804 lines
27 KiB
Python
804 lines
27 KiB
Python
#!/usr/bin/env python3
|
||
"""Extract The Nectar of Devotion (1970 Edition) from other.html.
|
||
|
||
The book has 3 intro segments (Front Flap, Preface, Introduction),
|
||
51 chapters organized in 4 parts, and a Concluding Words section.
|
||
|
||
Output:
|
||
lib/jsons/books/nod/intro.json — Book metadata
|
||
lib/jsons/books/nod/intro/*.json — Intro segments
|
||
lib/jsons/books/nod/chapters/*.json — Chapters + concluding words
|
||
|
||
Usage:
|
||
python3 scripts/nod_extract.py
|
||
"""
|
||
|
||
import json
|
||
import os
|
||
import re
|
||
import sys
|
||
|
||
from bs4 import BeautifulSoup, Tag
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# ScaGoudy → IAST Unicode mapping
|
||
# ---------------------------------------------------------------------------
|
||
SCAGOUDY_TO_IAST = {
|
||
"ä": "ā",
|
||
"Ä": "Ā",
|
||
"å": "ṛ",
|
||
"Å": "Ṛ",
|
||
"ñ": "ṣ",
|
||
"Ñ": "Ṣ",
|
||
"ö": "ṭ",
|
||
"Ö": "Ṭ",
|
||
"ë": "ṇ",
|
||
"Ë": "Ṇ",
|
||
"ò": "ḍ",
|
||
"Ò": "Ḍ",
|
||
"ç": "ś",
|
||
"Ç": "Ś",
|
||
"ì": "ṅ",
|
||
"ï": "ñ",
|
||
"ü": "ū",
|
||
"Ü": "Ū",
|
||
"é": "ī",
|
||
"É": "Ī",
|
||
"à": "ṁ",
|
||
"ù": "ḥ",
|
||
"è": "ṝ",
|
||
"û": "ṁl",
|
||
"ÿ": "ḷ",
|
||
}
|
||
|
||
|
||
def decode_scagoudy(text: str) -> str:
|
||
"""Replace ScaGoudy-encoded characters with proper IAST Unicode."""
|
||
for src, dst in SCAGOUDY_TO_IAST.items():
|
||
text = text.replace(src, dst)
|
||
return text
|
||
|
||
|
||
def decode_html_content(tag: Tag) -> str:
|
||
"""Extract inner HTML from a tag, decode ScaGoudy, preserve formatting tags."""
|
||
inner = "".join(str(child) for child in tag.children)
|
||
inner = decode_scagoudy(inner)
|
||
# Remove span tags with CSS classes (RTF conversion artifacts)
|
||
inner = re.sub(
|
||
r'<span[^>]*class="[^"]*"[^>]*>(.*?)</span>', r"\1", inner, flags=re.DOTALL
|
||
)
|
||
# Clean up Apple-tab-span leftovers (tab characters)
|
||
inner = inner.replace("\t", " ")
|
||
# Normalize whitespace
|
||
inner = re.sub(r"\s+", " ", inner).strip()
|
||
return inner
|
||
|
||
|
||
def decode_text(tag: Tag) -> str:
|
||
"""Get plain text from a tag with ScaGoudy decoding."""
|
||
return decode_scagoudy(tag.get_text().strip())
|
||
|
||
|
||
def get_css_class(tag: Tag) -> str:
|
||
"""Get the first CSS class of a tag, or empty string."""
|
||
classes = tag.get("class", [])
|
||
return classes[0] if classes else ""
|
||
|
||
|
||
def is_empty(tag: Tag) -> bool:
|
||
"""Check if a tag is empty or whitespace-only."""
