Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2026-05-31 02:55:19 -04:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'bbc-sport-headline-block-fix' of https://github.com/claybdavis/calibre

Browse Source
This commit is contained in:
Kovid Goyal
2026-05-19 07:26:53 +05:30
parent 19d2488ca5 507feb15fa
commit d8490c2208
2 changed files with 70 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
recipes/bbc.recipe
+35
View File
@@ -85,6 +85,20 @@ def serialize_contributor(contributor):
yield '<div>' + prepare_string_for_xml(contributor['subtitle']) + '</div>'
def extract_text_block_plaintext(text_block):
# Pull the plain text out of a nested {type: text, model: {blocks: [...]}}
# block. Used to recover a headline string from the 'headline' / 'topper'
# block types that BBC Sport articles now use in place of the top-level
# article['headline'] field.
chunks = []
for x in text_block.get('model', {}).get('blocks', []):
if x.get('type') == 'paragraph':
for f in x.get('model', {}).get('blocks', []):
if f.get('type') == 'fragment':
chunks.append(f.get('model', {}).get('text', ''))
return ''.join(chunks)
def parse_article_json(root, abort_article):
data = root['data']
has_media_experience = False
@@ -102,14 +116,35 @@ def parse_article_json(root, abort_article):
lines = []
if article.get('headline'):
lines.append('<h1>{}</h1>'.format(prepare_string_for_xml(article['headline'])))
elif article.get('metadata', {}).get('seoHeadline'):
# BBC Sport articles no longer populate the top-level 'headline' field;
# the visible headline is on metadata.seoHeadline (and is plain text).
lines.append('<h1>{}</h1>'.format(prepare_string_for_xml(article['metadata']['seoHeadline'])))
if article.get('contributor'):
lines.extend(serialize_contributor(article['contributor']))
headline_emitted = bool(lines and lines[0].startswith('<h1>'))
for block in article['content']['model']['blocks']:
bt = block.get('type')
if bt == 'image':
lines.extend(serialize_image(block))
elif bt == 'text':
lines.extend(serialize_text(block))
elif bt == 'headline' and not headline_emitted:
# New BBC Sport block-type containing the headline as a nested text block.
inner = block.get('model', {}).get('blocks') or [{}]
text = extract_text_block_plaintext(inner[0])
if text:
lines.insert(0, '<h1>{}</h1>'.format(prepare_string_for_xml(text)))
headline_emitted = True
elif bt == 'topper' and not headline_emitted:
# 'high-impact' BBC Sport article variant: headline lives under
# block.model.heading.blocks[0] rather than block.model.blocks[0].
heading = block.get('model', {}).get('heading') or {}
inner = heading.get('blocks') or [{}]
text = extract_text_block_plaintext(inner[0])
if text:
lines.insert(0, '<h1>{}</h1>'.format(prepare_string_for_xml(text)))
headline_emitted = True
return '<html><body id="main-content"><article>' + '\n'.join(lines) + '</article></body></html>'
recipes/bbc_fast.recipe
+35
View File
@@ -85,6 +85,20 @@ def serialize_contributor(contributor):
yield '<div>' + prepare_string_for_xml(contributor['subtitle']) + '</div>'
def extract_text_block_plaintext(text_block):
# Pull the plain text out of a nested {type: text, model: {blocks: [...]}}
# block. Used to recover a headline string from the 'headline' / 'topper'
# block types that BBC Sport articles now use in place of the top-level
# article['headline'] field.
chunks = []
for x in text_block.get('model', {}).get('blocks', []):
if x.get('type') == 'paragraph':
for f in x.get('model', {}).get('blocks', []):
if f.get('type') == 'fragment':
chunks.append(f.get('model', {}).get('text', ''))
return ''.join(chunks)
def parse_article_json(root, abort_article):
data = root['data']
has_media_experience = False
@@ -102,14 +116,35 @@ def parse_article_json(root, abort_article):
lines = []
if article.get('headline'):
lines.append('<h1>{}</h1>'.format(prepare_string_for_xml(article['headline'])))
elif article.get('metadata', {}).get('seoHeadline'):
# BBC Sport articles no longer populate the top-level 'headline' field;
# the visible headline is on metadata.seoHeadline (and is plain text).
lines.append('<h1>{}</h1>'.format(prepare_string_for_xml(article['metadata']['seoHeadline'])))
if article.get('contributor'):
lines.extend(serialize_contributor(article['contributor']))
headline_emitted = bool(lines and lines[0].startswith('<h1>'))
for block in article['content']['model']['blocks']:
bt = block.get('type')
if bt == 'image':
lines.extend(serialize_image(block))
elif bt == 'text':
lines.extend(serialize_text(block))
elif bt == 'headline' and not headline_emitted:
# New BBC Sport block-type containing the headline as a nested text block.
inner = block.get('model', {}).get('blocks') or [{}]
text = extract_text_block_plaintext(inner[0])
if text:
lines.insert(0, '<h1>{}</h1>'.format(prepare_string_for_xml(text)))
headline_emitted = True
elif bt == 'topper' and not headline_emitted:
# 'high-impact' BBC Sport article variant: headline lives under
# block.model.heading.blocks[0] rather than block.model.blocks[0].
heading = block.get('model', {}).get('heading') or {}
inner = heading.get('blocks') or [{}]
text = extract_text_block_plaintext(inner[0])
if text:
lines.insert(0, '<h1>{}</h1>'.format(prepare_string_for_xml(text)))
headline_emitted = True
return '<html><body id="main-content"><article>' + '\n'.join(lines) + '</article></body></html>'