import os
import re
FILES_LIST = "list.txt"
BSC_ROOT = "."
DOCS_OUTPUT = "docs"
RAW_BASE_URL = "https://raw.usercontent.amit.is-a.dev"
# Protected paths that should never be overwritten
PROTECTED_INDEX_FILES = {
os.path.normpath(os.path.join(DOCS_OUTPUT, f"semester_{i}", "index.md"))
for i in range(1, 9)
}
PROTECTED_INDEX_FILES.add(os.path.normpath(os.path.join(DOCS_OUTPUT, "index.md")))
ALGO_FOLDER_NAME = "algorithms" # folder name that contains algorithm .md files
# Directories excluded from index generation
IGNORED_FOLDERS = {"stylesheets", "overrides", "assets", ".vitepress", "node_modules", "public"}
# Language config mapped by file extension
SUPPORTED_LANGS = {
'.c': {'label': 'C', 'fence': 'c', 'style': 'c'},
'.cpp': {'label': 'C++', 'fence': 'cpp', 'style': 'c'},
'.h': {'label': 'C Header', 'fence': 'c', 'style': 'c'},
'.hpp': {'label': 'C++ Header', 'fence': 'cpp', 'style': 'c'},
'.r': {'label': 'R', 'fence': 'r', 'style': 'hash'},
'.py': {'label': 'Python', 'fence': 'python', 'style': 'hash'},
'.java': {'label': 'Java', 'fence': 'java', 'style': 'c'},
'.js': {'label': 'JavaScript', 'fence': 'javascript', 'style': 'c'},
'.ts': {'label': 'TypeScript', 'fence': 'typescript', 'style': 'c'},
'.sh': {'label': 'Bash', 'fence': 'bash', 'style': 'hash'},
}
def read_block_comment(lines, start):
result = []
i = start
n = len(lines)
first = lines[i].strip()
if first.startswith('/*') and '*/' in first:
inner = first[2: first.index('*/')].strip().strip('*').strip()
if inner:
result.append(inner)
return result, i + 1
inner = first[2:].strip().strip('*').strip()
if inner:
result.append(inner)
i += 1
while i < n:
line = lines[i].strip()
if '*/' in line:
text = line[: line.index('*/')].strip().strip('*').strip()
if text:
result.append(text)
return result, i + 1
text = line.strip('*').strip()
if text:
result.append(text)
i += 1
return result, i
def esc_yaml(s: str) -> str:
return s.replace("'", "''")
def esc_html(s: str) -> str:
"""Escape characters that break Vue template compilation in VitePress."""
return s.replace('&', '&').replace('<', '<').replace('>', '>')
def format_author(author: str) -> str:
"""Return a mailto markdown link if the author string contains an email."""
m = re.search(r'(.*?)\s*<(.*?)>', author)
if m:
name = esc_html(m.group(1).strip())
email = esc_html(m.group(2).strip())
return f"[{name}](mailto:{email})"
return esc_html(author)
def parse_c_style(content):
lines = content.splitlines()
n = len(lines)
i = 0
author = date = repo = license_str = problem_statement = ""
while i < n and not lines[i].strip():
i += 1
# Extract metadata block
if i < n and lines[i].strip().startswith('/*'):
block, i = read_block_comment(lines, i)
for line in block:
# Handles both newline-per-field and pipe-separated formats
parts = [p.strip() for p in line.split('|')]
for part in parts:
if ':' in part:
key, _, val = part.partition(':')
key = key.strip().lower()
val = val.strip()
if 'author' in key: author = val
elif 'date' in key: date = val
elif 'repo' in key: repo = val
elif 'license' in key: license_str = val
while i < n and not lines[i].strip():
i += 1
# Extract problem statement block (ensures it's not the actual code starting)
if i < n and lines[i].strip().startswith('/*'):
peek_block, peek_i = read_block_comment(lines, i)
block_text = ' '.join(peek_block)
if '#include' not in block_text and 'import ' not in block_text:
problem_statement = ' '.join(p for p in peek_block if p).strip()
i = peek_i
# Locate the beginning of actual source code
code_start = None
for j in list(range(i, n)) + list(range(0, i)):
line_strip = lines[j].strip()
if line_strip.startswith('#include') or line_strip.startswith('import '):
code_start = j
break
code = '\n'.join(lines[code_start:]).strip() if code_start is not None else content.strip()
return author, date, repo, license_str, problem_statement, code
def parse_hash_style(content):
lines = content.splitlines()
n = len(lines)
i = 0
author = date = repo = license_str = problem_statement = ""
while i < n and not lines[i].strip():
i += 1
# Extract metadata block
meta_lines = []
while i < n and lines[i].strip().startswith('#'):
meta_lines.append(lines[i].strip()[1:].strip())
i += 1
for raw_line in meta_lines:
for part in raw_line.split('|'):
if ':' in part:
key, _, val = part.partition(':')
key = key.strip().lower()
val = val.strip()
if 'author' in key: author = val
elif 'date' in key: date = val
elif 'repo' in key: repo = val
elif 'license' in key: license_str = val
while i < n and not lines[i].strip():
i += 1
# Extract problem statement block
ps_lines = []
while i < n and lines[i].strip().startswith('#'):
text = lines[i].strip()[1:].strip()
if text:
ps_lines.append(text)
i += 1
if ps_lines:
problem_statement = ' '.join(ps_lines).strip()
code = content.strip()
return author, date, repo, license_str, problem_statement, code
ALGO_ICON_SVG = ''
def parse_algo_md(content):
"""Parse a GitHub-style algorithm .md file and convert to VitePress format."""
