Skip to content

Commit fee188a

Browse files
ShaileshS1702ShaileshS1702
committed
fix numbering on website by using local file
1 parent 539798f commit fee188a

File tree

1 file changed

+119
-95
lines changed

1 file changed

+119
-95
lines changed

‎filters/pandoc-section-links‎

Lines changed: 119 additions & 95 deletions
Original file line numberDiff line numberDiff line change
@@ -18,103 +18,127 @@ def process_pandoc_json(doc):
1818
current_section = 1 # Start from 1 since we'll add offset to get 0
1919
section_counter_found = False
2020

21-
def process_element(elem):
22-
nonlocal section_offset, current_section, section_counter_found
23-
24-
if isinstance(elem, dict):
25-
# Handle RawBlock LaTeX commands
26-
if elem.get('t') == 'RawBlock' and elem.get('c', [None, None])[0] == 'latex':
27-
latex_content = elem['c'][1]
28-
29-
# Check for \\setcounter{section}{-1}
30-
if r'\setcounter{section}{-1}' in latex_content:
31-
section_offset = -1
32-
section_counter_found = True
33-
# Remove this command from output
34-
return {'t': 'Null'}
35-
36-
# Handle RawInline LaTeX commands
37-
elif elem.get('t') == 'RawInline' and elem.get('c', [None, None])[0] == 'latex':
38-
latex_content = elem['c'][1]
39-
40-
# Check for \\setcounter{section}{-1}
41-
if r'\setcounter{section}{-1}' in latex_content:
42-
section_offset = -1
43-
section_counter_found = True
44-
# Remove this command from output
45-
return {'t': 'Str', 'c': ''}
46-
47-
# Handle Headers (sections)
48-
elif elem.get('t') == 'Header':
49-
level = elem['c'][0]
50-
attr = elem['c'][1]
51-
content = elem['c'][2]
52-
53-
if level == 1: # Top-level section
54-
# Calculate the adjusted section number (starting from 0 if offset is -1)
55-
adjusted_section = current_section + section_offset
56-
current_section += 1
57-
58-
# Store label mapping if this section has a label
59-
section_id = attr[0] if attr[0] else None
60-
if section_id:
61-
section_labels[section_id] = adjusted_section
62-
63-
# Clean up section title - remove § symbols and numbers
64-
if content:
65-
# Extract the existing text content
66-
existing_text = ""
67-
for item in content:
68-
if isinstance(item, dict) and item.get('t') == 'Str':
69-
existing_text += item['c']
70-
elif isinstance(item, dict) and item.get('t') == 'Space':
71-
existing_text += " "
72-
73-
# Remove any existing § symbol and number from the beginning
74-
if existing_text.startswith('§'):
75-
import re
76-
existing_text = re.sub(r'^§\d+\s*', '', existing_text)
77-
78-
# Create new content with just the clean title (no § symbol)
79-
new_content = [{'t': 'Str', 'c': existing_text}]
80-
elem['c'][2] = new_content
81-
82-
return elem
83-
84-
# Handle Links (cross-references)
85-
elif elem.get('t') == 'Link':
86-
attr = elem['c'][0]
87-
content = elem['c'][1]
88-
target = elem['c'][2]
89-
url = target[0]
90-
title = target[1]
91-
92-
# Check if this is an internal reference (starts with #)
93-
if url.startswith('#'):
94-
ref_id = url[1:] # Remove the #
95-
if ref_id in section_labels:
96-
# Update the link text to show correct section number
97-
adjusted_num = section_labels[ref_id]
98-
# Replace the content with the correct section number
99-
elem['c'][1] = [{'t': 'Str', 'c': f'§{adjusted_num}'}]
100-
101-
return elem
102-
103-
# Recursively process other elements
104-
elif isinstance(elem, dict):
105-
for key, value in elem.items():
106-
if isinstance(value, (list, dict)):
107-
elem[key] = process_element(value)
108-
109-
elif isinstance(elem, list):
110-
return [process_element(item) for item in elem]
111-
112-
return elem
21+
22+
# Two-pass approach: first gather headers and labels, then transform
23+
def gather_labels(blocks):
24+
"""Walk blocks and gather label -> hierarchical-number strings.
25+
We maintain counters per header level. For example, for a subsection
26+
we produce '13.6'. The section counter (level 1) gets the section_offset applied.
27+
"""
28+
labels = {}
29+
# counters indexed by level-1: counters[0]=section, [1]=subsection, ...
30+