|
||
text = tag.get_text().strip()
|
||
return not text or text == "\xa0"
|
||
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# CSS class categorization for NoD
|
||
# ---------------------------------------------------------------------------
|
||
|
||
# Bookmark class — used for section boundary detection
|
||
BOOKMARK_CLASS = "p13"
|
||
|
||
# Book title header class
|
||
BOOK_HEADER_CLASS = "p12"
|
||
|
||
# Body paragraph classes
|
||
BODY_CLASSES = frozenset(["p6", "p35", "p30"])
|
||
|
||
# Section heading (bold) — "Chapter One", "Preface", etc.
|
||
SECTION_HEADING_CLASS = "p31"
|
||
|
||
# Chapter title (bold) — "Characteristics of Pure Devotional Service", etc.
|
||
CHAPTER_TITLE_CLASS = "p27"
|
||
|
||
# Part header class — "PART ONE", "PART TWO", etc.
|
||
PART_HEADER_CLASS = "p34"
|
||
|
||
# Sub-section heading (bold) within chapters
|
||
SUB_HEADING_CLASS = "p36"
|
||
|
||
# Italic sub-headings
|
||
ITALIC_SUB_HEADING_CLASSES = frozenset(["p37", "p38"])
|
||
|
||
# Title page / copyright text classes (in intro area)
|
||
TITLE_PAGE_CLASS = "p3"
|
||
COPYRIGHT_CLASS = "p32"
|
||
|
||
# Sanskrit verse (italic with br tags)
|
||
SANSKRIT_VERSE_CLASS = "p33"
|
||
|
||
# Classes to skip entirely (empty/decorative)
|
||
SKIP_CLASSES = frozenset(["p9", "p15"])
|
||
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# Intro segment definitions
|
||
# ---------------------------------------------------------------------------
|
||
|
||
INTRO_SEGMENTS = [
|
||
{
|
||
"slug": "front-flap",
|
||
"title": "Front Flap",
|
||
"order": 1,
|
||
"bookmark_pattern": "NoD-1970: Front Flap",
|
||
},
|
||
{
|
||
"slug": "preface",
|
||
"title": "Preface",
|
||
"order": 2,
|
||
"bookmark_pattern": "NoD-1970: Preface",
|
||
},
|
||
{
|
||
"slug": "introduction",
|
||
"title": "Introduction",
|
||
"order": 3,
|
||
"bookmark_pattern": "NoD-1970: Introduction",
|
||
},
|
||
]
|
||
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# Extraction helpers
|
||
# ---------------------------------------------------------------------------
|
||
|
||
|
||
def find_nod_boundaries(all_paragraphs: list) -> tuple[int, int]:
|
||
"""Find the start and end paragraph indices of the NoD content.
|
||
|
||
Start: <p class="p12"> containing "Nectar of Devotion"
|
||
End: next <p class="p12"> after start, or end of file
|
||
"""
|
||
start_idx = None
|
||
end_idx = len(all_paragraphs)
|
||
|
||
for i, p in enumerate(all_paragraphs):
|
||
cls = get_css_class(p)
|
||
if cls != BOOK_HEADER_CLASS:
|
||
continue
|
||
text = decode_text(p)
|
||
if "Nectar of Devotion" in text:
|
||
start_idx = i
|
||
elif start_idx is not None:
|
||
# Next p12 after NoD start = end of NoD
|
||
end_idx = i
|
||
break
|
||
|
||
if start_idx is None:
|
||
print("ERROR: Could not find 'The Nectar of Devotion' header in other.html!")
|
||
sys.exit(1)
|
||
|
||
return start_idx, end_idx
|
||
|
||
|
||
def find_all_bookmarks(all_paragraphs: list, start_idx: int, end_idx: int) -> list:
|
||
"""Find all p13 bookmark tags in the NoD range.
|
||
|
||
Returns list of (paragraph_index, bookmark_text).
|
||
"""
|
||
bookmarks = []
|
||
for i in range(start_idx, end_idx):
|
||
p = all_paragraphs[i]
|
||
cls = get_css_class(p)
|
||
if cls == BOOKMARK_CLASS:
|
||
text = decode_text(p)
|
||
if "NoD" in text:
|
||
bookmarks.append((i, text))
|
||
return bookmarks
|
||
|
||
|
||
def parse_chapter_bookmark(text: str) -> tuple[int, str] | None:
|
||
"""Parse a chapter bookmark like 'NoD 1-1970: Characteristics of Pure Devotional Service'.
|
||
|
||
Returns (chapter_number, title) or None if not a chapter bookmark.
