231 lines
7.2 KiB
Python
231 lines
7.2 KiB
Python
import urllib.request
|
||
import os
|
||
import re
|
||
import json
|
||
import jinja2
|
||
|
||
|
||
STATIC_FOLDER_PATH = './static' # without trailing slash
|
||
PUBLIC_STATIC_FOLDER_PATH = '/static' # without trailing slash
|
||
WRAPPING_TEMPLATES_DIR = './wrapping-templates'
|
||
|
||
# This uses a low quality copy of all the images
|
||
# (using a folder with the name "images-small",
|
||
# which stores a copy of all the images generated with:
|
||
# $ mogrify -quality 5% -adaptive-resize 25% -remap pattern:gray50 * )
|
||
fast = False
|
||
|
||
def API_request(url, pagename):
|
||
"""
|
||
url = API request url (string)
|
||
data = { 'query':
|
||
'pages' :
|
||
pageid : {
|
||
'links' : {
|
||
'?' : '?'
|
||
'title' : 'pagename'
|
||
}
|
||
}
|
||
}
|
||
}
|
||
"""
|
||
response = urllib.request.urlopen(url).read()
|
||
data = json.loads(response)
|
||
|
||
# Save response as JSON to be able to inspect API call
|
||
json_file = f'{ STATIC_FOLDER_PATH }/{ pagename }.json'
|
||
print('Saving JSON:', json_file)
|
||
with open(json_file, 'w') as out:
|
||
out.write(json.dumps(data, indent=4))
|
||
out.close()
|
||
|
||
return data
|
||
|
||
def download_media(html, images, wiki):
|
||
"""
|
||
html = string (HTML)
|
||
images = list of filenames (str)
|
||
"""
|
||
# check if 'images/' already exists
|
||
if not os.path.exists(f'{ STATIC_FOLDER_PATH }/images'):
|
||
os.makedirs(f'{ STATIC_FOLDER_PATH }/images')
|
||
|
||
# download media files
|
||
for filename in images:
|
||
filename = filename.replace(' ', '_') # safe filenames
|
||
|
||
# check if the image is already downloaded
|
||
# if not, then download the file
|
||
if not os.path.isfile(f'{ STATIC_FOLDER_PATH }/images/{ filename }'):
|
||
|
||
# first we search for the full filename of the image
|
||
url = f'{ wiki }/api.php?action=query&list=allimages&aifrom={ filename }&format=json'
|
||
response = urllib.request.urlopen(url).read()
|
||
data = json.loads(response)
|
||
|
||
# we select the first search result
|
||
# (assuming that this is the image we are looking for)
|
||
image = data['query']['allimages'][0]
|
||
|
||
# then we download the image
|
||
image_url = image['url']
|
||
image_filename = image['name']
|
||
print('Downloading:', image_filename)
|
||
image_response = urllib.request.urlopen(image_url).read()
|
||
|
||
# and we save it as a file
|
||
image_path = f'{ STATIC_FOLDER_PATH }/images/{ image_filename }'
|
||
out = open(image_path, 'wb')
|
||
out.write(image_response)
|
||
out.close()
|
||
|
||
import time
|
||
time.sleep(3) # do not overload the server
|
||
|
||
# replace src link
|
||
image_path = f'{ PUBLIC_STATIC_FOLDER_PATH }/images/{ filename }' # here the images need to link to the / of the domain, for flask :/// confusing! this breaks the whole idea to still be able to make a local copy of the file
|
||
matches = re.findall(rf'src="/book/images/.*?px-{ filename }"', html) # for debugging
|
||
if matches:
|
||
html = re.sub(rf'src="/book/images/.*?px-{ filename }"', f'src="{ image_path }"', html)
|
||
else:
|
||
matches = re.findall(rf'src="/book/images/.*?{ filename }"', html) # for debugging
|
||
html = re.sub(rf'src="/book/images/.*?{ filename }"', f'src="{ image_path }"', html)