lines = content.splitlines()
title = ""
problem_statement = ""
body_lines = []
i = 0
n = len(lines)
# Extract title from first # heading
while i < n:
line = lines[i]
if line.startswith("# "):
title = line[2:].strip()
i += 1
break
i += 1
# Parse rest: find problem statement in > blockquote under ### Problem Statement
in_problem_section = False
ps_lines = []
while i < n:
line = lines[i]
stripped = line.strip()
if stripped.lower().startswith("### problem statement"):
in_problem_section = True
i += 1
continue
if in_problem_section:
# Allow blank lines between heading and blockquote
if stripped == "":
i += 1
continue
if stripped.startswith("> "):
ps_lines.append(stripped[2:].strip())
i += 1
continue
elif stripped == ">":
i += 1
continue
else:
# Non-blockquote, non-blank line ends problem section
in_problem_section = False
problem_statement = " ".join(ps_lines).strip()
body_lines.append(line)
else:
body_lines.append(line)
i += 1
if in_problem_section and ps_lines:
problem_statement = " ".join(ps_lines).strip()
return title, problem_statement, body_lines
def build_algo_md(filename_base, title, problem_statement, body_lines, source_path=""):
"""Build VitePress .md from parsed algo content."""
desc = problem_statement if problem_statement else f"Algorithm — {title}"
fm = [
"---",
f"title: '{ALGO_ICON_SVG} {esc_yaml(title)}'",
f"description: '{esc_yaml(desc)}'",
f"source: '{source_path}'",
"---",
"",
]
body = [f"# {title}", ""]
if problem_statement:
body += [
"### Problem Statement",
"",
"::: tip Problem Statement",
esc_html(problem_statement),
":::",
"",
]
# Replace > blockquote problem section in body with nothing (already handled above)
# Just append the remaining body lines (Algorithm, Pseudocode, Complexity etc.)
skip_next_blockquote = False
cleaned = []
blines = body_lines[:]
j = 0
while j < len(blines):
l = blines[j]
s = l.strip()
if s.lower().startswith("### problem statement"):
# skip until blockquote ends
j += 1
while j < len(blines) and (blines[j].strip().startswith(">") or blines[j].strip() == ""):
j += 1
continue
cleaned.append(l)
j += 1
# Remove leading blank lines from cleaned
while cleaned and not cleaned[0].strip():
cleaned.pop(0)
body += cleaned
return "\n".join(fm + body)
def build_md(filename, lang_label, fence_lang, author, date, repo, license_str,
problem_statement, code, raw_url, github_url, rel_url):
desc = problem_statement if problem_statement else f"{lang_label} program — {filename}"
icon_svg = ''
fm_lines = [
"---",
f"title: '{icon_svg} {esc_yaml(filename)}'",
f"description: '{esc_yaml(desc)}'",
f"source: '{rel_url}'",
"---",
]
body = [
"",
f"# {filename}",
""
]
has_meta = any([author, date, license_str])
if has_meta:
body.extend([
"### Metadata",
""
])
if author:
body.append(f"- **Author** — {format_author(author)}")
if date:
body.append(f"- **Last updated** — {esc_html(date)}")
if license_str:
body.append(
f"- **License** — [{esc_html(license_str)}]"
f"(https://github.com/notamitgamer/bsc/blob/main/LICENSE)"
)
body.append("")
if problem_statement:
body += [
"",
"### Problem Statement",
"",
f"::: tip {filename}",
esc_html(problem_statement),
":::",
"",
]
body += [
"## Source Code",
"",
f"```{fence_lang} [{filename}]",
code,
"```",
"---",
# Action links — minimal, right-aligned feel via HTML
'