|
||
"""
|
||
m = re.match(r"NoD\s+(\d+)-1970:\s*(.+)", text.strip())
|
||
if m:
|
||
return int(m.group(1)), m.group(2).strip()
|
||
return None
|
||
|
||
|
||
def parse_intro_bookmark(text: str) -> str | None:
|
||
"""Parse an intro bookmark like 'NoD-1970: Preface'.
|
||
|
||
Returns the segment name or None if not an intro bookmark.
|
||
"""
|
||
m = re.match(r"NoD-1970:\s*(.+)", text.strip())
|
||
if m:
|
||
return m.group(1).strip()
|
||
return None
|
||
|
||
|
||
def extract_paragraphs(
|
||
all_paragraphs: list,
|
||
start_idx: int,
|
||
end_idx: int,
|
||
is_intro: bool = False,
|
||
) -> list[str]:
|
||
"""Extract paragraphs for a section (chapter or intro segment).
|
||
|
||
Handles all CSS classes within the NoD content.
|
||
"""
|
||
result: list[str] = []
|
||
|
||
for i in range(start_idx, end_idx):
|
||
p = all_paragraphs[i]
|
||
cls = get_css_class(p)
|
||
|
||
# Skip empty paragraphs
|
||
if is_empty(p):
|
||
continue
|
||
|
||
# Skip noise classes
|
||
if cls in SKIP_CLASSES:
|
||
continue
|
||
|
||
# Skip bookmark tags
|
||
if cls == BOOKMARK_CLASS:
|
||
continue
|
||
|
||
# Skip book title header
|
||
if cls == BOOK_HEADER_CLASS:
|
||
continue
|
||
|
||
# Part headers — include as bold paragraph
|
||
if cls == PART_HEADER_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Section heading (p31) — include as paragraph
|
||
if cls == SECTION_HEADING_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Chapter title (p27) — include as bold paragraph
|
||
if cls == CHAPTER_TITLE_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Sub-section headings within chapters (p36, bold)
|
||
if cls == SUB_HEADING_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Italic sub-headings (p37, p38)
|
||
if cls in ITALIC_SUB_HEADING_CLASSES:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Title page / copyright text (in intro area)
|
||
if cls == TITLE_PAGE_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Copyright text
|
||
if cls == COPYRIGHT_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Sanskrit verse (italic with br tags) — format as centered italic div
|
||
# Strip all inner <i>/<\/i> tags before re-wrapping to avoid nesting
|
||
if cls == SANSKRIT_VERSE_CLASS:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
stripped = re.sub(r"</?i>", "", html)
|
||
result.append(f'<div style="text-align:center"><i>{stripped}</i></div>')
|
||
continue
|
||
|
||
# Body paragraphs (p6, p35, p30)
|
||
if cls in BODY_CLASSES:
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
continue
|
||
|
||
# Any other class — try to extract content (be permissive)
|
||
html = decode_html_content(p)
|
||
if html:
|
||
result.append(html)
|
||
|
||
return result
|
||
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# Main extraction logic
|
||
# ---------------------------------------------------------------------------
|
||
|
||
|
||
def extract_nod(soup: BeautifulSoup) -> dict:
|
||
"""Extract the entire Nectar of Devotion from the parsed HTML.