counters = [0, 0, 0]
31+
32+
def walk(bs):
33+
for b in bs:
34+
if isinstance(b, dict) and b.get('t') == 'RawBlock' and isinstance(b.get('c'), list) and b['c'][0] == 'latex':
35+
text = b['c'][1]
36+
if r'\\setcounter{section}{-1}' in text or r'\setcounter{section}{-1}' in text:
37+
# set the offset so sections are effectively zero-based
38+
nonlocal_set_offset()
39+
if isinstance(b, dict) and b.get('t') == 'Header':
40+
level = b['c'][0]
41+
attr = b['c'][1]
42+
# ensure counters list long enough
43+
if level-1 >= len(counters):
44+
counters.extend([0] * (level - len(counters)))
45+
# increment this level and zero deeper levels
46+
counters[level-1] += 1
47+
for i in range(level, len(counters)):
48+
counters[i] = 0
49+
# compute hierarchical number string
50+
# apply section_offset only to top-level counter
51+
parts = []
52+
for i in range(level):
53+
val = counters[i]
54+
if i == 0:
55+
val = val + section_offset
56+
parts.append(str(val))
57+
numstr = '.'.join(parts)
58+
if attr and attr[0]:
59+
labels[attr[0]] = numstr
60+
# recurse into possible nested lists
61+
if isinstance(b, dict):
62+
for key in ('c', 'content', 'blocks'):
63+
if key in b and isinstance(b[key], list):
64+
walk(b[key])
65+
66+
# helper to set section_offset from inner scope
67+
def nonlocal_set_offset():
68+
nonlocal section_offset
69+
section_offset = -1
70+
71+
walk(blocks)
72+
return labels
73+
74+
def transform_blocks(blocks):
75+
"""Transform headers (clean titles) and replace link texts using section_labels mapping.
76+
section_labels values are strings like '13.6' now.
77+
"""
78+
# counters for producing header numbers during transform (keep consistent with gather)
79+
counters = [0, 0, 0]
80+
81+
def walk_and_transform(bs):
82+
new = []
83+
for b in bs:
84+
if isinstance(b, dict) and b.get('t') == 'Header':
85+
level = b['c'][0]
86+
content = b['c'][2]
87+
# ensure counters long enough
88+
if level-1 >= len(counters):
89+
counters.extend([0] * (level - len(counters)))
90+
counters[level-1] += 1
91+
for i in range(level, len(counters)):
92+
counters[i] = 0
93+
# clean title text regardless of level
94+
existing_text = ''
95+
for item in content:
96+
if isinstance(item, dict) and item.get('t') == 'Str':
97+
existing_text += item['c']
98+
elif isinstance(item, dict) and item.get('t') == 'Space':
99+
existing_text += ' '
100+
if existing_text.startswith('§'):
101+
existing_text = re.sub(r'^§\d+(?:\.\d+)*\s*', '', existing_text)
102+
b['c'][2] = [{'t': 'Str', 'c': existing_text}]
103+
# Transform links and recurse
104+
def transform_element(elem):
105+
if isinstance(elem, dict):
106+
if elem.get('t') == 'Link':
107+
target = elem['c'][2]
108+
url = target[0]
109+
if isinstance(url, str) and url.startswith('#'):
110+
ref_id = url[1:]
111+
if ref_id in section_labels:
112+
num = section_labels[ref_id]
113+
# set link text to the hierarchical number string (no §)
114+
elem['c'][1] = [{'t': 'Str', 'c': num}]
115+
return elem
116+
for k, v in list(elem.items()):
117+
if isinstance(v, list):
118+
elem[k] = [transform_element(x) for x in v]
119+
elif isinstance(elem, list):
120+
return [transform_element(x) for x in elem]
121+
return elem
122+
123+
# Recurse into nested lists
124+
if isinstance(b, dict):
125+
for key in ('c', 'content', 'blocks'):
126+
if key in b and isinstance(b[key], list):
127+
b[key] = walk_and_transform(b[key])
128+
b = transform_element(b)
129+
new.append(b)
130+
return new
131+
132+
return walk_and_transform(blocks)
113133

114-
# Process the document
134+
# Two-pass processing: gather labels first, then transform blocks
115135
if 'blocks' in doc:
116-
doc['blocks'] = process_element(doc['blocks'])
117-
136+
# gather labels (fills section_labels)
137+
gathered = gather_labels(doc['blocks'])
138+
section_labels.update(gathered)
139+
# transform blocks using the gathered labels
140+
current_section = 1
141+
doc['blocks'] = transform_blocks(doc['blocks'])
118142
return doc
119143

120144
def main():

0 commit comments

Comments
 (0)