Changeset 35127
- Timestamp:
- 2021-04-27T14:11:45+12:00 (3 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
main/trunk/model-sites-dev/eurovision-lod/collect/eurovision/prepare/errata-categories/escwikipedia.py
r35117 r35127 302 302 ## debug_output_country_year_recs(country_year_recs) 303 303 304 # Splice in logo image for that year into each country_year_rec 305 #infobox_table = esc_year_soup.find("table",{"class": "infobox"}) 306 #infobox_td = infobox_table.find("td",{"class": "infobox-image"}) 307 #infobox_img = infobox_td.find("img") 308 309 #infobox_img = esc_year_soup.select("table.infobox td.infobox-image img") 310 311 #infobox_img = esc_year_soup.select("table.infobox td.infobox-image img") 312 #infobox_img = esc_year_soup.select("table.infobox td.infobox-image") 313 # infobox_img = esc_year_soup.select("table.infobox") 314 315 #infobox_table = esc_year_soup.find("table",{"class": "infobox"}) 316 #infobox_img = infobox_table.tbody.tr[0].td[0].a.img 317 318 # Looks like Wikipedia has changed some of its infobox CSS 319 # If processing newer downloads, then the following is probably the select 320 # statement to use 321 322 # new school! 323 #infobox_img = esc_year_soup.select("table.infobox td.infobox-image img") 324 325 # old school 326 infobox_imgs = esc_year_soup.select("table.infobox tr td a.image img") 327 328 if (len(infobox_imgs) == 0): 329 print("****") 330 print("****!!! No ESC Logo image found!!!") 331 print("****") 332 else: 333 # Some pages include addition image graphics, such as a map showing country entrants 334 # => Want the first one 335 infobox_logo_img = infobox_imgs[0]; 336 337 for country_year_key in country_year_recs.keys(): 338 country_year_rec = country_year_recs.get(country_year_key) 339 340 country_year_rec["YearLogoImg"] = str(infobox_logo_img) 341 342 304 343 return country_year_recs 305 344
Note:
See TracChangeset
for help on using the changeset viewer.