7SbwUBxHxlhEFLoxy7zSiduyo8F2EddB2Y,4021
68
68
  warn/scrapers/wi.py,sha256=ClEzXkwZbop0W4fkQgsb5oHAPUrb4luUPGV-jOKwkcg,4855
69
- warn_scraper-1.2.56.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
70
- warn_scraper-1.2.56.dist-info/METADATA,sha256=vf8OqPrde5xVN13Aam1CyVG7svsDyZzfx_jr-Alvq3A,2025
71
- warn_scraper-1.2.56.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
72
- warn_scraper-1.2.56.dist-info/entry_points.txt,sha256=poh_oSweObGlBSs1_2qZmnTodlOYD0KfO7-h7W2UQIw,47
73
- warn_scraper-1.2.56.dist-info/top_level.txt,sha256=gOhHgNEkrUvajlzoKkVOo-TlQht9MoXnKOErjzqLGHo,11
74
- warn_scraper-1.2.56.dist-info/RECORD,,
69
+ warn_scraper-1.2.57.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
70
+ warn_scraper-1.2.57.dist-info/METADATA,sha256=Z__kggmqalQaYaWaynUtNFXqNiBW9X8u7SDh0Y3qjkI,2025
71
+ warn_scraper-1.2.57.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
72
+ warn_scraper-1.2.57.dist-info/entry_points.txt,sha256=poh_oSweObGlBSs1_2qZmnTodlOYD0KfO7-h7W2UQIw,47
73
+ warn_scraper-1.2.57.dist-info/top_level.txt,sha256=gOhHgNEkrUvajlzoKkVOo-TlQht9MoXnKOErjzqLGHo,11
74
+ warn_scraper-1.2.57.dist-info/RECORD,,