|
||
|
||
Returns a dict with:
|
||
"intro_segments": list of intro segment dicts
|
||
"chapters": list of chapter dicts
|
||
"concluding_words": concluding words dict
|
||
"parts": list of part assignment dicts
|
||
"chapter_listing": list of {number, title} for intro.json
|
||
"""
|
||
all_paragraphs = list(soup.find_all("p"))
|
||
|
||
# Step 1: Find NoD boundaries
|
||
nod_start, nod_end = find_nod_boundaries(all_paragraphs)
|
||
print(f" NoD content: paragraphs {nod_start} to {nod_end}")
|
||
print(f" Total paragraphs in range: {nod_end - nod_start}")
|
||
|
||
# Step 2: Find all bookmarks
|
||
bookmarks = find_all_bookmarks(all_paragraphs, nod_start, nod_end)
|
||
print(f" Found {len(bookmarks)} NoD bookmarks")
|
||
|
||
# Step 3: Categorize bookmarks into intro vs chapter
|
||
intro_bookmarks = [] # (para_idx, segment_name)
|
||
chapter_bookmarks = [] # (para_idx, chapter_num, title)
|
||
|
||
for pos, text in bookmarks:
|
||
ch = parse_chapter_bookmark(text)
|
||
if ch:
|
||
chapter_bookmarks.append((pos, ch[0], ch[1]))
|
||
else:
|
||
intro_name = parse_intro_bookmark(text)
|
||
if intro_name:
|
||
intro_bookmarks.append((pos, intro_name))
|
||
|
||
print(f" Intro bookmarks: {len(intro_bookmarks)}")
|
||
print(f" Chapter bookmarks: {len(chapter_bookmarks)}")
|
||
|
||
# Step 4: Track PART assignments by scanning for p34 tags
|
||
# We need to know which chapters belong to which part.
|
||
part_assignments = {} # chapter_num -> part_num
|
||
parts_data = [] # list of {number, title, chapters}
|
||
|
||
# Walk through all paragraphs in NoD range and track parts
|
||
# We process this before extracting chapters so we know part assignments
|
||
part_positions = [] # (para_idx, part_title)
|
||
for i in range(nod_start, nod_end):
|
||
p = all_paragraphs[i]
|
||
cls = get_css_class(p)
|
||
if cls == PART_HEADER_CLASS:
|
||
text = decode_text(p)
|
||
if text:
|
||
part_positions.append((i, text))
|
||
|
||
print(f" Part headers found: {len(part_positions)}")
|
||
for pos, title in part_positions:
|
||
print(f" Line ~{pos}: {title}")
|
||
|
||
# Assign parts to chapters based on position
|
||
# A chapter belongs to the most recent PART header before its bookmark
|
||
part_number_map = {}
|
||
for pos, title in part_positions:
|
||
# Extract part number from title like "PART ONE", "PART TWO", etc.
|
||
part_num = None
|
||
title_upper = title.upper()
|
||
if "ONE" in title_upper or "I" == title_upper.replace("PART", "").strip():
|
||
part_num = 1
|
||
elif "TWO" in title_upper or "II" == title_upper.replace("PART", "").strip():
|
||
part_num = 2
|
||
elif "THREE" in title_upper or "III" == title_upper.replace("PART", "").strip():
|
||
part_num = 3
|
||
elif "FOUR" in title_upper or "IV" == title_upper.replace("PART", "").strip():
|
||
part_num = 4
|
||
if part_num:
|
||
part_number_map[pos] = part_num
|
||
|
||
# Now assign each chapter to a part
|
||
sorted_part_positions = sorted(part_number_map.keys())
|
||
for ch_pos, ch_num, ch_title in chapter_bookmarks:
|
||
# Find the most recent PART header before this chapter's bookmark
|
||
assigned_part = None
|
||
for pp in sorted_part_positions:
|
||
if pp < ch_pos:
|
||
assigned_part = part_number_map[pp]
|
||
else:
|
||
