When adding markdown (.md) or textile (.textile) files that contain references to images, automatically add them as txtz with the images

This commit is contained in:
Kovid Goyal 2021-08-31 20:56:44 +05:30
parent 1a54a69098
commit 282477bcfa
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
3 changed files with 30 additions and 21 deletions

View File

@ -54,21 +54,19 @@ class TXT2TXTZ(FileTypePlugin):
'containing Markdown or Textile references to images. The referenced '
'images as well as the TXT file are added to the archive.')
version = numeric_version
file_types = {'txt', 'text'}
file_types = {'txt', 'text', 'md', 'markdown', 'textile'}
supported_platforms = ['windows', 'osx', 'linux']
on_import = True
def _get_image_references(self, txt, base_dir):
from calibre.ebooks.txt.processor import get_images_from_polyglot_text
return get_images_from_polyglot_text(txt, base_dir)
def run(self, path_to_ebook):
from calibre.ebooks.metadata.opf2 import metadata_to_opf
from calibre.ebooks.txt.processor import get_images_from_polyglot_text
with open(path_to_ebook, 'rb') as ebf:
txt = ebf.read().decode('utf-8', 'replace')
base_dir = os.path.dirname(path_to_ebook)
images = self._get_image_references(txt, base_dir)
ext = path_to_ebook.rpartition('.')[-1].lower()
images = get_images_from_polyglot_text(txt, base_dir, ext)
if images:
# Create TXTZ and put file plus images inside of it.

View File

@ -152,7 +152,9 @@ class TXTInput(InputFormatPlugin):
zf.extractall('.')
for x in walk('.'):
if os.path.splitext(x)[1].lower() in ('.txt', '.text'):
ext = os.path.splitext(x)[1].lower()
if ext in ('.txt', '.text', '.textile', '.md', '.markdown'):
file_ext = ext
with open(x, 'rb') as tf:
txt += tf.read() + b'\n\n'
if os.path.exists('metadata.opf'):
@ -168,10 +170,17 @@ class TXTInput(InputFormatPlugin):
if txt_formatting is not None and txt_formatting.text:
txt_formatting = txt_formatting.text.strip()
if txt_formatting in ('plain', 'textile', 'markdown') and options.formatting_type == 'auto':
log.info(f'Using metadata from TXTZ archive to set text formmating type to: {txt_formatting}')
log.info(f'Using metadata from TXTZ archive to set text formating type to: {txt_formatting}')
options.formatting_type = txt_formatting
if txt_formatting != 'plain':
options.paragraph_type = 'off'
if options.formatting_type == 'auto':
if file_ext == 'textile':
options.formatting_type = txt_formatting
options.paragraph_type = 'off'
elif file_ext in ('md', 'markdown'):
options.formatting_type = txt_formatting
options.paragraph_type = 'off'
else:
if getattr(stream, 'name', None):
base_dir = os.path.dirname(stream.name)

View File

@ -343,7 +343,7 @@ def detect_formatting_type(txt):
return 'heuristic'
def get_images_from_polyglot_text(txt: str, base_dir: str = '') -> set:
def get_images_from_polyglot_text(txt: str, base_dir: str = '', file_ext: str = 'txt') -> set:
from calibre.ebooks.oeb.base import OEB_IMAGES
from calibre import guess_type
if not base_dir:
@ -354,17 +354,19 @@ def get_images_from_polyglot_text(txt: str, base_dir: str = '') -> set:
if path and not os.path.isabs(path) and guess_type(path)[0] in OEB_IMAGES and os.path.exists(os.path.join(base_dir, path)):
images.add(path)
# Textile
for m in re.finditer(r'(?mu)(?:[\[{])?\!(?:\. )?(?P<path>[^\s(!]+)\s?(?:\(([^\)]+)\))?\!(?::(\S+))?(?:[\]}]|(?=\s|$))', txt):
path = m.group('path')
check_path(path)
if file_ext in ('txt', 'text', 'textile'):
# Textile
for m in re.finditer(r'(?mu)(?:[\[{])?\!(?:\. )?(?P<path>[^\s(!]+)\s?(?:\(([^\)]+)\))?\!(?::(\S+))?(?:[\]}]|(?=\s|$))', txt):
path = m.group('path')
check_path(path)
# Markdown
from markdown import Markdown
html = HTML_TEMPLATE % ('', Markdown().convert(txt))
from html5_parser import parse
root = parse(html)
for img in root.iterdescendants('img'):
path = img.get('src')
check_path(path)
if file_ext in ('txt', 'text', 'md', 'markdown'):
# Markdown
from markdown import Markdown
html = HTML_TEMPLATE % ('', Markdown().convert(txt))
from html5_parser import parse
root = parse(html)
for img in root.iterdescendants('img'):
path = img.get('src')
check_path(path)
return images