|
||
# print(f'{filename}: {matches}\n------') # for debugging: each image should have the correct match!
|
||
|
||
return html
|
||
|
||
def add_item_inventory_links(html):
|
||
"""
|
||
html = string (HTML)
|
||
"""
|
||
# Find all references in the text to the item index
|
||
pattern = r'Item \d\d\d'
|
||
matches = re.findall(pattern, html)
|
||
index = {}
|
||
new_html = ''
|
||
from nltk.tokenize import sent_tokenize
|
||
for line in sent_tokenize(html):
|
||
for match in matches:
|
||
if match in line:
|
||
number = match.replace('Item ', '').strip()
|
||
if not number in index:
|
||
index[number] = []
|
||
count = 1
|
||
else:
|
||
count = index[number][-1] + 1
|
||
index[number].append(count)
|
||
item_id = f'ii-{ number }-{ index[number][-1] }'
|
||
line = line.replace(match, f'Item <a id="{ item_id }" href="#Item_Index">{ number }</a>')
|
||
|
||
# the line is pushed back to the new_html
|
||
new_html += line + ' '
|
||
|
||
# Also add a <span> around the index nr to style it
|
||
matches = re.findall(r'<li>\d\d\d', new_html)
|
||
for match in matches:
|
||
new_html = new_html.replace(match, f'<li><span class="item_nr">{ match }</span>')
|
||
|
||
# import json
|
||
# print(json.dumps(index, indent=4))
|
||
|
||
return new_html
|
||
|
||
def tweaking(html):
|
||
"""
|
||
html = string (HTML)
|
||
"""
|
||
html = html.replace('<a href="#X,_y,_z_(4_filmstills)"', '<a href="#x,_y,_z_(4_filmstills)"') # change the anchor link in the TOC to lowercase
|
||
html = html.replace('<a href="#Rehearsal_as_the_%E2%80%98Other%E2%80%99_to_Hypercomputation"', '<a href="#Rehearsal_as_the_‘Other’_to_Hypercomputation"') # change the anchor link in the TOC to lowercase
|
||
html = html.replace('<a href="#We_hardly_encounter_anything_that_didn%E2%80%99t_really_matter"', '<a href="#We_hardly_encounter_anything_that_didn’t_really_matter"') # change the anchor link in the TOC to lowercase
|
||
return html
|
||
|
||
def clean_up(html):
|
||
"""
|
||
html = string (HTML)
|
||
"""
|
||
html = re.sub(r'\[.*edit.*\]', '', html) # remove the [edit]
|
||
html = re.sub(r'href="/book/index.php\?title=', 'href="#', html) # remove the internal wiki links
|
||
html = re.sub(r'[(?=\d)', '', html) # remove left footnote bracket [
|
||
html = re.sub(r'(?<=\d)]', '', html) # remove right footnote bracket ]
|
||
return html
|
||
|
||
def fast_loader(html):
|
||
"""
|
||
html = string (HTML)
|
||
"""
|
||
if fast == True:
|
||
html = html.replace('/images/', '/images-small/')
|
||
print('--- rendered in FAST mode ---')
|
||
|
||
return html
|
||
|
||
def parse_page(pagename, wiki):
|
||
"""
|
||
pagename = string
|
||
html = string (HTML)
|
||
"""
|
||
parse = f'{ wiki }/api.php?action=parse&page={ pagename }&pst=True&format=json'
|
||
data = API_request(parse, pagename)
|
||
# print(json.dumps(data, indent=4))
|
||
if 'parse' in data:
|
||
html = data['parse']['text']['*']
|
||
images = data['parse']['images']
|
||
html = download_media(html, images, wiki)
|
||
html = clean_up(html)
|
||
html = add_item_inventory_links(html)
|
||
html = tweaking(html)
|
||
html = fast_loader(html)
|
||
else:
|
||
html = None
|
||
|
||
return html
|
||
|
||
def save(html, pagename, publication_unfolded):
|
||
"""
|
||
html = string (HTML)
|
||
pagename = string
|
||
"""
|
||
if html:
|
||
|
||
# save final page that will be used with PagedJS
|
||
template_file = open(f'{ STATIC_FOLDER_PATH }/{ WRAPPING_TEMPLATES_DIR }/template.html').read()
|
||
template = jinja2.Template(template_file)
|
||
html = template.render(publication_unfolded=publication_unfolded, title=pagename)
|
||
|
||
html_file = f'{ STATIC_FOLDER_PATH }/{ pagename }.html'
|
||
print('Saving HTML:', html_file)
|
||
with open(html_file, 'w') as out:
|
||
out.write(html)
|
||
out.close()
|
||
|
||
# save extra html page for debugging
|
||
template_file = open(f'{ STATIC_FOLDER_PATH }/{ WRAPPING_TEMPLATES_DIR }/template.inspect.html').read()
|
||
template = jinja2.Template(template_file)
|
||
html = template.render(publication_unfolded=publication_unfolded, title=pagename)
|
||
|
||
html_file = f'{ STATIC_FOLDER_PATH }/{ pagename }.inspect.html'
|
||
print('Saving HTML:', html_file)
|
||
with open(html_file, 'w') as out:
|
||
out.write(html)
|
||
out.close()
|
||
|
||
def update_material_now(pagename, wiki):
|
||
"""
|
||
pagename = string
|
||
publication_unfolded = string (HTML)
|
||
"""
|
||
publication_unfolded = parse_page(pagename, wiki)
|
||
|
||
return publication_unfolded
|
||
|
||
# ---
|
||
|
||
if __name__ == "__main__":
|
||
|
||
wiki = 'https://possiblebodies.constantvzw.org/book' # remove tail slash '/'
|
||
pagename = 'Unfolded'
|
||
|
||
publication_unfolded = update_material_now(pagename, wiki) # download the latest version of the page
|
||
save(publication_unfolded, pagename, publication_unfolded) # save the page to file
|
||
|