Homec4science
Diffusion Slides (master) Browse

R3683/Slides/scripts/organize_image.py
18ad8d1822eamaster

Last Change
Raw File

organize_image.py
View Options

#!/usr/bin/python
"""
This executable permits to organize the image
links in a Jupyter notebook.
"""
################################################################
import argparse
import Slides.cell_manager as cell_manager
import os
import pyparsing as pp
################################################################
list_images = set()
def populate_list_images(image_dir):
if not os.path.isdir(image_dir):
os.mkdir(image_dir)
for i in os.listdir(image_dir):
i = os.path.join(image_dir, i)
i = os.path.abspath(i)
i = os.path.relpath(i, os.getcwd())
list_images.add(i)
################################################################
# parsing rules for attachments
def _print(label):
return lambda x: print(label + ':', x)
valid_characters = pp.printables
for c in "[]!()":
valid_characters = valid_characters.replace(c, '')
markdown_image_link = (pp.Literal("![") +
pp.Word(valid_characters) +
pp.Literal("]") +
pp.Literal("(") +
pp.Word(valid_characters) +
pp.Literal(")"))
markdown_image_link.addParseAction(
lambda x: {'alt': x[1].strip(), 'src': x[4].strip(),
'link': ''.join(x)})
valid_characters = pp.printables
for c in "<>\"='":
valid_characters = valid_characters.replace(c, '')
valid_characters += ' '
def html_image_option(name):
opt = pp.Literal(f'{name}="') + pp.Word(valid_characters) + pp.Literal('"')
opt.addParseAction(
lambda _, b, x: {name: x[1].strip()})
opt2 = pp.Literal(f'{name}=\'') + \
pp.Word(valid_characters) + pp.Literal('\'')
opt2.addParseAction(
lambda _, b, x: {name: x[1].strip()})
return opt | opt2
html_image_body = pp.OneOrMore(html_image_option('alt') |
html_image_option('src') |
html_image_option('width'))
html_image_link = (pp.Literal("<img") +
html_image_body + pp.Literal(">").addParseAction(
lambda _, b, x: {'end': b}))
def gather_parsed(original, _l, tokens):
res = {}
end = tokens[-1]['end']+1
res['link'] = original[_l:end]
for e in tokens[1:-1]:
res.update(e)
return res
html_image_link.addParseAction(gather_parsed)
image_link = html_image_link | markdown_image_link
images = pp.ZeroOrMore(
pp.SkipTo(image_link).suppress() + image_link)
################################################################
def create_image_file(fname, data, image_dir=None):
try:
os.mkdir(image_dir)
except FileExistsError:
pass
base, ext = os.path.splitext(fname)
fname = os.path.join(base.replace(' ', '_') + ext)
fname = os.path.join(image_dir, fname)
while os.path.exists(fname):
base, ext = os.path.splitext(fname)
fname = os.path.join(base+'_' + ext)
with open(fname, 'wb') as f:
import base64
data = base64.b64decode(data)
f.write(data)
return fname
################################################################
def extract_attachement(cell, **kwargs):
cited_attachments = {}
links = images.parseString(cell.source)
for _l in links:
fname = _l['src']
if not fname.startswith('attachment:'):
continue
fname = fname.replace('attachment:', '')
cited_attachments[fname] = _l
for name, a in cell.attachments.items():
for _type, _file in a.items():
if name in cited_attachments:
print(f'found_matching_attachment: {name}: {_type}')
fname = create_image_file(name, _file, **kwargs)
att = cited_attachments[name]
print(att)
cell.source = cell.source.replace(
att['link'],
f'<img alt="{att["alt"]}" src="./{fname}">')
################################################################
def check_images(cell, **kwargs):
links = images.parseString(cell.source)
if links:
print(cell.source)
print([e['src'] for e in links])
for _l in links:
src = _l['src']
if not os.path.exists(src):
raise RuntimeError(f"not existing image: {src}")
################################################################
cited_images = []
def list_images_links(cell, **kwargs):
print(cell.source)
links = images.parseString(cell.source)
global cited_images
print(links)
cited_images += [e['src'] for e in links]
################################################################
def main():
parser = argparse.ArgumentParser(description='Jupyter image organizer')
parser.add_argument("--image_dir", type=str, default='images',
help='Specify the directory where to store images')
parser.add_argument("notebooks", nargs='*',
help="The notebooks to associate in a presentation")
parser.add_argument("-o", "--out_filename", type=str, default="talk.ipynb",
help="name of notebook to be generated")
parser.add_argument("--extract_attachements", action='store_true',
help="demands to extract attachements images")
parser.add_argument("--check_images", action='store_true',
help="check if all images are present")
parser.add_argument("--clean_images", action='store_true',
help="delete unused images")
arguments = parser.parse_args()
populate_list_images(arguments.image_dir)
def analyze_images(i, cell):
# print(f'{i}: {list(cell.keys())}')
if 'attachments' in cell and arguments.extract_attachements:
extract_attachement(cell, image_dir=arguments.image_dir)
if arguments.check_images:
check_images(cell, image_dir=arguments.image_dir)
if arguments.clean_images:
list_images_links(cell, image_dir=arguments.image_dir)
return cell
try:
cell_manager.mergeNotebooks(arguments.notebooks,
modify_functor=analyze_images,
out_filename=arguments.out_filename)
except cell_manager.NotFoundNotebooks:
print('Could not find the notebooks to treat: check path provided')
return
global cited_images
cited_images = [os.path.abspath(e) for e in cited_images]
cited_images = [os.path.relpath(e, os.getcwd()) for e in cited_images]
print(cited_images)
if arguments.clean_images:
print(f'found {len(list_images)} image files')
print(f'found {len(cited_images)} cited images')
for fname in list_images:
if fname not in cited_images:
print(f"should_delete_file: {fname}")
################################################################
if __name__ == "__main__":
main()
c4science · Help