break
|
||
if assigned_part:
|
||
part_assignments[ch_num] = assigned_part
|
||
|
||
# Build parts_data
|
||
for part_num in sorted(set(part_assignments.values())):
|
||
chs = sorted([ch for ch, pn in part_assignments.items() if pn == part_num])
|
||
part_title = f"Part {'One' if part_num == 1 else 'Two' if part_num == 2 else 'Three' if part_num == 3 else 'Four'}"
|
||
parts_data.append(
|
||
{
|
||
"number": part_num,
|
||
"title": part_title,
|
||
"chapters": chs,
|
||
}
|
||
)
|
||
|
||
# Step 5: Extract intro segments
|
||
intro_segments = []
|
||
for seg_def in INTRO_SEGMENTS:
|
||
# Find matching bookmark
|
||
bookmark_pos = None
|
||
for pos, name in intro_bookmarks:
|
||
if (
|
||
seg_def["bookmark_pattern"].endswith(name)
|
||
or name == seg_def["bookmark_pattern"].split(": ", 1)[-1]
|
||
):
|
||
bookmark_pos = pos
|
||
break
|
||
|
||
if bookmark_pos is None:
|
||
print(f" WARNING: Could not find bookmark for '{seg_def['slug']}'")
|
||
continue
|
||
|
||
# Content starts after the bookmark
|
||
content_start = bookmark_pos + 1
|
||
|
||
# Content ends at the next bookmark (intro or chapter) or PART header
|
||
content_end = nod_end
|
||
all_section_starts = sorted(
|
||
[pos for pos, _ in intro_bookmarks]
|
||
+ [pos for pos, _, _ in chapter_bookmarks]
|
||
)
|
||
for s in all_section_starts:
|
||
if s > bookmark_pos:
|
||
content_end = s
|
||
break
|
||
|
||
paragraphs = extract_paragraphs(
|
||
all_paragraphs, content_start, content_end, is_intro=True
|
||
)
|
||
|
||
seg = {
|
||
"slug": seg_def["slug"],
|
||
"title": seg_def["title"],
|
||
"type": "intro-segment",
|
||
"book": "nod",
|
||
"order": seg_def["order"],
|
||
"paragraphs": paragraphs,
|
||
}
|
||
intro_segments.append(seg)
|
||
|
||
# Step 6: Extract chapters
|
||
chapters = []
|
||
for idx, (ch_pos, ch_num, ch_title) in enumerate(chapter_bookmarks):
|
||
content_start = ch_pos + 1
|
||
|
||
# Content ends at next chapter bookmark or "Concluding Words" p27 or nod_end
|
||
if idx + 1 < len(chapter_bookmarks):
|
||
content_end = chapter_bookmarks[idx + 1][0]
|
||
else:
|
||
# Last chapter (51) — content ends at "Concluding Words" p27 heading
|
||
# or at nod_end. We need to find where "Concluding Words" starts.
|
||
# Look for p27 containing "Concluding Words" after the last chapter bookmark.
|
||
concluding_pos = None
|
||
for i in range(ch_pos + 1, nod_end):
|
||
p = all_paragraphs[i]
|
||
cls = get_css_class(p)
|
||
if cls == CHAPTER_TITLE_CLASS:
|
||
text = decode_text(p)
|
||
if "Concluding Words" in text:
|
||
concluding_pos = i
|
||
break
|
||
|
||
if concluding_pos:
|
||
content_end = concluding_pos
|
||
else:
|
||
content_end = nod_end
|
||
|
||
# For the PART header: if a PART header appears right before this
|
||
# chapter's bookmark, include it in this chapter's content range.
|
||
# The PART header might be 1-3 paragraphs before the bookmark (with
|
||
# possible p15 empty lines in between).
|
||
actual_start = content_start
|
||
for pp in sorted_part_positions:
|
||
# Check if this PART header is between the previous section end and
|
||
# this chapter's bookmark
|
||
if pp < ch_pos and pp >= content_start - 5:
|
||
# Check that there's no other bookmark between pp and ch_pos
|
||
