Browse Source

Amélioration de la génération de PDF

- Mise à jour automatique du nombre de pages dans le CSS
- Intégration des bookmarks dans le PDF final
prod
Youen 1 year ago
parent
commit
ca4d9a1518
  1. 10
      Makefile
  2. 24
      sphinx-tools/count_pdf_pages.py
  3. 28
      sphinx-tools/update_pdf_bookmarks.py

10
Makefile

@ -20,13 +20,19 @@ help:
@$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O) @$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
# Default make weasyprint target has a bug so we have to call it manually on the generated index.html file # Default make weasyprint target has a bug so we have to call it manually on the generated index.html file
# Then we execute count_pdf_pages.py that will count the number of pages, and update print-theme.css
# Then we generate index.pdf again (with the correct number of pages in the footer)
# After that, extract table of content data from the pdf and generate a .tex file # After that, extract table of content data from the pdf and generate a .tex file
# Then generate a toc.pdf from the .tex file # Then generate a toc.pdf from the .tex file
# And put them together in a .pdf file # And put them together in a .pdf file
# Finally, we generate the final PDF by adding the bookmarks (for easier navigation in the PDF)
# After all, clean up # After all, clean up
pdf: weasyprint pdf: weasyprint
weasyprint build/weasyprint/index.html build/weasyprint/index.pdf -s source/css/print-theme.css
python3 sphinx-tools/count_pdf_pages.py
weasyprint build/weasyprint/index.html build/weasyprint/index.pdf -s source/css/print-theme.css weasyprint build/weasyprint/index.html build/weasyprint/index.pdf -s source/css/print-theme.css
./pdftoc-to-latex build/weasyprint/index.pdf > build/weasyprint/toc.tex ./pdftoc-to-latex build/weasyprint/index.pdf > build/weasyprint/toc.tex
- pdflatex -interaction nonstopmode -output-directory=build/weasyprint build/weasyprint/toc.tex - pdflatex -interaction nonstopmode -output-directory=build/weasyprint build/weasyprint/toc.tex
pdftk A=build/weasyprint/index.pdf B=build/weasyprint/toc.pdf cat A1 B A2-end output build/weasyprint/vheliotech.pdf pdftk A=build/weasyprint/index.pdf B=build/weasyprint/toc.pdf cat A1 B A2-end output build/weasyprint/vheliotech-without-bookmarks.pdf
rm build/weasyprint/index.pdf build/weasyprint/toc.tex build/weasyprint/toc.pdf build/weasyprint/GuidedemontageVheliotech.pdf python3 sphinx-tools/update_pdf_bookmarks.py build/weasyprint/index.pdf build/weasyprint/vheliotech-without-bookmarks.pdf build/weasyprint/vheliotech.pdf
#rm build/weasyprint/index.pdf build/weasyprint/toc.tex build/weasyprint/toc.pdf build/weasyprint/GuidedemontageVheliotech.pdf

24
sphinx-tools/count_pdf_pages.py

@ -0,0 +1,24 @@
import subprocess
import re
pdf_filename = 'build/weasyprint/index.pdf'
css_filename = 'source/css/print-theme.css'
additional_pages = 2
# count pages in index.pdf
pdfinfo = subprocess.run(['pdfinfo', pdf_filename], stdout=subprocess.PIPE)
pages_match = re.search('\\nPages:\s+([0-9]+)\\n', pdfinfo.stdout.decode())
num_pages = int(pages_match.group(1))
print('index.pdf: ' + str(num_pages) + ' pages')
num_pages = num_pages + additional_pages # account for table of content that will be added later
# update the CSS file with the correct number of pages
with open(css_filename) as css_file:
css = css_file.read()
css = re.sub('content: counter\(page\) "/[0-9]+";', 'content: counter(page) "/'+str(num_pages)+'";', css)
with open(css_filename, 'w') as css_file:
css_file.write(css)

28
sphinx-tools/update_pdf_bookmarks.py

@ -0,0 +1,28 @@
import sys
import subprocess
import re
extract_bookmarks_from = sys.argv[1]
source_pdf_filename = sys.argv[2]
output_filename = sys.argv[3]
bookmarks_filename = extract_bookmarks_from.replace('.pdf', '.txt')
assert(bookmarks_filename != extract_bookmarks_from)
# extract PDF metadata into a text file
subprocess.run(['pdftk', extract_bookmarks_from, 'dump_data', 'output', bookmarks_filename])
with open(bookmarks_filename) as bookmarks_file:
metadata = bookmarks_file.read()
# Offset page numbers
def replaceBookmarkPageNumber(match):
initial_page = int(match.group(1))
return 'BookmarkPageNumber: ' + str(initial_page + 2)
metadata = re.sub('BookmarkPageNumber:\s+([0-9]+)', replaceBookmarkPageNumber, metadata)
with open(bookmarks_filename, 'w') as bookmarks_file:
bookmarks_file.write(metadata)
# generate the output PDF
subprocess.run(['pdftk', source_pdf_filename, 'update_info', bookmarks_filename, 'output', output_filename])
Loading…
Cancel
Save