has_other_bookmark = False
|
||
for bpos, _ in bookmarks:
|
||
if bpos > pp and bpos < ch_pos:
|
||
has_other_bookmark = True
|
||
break
|
||
if not has_other_bookmark:
|
||
actual_start = pp
|
||
|
||
paragraphs = extract_paragraphs(all_paragraphs, actual_start, content_end)
|
||
|
||
ch = {
|
||
"book": "nod",
|
||
"chapter": ch_num,
|
||
"title": ch_title,
|
||
"type": "chapter",
|
||
"paragraphs": paragraphs,
|
||
}
|
||
chapters.append(ch)
|
||
|
||
# Step 7: Extract "Concluding Words"
|
||
concluding_words = None
|
||
# Find the "Concluding Words" p27 heading after chapter 51
|
||
last_ch_pos = chapter_bookmarks[-1][0] if chapter_bookmarks else nod_start
|
||
for i in range(last_ch_pos, nod_end):
|
||
p = all_paragraphs[i]
|
||
cls = get_css_class(p)
|
||
if cls == CHAPTER_TITLE_CLASS:
|
||
text = decode_text(p)
|
||
if "Concluding Words" in text:
|
||
# Extract from this heading to nod_end
|
||
paragraphs = extract_paragraphs(all_paragraphs, i, nod_end)
|
||
concluding_words = {
|
||
"book": "nod",
|
||
"chapter": None,
|
||
"title": "Concluding Words",
|
||
"type": "conclusion",
|
||
"slug": "concluding-words",
|
||
"paragraphs": paragraphs,
|
||
}
|
||
break
|
||
|
||
# Step 8: Build chapter listing for intro.json
|
||
chapter_listing = [
|
||
{"number": ch["chapter"], "title": ch["title"]} for ch in chapters
|
||
]
|
||
|
||
return {
|
||
"intro_segments": intro_segments,
|
||
"chapters": chapters,
|
||
"concluding_words": concluding_words,
|
||
"parts": parts_data,
|
||
"chapter_listing": chapter_listing,
|
||
}
|
||
|
||
|
||
# ---------------------------------------------------------------------------
|
||
# Output
|
||
# ---------------------------------------------------------------------------
|
||
|
||
|
||
def write_json(data, filepath):
|
||
"""Write data as formatted JSON to filepath, creating directories as needed."""
|
||
os.makedirs(os.path.dirname(filepath), exist_ok=True)
|
||
with open(filepath, "w", encoding="utf-8") as f:
|
||
json.dump(data, f, ensure_ascii=False, indent=2)
|
||
|
||
|
||
def main():
|
||
base_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
|
||
html_path = os.path.join(base_dir, "public", "html", "other.html")
|
||
output_base = os.path.join(base_dir, "lib", "jsons", "books", "nod")
|
||
|
||
if not os.path.exists(html_path):
|
||
print(f"ERROR: {html_path} not found!")
|
||
sys.exit(1)
|
||
|
||
print(f"Reading {html_path}...")
|
||
with open(html_path, "r", encoding="utf-8") as f:
|
||
soup = BeautifulSoup(f, "html.parser")
|
||
|
||
print("Extracting The Nectar of Devotion (1970 Edition)...\n")
|
||
result = extract_nod(soup)
|
||
|
||
intro_segments = result["intro_segments"]
|
||
chapters = result["chapters"]
|
||
concluding_words = result["concluding_words"]
|
||
parts = result["parts"]
|
||
chapter_listing = result["chapter_listing"]
|
||
|
||
# --- Write intro.json ---
|
||
intro_meta = {
|
||
"book": "nod",
|
||
"title": "The Nectar of Devotion",
|
||
"subtitle": "A Summary Study of Śrīla Rūpa Gosvāmī's Bhakti-rasāmṛta-sindhu",
|
||
"author": "His Divine Grace A.C. Bhaktivedanta Swami Prabhupāda",
|
||
"year": 1970,
|
||
"description": "The complete science of bhakti-yoga, or devotional service to Lord Kṛṣṇa.",
|
||
"totalChapters": 51,
|
||
"parts": parts,
|
||
"chapters": chapter_listing,
|
||
}
|
||
intro_path = os.path.join(output_base, "intro.json")
|
||
write_json(intro_meta, intro_path)
|
||
print(f"\n Written: {intro_path}")
|
||
|
||
# --- Write intro segments ---
|
||
print(f"\n--- Intro Segments ---")
|
||
for seg in intro_segments:
|
||
filepath = os.path.join(output_base, "intro", f"{seg['slug']}.json")
|
||
write_json(seg, filepath)
|
||
print(
|
||
f" {seg['order']}. {seg['title']}: {len(seg['paragraphs'])} paragraphs → {filepath}"
|
||
)
|
||
|
||
# --- Write chapters ---
|
||
print(f"\n--- Chapters ---")
|
||
for ch in chapters:
|
||
ch_num_str = str(ch["chapter"]).zfill(2)
|
||
filepath = os.path.join(output_base, "chapters", f"{ch_num_str}.json")
|
||
write_json(ch, filepath)
|
||
part_info = ""
|
||
for p in parts:
|
||
if ch["chapter"] in p["chapters"]:
|
||
part_info = f" [Part {p['number']}]"
|
||
break
|
||
print(
|
||
f" Ch {ch['chapter']:>2}: {len(ch['paragraphs']):>3} paragraphs — "
|
||
f"{ch['title']}{part_info}"
|
||
)
|
||
|
||
# --- Write concluding words ---
|
||
if concluding_words:
|
||
filepath = os.path.join(output_base, "chapters", "concluding-words.json")
|
||
write_json(concluding_words, filepath)
|
||
print(
|
||
f"\n Concluding Words: {len(concluding_words['paragraphs'])} paragraphs → {filepath}"
|
||
)
|
||
|
||
# --- Summary ---
|
||
total_intro_paras = sum(len(s["paragraphs"]) for s in intro_segments)
|
||
total_chapter_paras = sum(len(ch["paragraphs"]) for ch in chapters)
|
||
total_concluding_paras = (
|
||
len(concluding_words["paragraphs"]) if concluding_words else 0
|
||
)
|
||
total_paras = total_intro_paras + total_chapter_paras + total_concluding_paras
|
||
|
||
print(f"\n{'=' * 60}")
|
||
print(f" EXTRACTION SUMMARY")
|
||
print(f"{'=' * 60}")
|
||
print(f" Intro segments: {len(intro_segments)}")
|
||
print(f" Chapters: {len(chapters)}")
|
||
print(f" Concluding Words: {'Yes' if concluding_words else 'No'}")
|
||
print(f" Parts: {len(parts)}")
|
||
for p in parts:
|
||
print(
|
||
f" Part {p['number']} ({p['title']}): chapters {p['chapters'][0]}-{p['chapters'][-1]} ({len(p['chapters'])} chapters)"
|
||
)
|
||
print(f" Total paragraphs: {total_paras}")
|
||
print(f" Intro: {total_intro_paras}")
|
||
print(f" Chapters: {total_chapter_paras}")
|
||
print(f" Concluding: {total_concluding_paras}")
|
||
print(f" Avg paras/chapter: {total_chapter_paras / len(chapters):.1f}")
|
||
print(f" Output directory: {output_base}/")
|
||
|
||
# --- Samples ---
|
||
print(f"\n--- Sample: Chapter 1 ---")
|
||
if chapters:
|
||
ch1 = chapters[0]
|
||
for i, p in enumerate(ch1["paragraphs"][:4]):
|
||
preview = re.sub(r"<[^>]+>", "", p)[:100]
|
||
print(f" [{i}] {preview}...")
|
||
|
||
print(f"\n--- Sample: Chapter {chapters[-1]['chapter']} (last) ---")
|
||
if chapters:
|
||
last = chapters[-1]
|
||
for i, p in enumerate(last["paragraphs"][:3]):
|
||
preview = re.sub(r"<[^>]+>", "", p)[:100]
|
||
print(f" [{i}] {preview}...")
|
||
|
||
if concluding_words:
|
||
print(f"\n--- Sample: Concluding Words ---")
|
||
for i, p in enumerate(concluding_words["paragraphs"][:3]):
|
||
preview = re.sub(r"<[^>]+>", "", p)[:100]
|
||
print(f" [{i}] {preview}...")
|
||
|
||
# --- Quality Checks ---
|
||
print(f"\n--- Quality Checks ---")
|
||
|
||
# Check all 51 chapters present
|
||
expected_chapters = set(range(1, 52))
|
||
actual_chapters = set(ch["chapter"] for ch in chapters)
|
||
missing = expected_chapters - actual_chapters
|
||
extra = actual_chapters - expected_chapters
|
||
if missing:
|
||
print(f" ⚠ Missing chapters: {sorted(missing)}")
|
||
if extra:
|
||
print(f" ⚠ Extra chapters: {sorted(extra)}")
|
||
if not missing and not extra:
|
||
print(f" ✅ All 51 chapters present")
|
||
|
||
# Check all 3 intro segments present
|
||
expected_intros = {"front-flap", "preface", "introduction"}
|
||
actual_intros = {s["slug"] for s in intro_segments}
|
||
missing_intros = expected_intros - actual_intros
|
||
if missing_intros:
|
||
print(f" ⚠ Missing intro segments: {missing_intros}")
|
||
else:
|
||
print(f" ✅ All 3 intro segments present")
|
||
|
||
# Check "Thus ends" markers in chapters
|
||
chapters_with_end = 0
|
||
for ch in chapters:
|
||
has_end = any("Thus end" in p for p in ch["paragraphs"])
|
||
if has_end:
|
||
chapters_with_end += 1
|
||
print(f" ℹ {chapters_with_end}/{len(chapters)} chapters have 'Thus ends' markers")
|
||
|
||
# Check empty chapters
|
||
empty_chapters = [ch["chapter"] for ch in chapters if len(ch["paragraphs"]) == 0]
|
||
if empty_chapters:
|
||
print(f" ⚠ Empty chapters (0 paragraphs): {empty_chapters}")
|
||
else:
|
||
print(f" ✅ No empty chapters")
|
||
|
||
# Check ScaGoudy decoding — look for common encoded characters that should have been decoded
|
||
# Note: ñ is excluded because it's a valid IAST output character (ï→ñ mapping)
|
||
encoded_chars = set("äåöëòçüéàùè")
|
||
suspect_count = 0
|
||
for ch in chapters:
|
||
for p in ch["paragraphs"]:
|
||
# Only flag if the encoded chars appear outside HTML tags
|
||
plain = re.sub(r"<[^>]+>", "", p)
|
||
for c in encoded_chars:
|
||
if c in plain:
|
||
# Many of these are legitimate uses (e.g., German loanwords)
|
||
# but in this context they almost always indicate ScaGoudy encoding
|
||
suspect_count += 1
|
||
break
|
||
if suspect_count == 0:
|
||
print(f" ✅ No suspected undecoded ScaGoudy characters found")
|
||
else:
|
||
print(
|
||
f" ℹ {suspect_count} paragraphs contain characters that may be undecoded ScaGoudy"
|
||
)
|
||
|
||
# Check part assignments
|
||
all_part_chapters = set()
|
||
for p in parts:
|
||
all_part_chapters.update(p["chapters"])
|
||
unassigned = [
|
||
ch["chapter"] for ch in chapters if ch["chapter"] not in all_part_chapters
|
||
]
|
||
if unassigned:
|
||
print(f" ⚠ Chapters not assigned to any part: {unassigned}")
|
||
else:
|
||
print(f" ✅ All chapters assigned to parts")
|
||
|
||
# Verify ScaGoudy decoding with a spot check on known text
|
||
print(f"\n--- ScaGoudy Decoding Spot Check ---")
|
||
# Chapter 1 should contain "Kṛṣṇa" (encoded as "Kåñëa")
|
||
if chapters:
|
||
ch1_text = " ".join(chapters[0]["paragraphs"][:5])
|
||
if "Kṛṣṇa" in ch1_text:
|
||
print(f" ✅ 'Kṛṣṇa' correctly decoded in Chapter 1")
|
||
elif "Kåñëa" in ch1_text:
|
||
print(f" ⚠ 'Kåñëa' still encoded — ScaGoudy decoding may have failed!")
|
||
else:
|
||
print(
|
||
f" ℹ 'Kṛṣṇa' not found in Chapter 1 first 5 paragraphs (may be normal)"
|
||
)
|
||
|
||
# Check for "Śrīmad-Bhāgavatam" (encoded as "Çrémad-Bhägavatam")
|
||
if "Śrīmad-Bhāgavatam" in ch1_text:
|
||
print(f" ✅ 'Śrīmad-Bhāgavatam' correctly decoded")
|
||
elif "Çrémad-Bhägavatam" in ch1_text:
|
||
print(f" ⚠ ScaGoudy decoding incomplete for 'Çrémad-Bhägavatam'")
|
||
|
||
print(f"\nDone! ✨")
|
||
|
||
|
||
if __name__ == "__main__":
|
||
main()
|