| name | dns-bhat-book-summarizer |
| description | Generates chapter-wise English summaries with source pointers and Eke romanisation files for DNS Bhat Kannada linguistics books in the ettuge repository. Use this skill whenever the user asks to: summarize a DNS Bhat book, create an English summary for a book in the dnsbhat folder, generate Eke transliteration of Kannada content, add chapter pointers to an -en.md file, create a claude-prompt for a DNS Bhat book, cross-link Kannada and English versions of a book, or produce a *-kn-eke.md romanisation file. Triggers on phrases like "summarize book NN", "create English summary", "generate Eke", "add chapter pointers", "create prompt for book", or any reference to producing -en.md / -kn-eke.md files for the DNS Bhat collection.
|
DNS Bhat Book Summarizer
You produce three kinds of output for DNS Bhat linguistics books in:
/Users/vishwas/code/ettuge/src/main/md/kannada/dnsbhat/
{NN}-{slug}-kn.md — Structured Kannada source: paragraph breaks, TOC, <a id> anchors at each chapter/section. Created from the raw -book.md or -blog.md if not already present.
{NN}-{slug}-en.md — Chapter-wise English summary with cross-links to -kn.md anchors and Eke romanisation
{NN}-{slug}-kn-eke.md — Eke romanisation of the Kannada content, structured with the same section anchors as -kn.md
The -kn.md is the canonical readable version — it's what the English summary links into. The raw -book.md or -blog.md is the OCR archive and is not modified.
Step 1: Identify source material
Before writing anything, read the target book folder:
/Users/vishwas/code/ettuge/src/main/md/kannada/dnsbhat/{NN}-{slug}/
Check which files exist:
README.md — folder-level README for GitHub browsing (create if absent; see README format below)
*-book.md — full book text (highest quality, use first)
*-blog.md — blog posts (use if no book.md)
*-kn.md — structured Kannada (use for chapter structure)
*-djvu.md — DjVu OCR (fallback, may be noisy)
*-en.md — existing English summary (update if present, don't overwrite good content)
*-kn-eke.md — existing Eke file (update if present)
*-claude-prompt.md — AI prompt (separate from what we produce here)
README.md format (create if absent):
# [NN] — {Kannada title}
**{English title}**
> {One-sentence description of the book's argument and significance.}
**{✅ Fully processed / ⚠️ Partial / ❌ Not collected}** · {year} · {publisher} · {N} pages
---
## Files in This Folder
| File | Contents |
|------|----------|
| [`{NN}-{slug}-book.md`](./{NN}-{slug}-book.md) | Raw OCR text — ... |
| [`{NN}-{slug}-kn-eke.md`](./{NN}-{slug}-kn-eke.md) | Eke romanisation |
| [`{NN}-{slug}-en.md`](./{NN}-{slug}-en.md) | English summaries |
| [`{NN}-{slug}-claude-prompt.md`](./{NN}-{slug}-claude-prompt.md) | AI context primer |
---
## Where to Start
- **Don't read Kannada?** → [`{NN}-{slug}-en.md`](./{NN}-{slug}-en.md)
- **Want the phonetics?** → [`{NN}-{slug}-kn-eke.md`](./{NN}-{slug}-kn-eke.md)
- **AI context primer?** → [`{NN}-{slug}-claude-prompt.md`](./{NN}-{slug}-claude-prompt.md)
- **Full Kannada text?** → [`{NN}-{slug}-book.md`](./{NN}-{slug}-book.md)
---
[← Back to catalogue](../README.md)
Read the primary source file and *-kn.md if it exists (for chapter structure). For books with a *-claude-prompt.md, read it too — it has the chapter outline.
Step 2: Produce the structured Kannada file (*-kn.md) — if absent
If no -kn.md exists, create one from the raw -book.md or -blog.md. The -kn.md is the primary readable, linkable version — do not modify the raw source file.
Paragraph detection in -book.md OCR output:
- Paragraphs typically begin with a 4-space indent
- Lines are hard-wrapped at ~65 chars — a paragraph is multiple consecutive lines followed by a blank line or the next indented line
- Chapter headings are usually short standalone lines (often numbered or in all-caps Kannada)
- Section subheadings may have a number prefix (e.g.,
೧., ೨.) or be formatted distinctively
-kn.md structure:
# {Kannada title}
**ಲೇಖಕರು:** ಡಿ. ಎನ್. ಶಂಕರ ಬಟ್
**ಪ್ರಕಟಣೆ:** {year}, {publisher}
> ಮೂಲ ಪಠ್ಯ: [`{NN}-{slug}-book.md`](./{NN}-{slug}-book.md)
> ಇಂಗ್ಲಿಶ್ ವಿಶ್ಲೇಷಣೆ: [`{NN}-{slug}-en.md`](./{NN}-{slug}-en.md)
---
## ಒಳಪಿಡಿ
- [ಅಧ್ಯಾಯ ೧ — {title}](#ch1)
- [೧.೧ {section}](#sec-1-1)
---
<a id="ch1"></a>
# ಅಧ್ಯಾಯ ೧ — {title}
[↑ ಒಳಪಿಡಿಗೆ ಹಿಂತಿರುಗಿ](#ಒಳಪಿಡಿ)
---
<a id="sec-1-1"></a>
## ೧.೧ {section title}
{paragraph text with blank lines between paragraphs}
See book 08's -kn.md as the reference implementation.
CI-generated chapter pages (automatic — no action needed)
Once full.md (i.e. book/kn/full.md in the Phase-20 taxonomy) contains <a id="chN"> or <a id="adhyAya-N"> anchors, the GitHub CI chapter-splitter generates browsable per-chapter pages on every push:
ch0.md — preamble + TOC with (#chN) links rewritten to page-relative chN links, plus a **Chapters:** Ch 1 · Ch 2 · ... bar at the bottom. If the preamble is > 1 500 chars the bar also appears at the very top.
ch1.md … chN.md — one file per chapter; opens with > ← Ch N-1 · [Contents](ch0) · Ch N+1 → prev/next nav; if the chapter body is > 1 500 chars a full **Chapters:** Ch 1 · Ch 2 · ... quick-jump bar appears immediately below the nav.
You do not create these files — CI writes them. Your only job is to put the right anchors into full.md.
Minimum to trigger chapterization: ≥ 2 <a id="chN"> (or adhyAya-N) anchors in the body. Single-anchor files are skipped.
Step 3: Produce the Eke romanisation file (*-kn-eke.md)
The Eke file has:
- Same chapter/section structure as the Kannada source
- Section anchors (
<a id="ch1"></a>, <a id="sec-1-1"></a>) matching the Kannada file
- A table of contents (ಒಳಪಿಡಿ in Eke:
oLapiDi) with links to anchors
- Cross-links at the top to
-kn.md and -en.md
- Key passages transliterated; summaries in Eke for long sections
Header format:
# {title in Eke}
**lEkhakaru:** Di. en. Sankara bhaT
**saraNi / Source:** {series if any}
**prakaTaNe / Published:** {year and publisher}
> mUla paThya (kannaDa lipi): [{NN}-{slug}-kn.md](./{NN}-{slug}-kn.md)
> ingliS viSlEShaNe: [{NN}-{slug}-en.md](./{NN}-{slug}-en.md)
---
## oLapiDi
- [adhyAya 1 — {chapter title in Eke}](#ch1)
- [1.1 {section in Eke}](#sec-1-1)
For each section:
<a id="ch1"></a>
# adhyAya 1 — {title in Eke}
[↑ oLapiDige hintirugi](#oLapiDi)
---
<a id="sec-1-1"></a>
## 1.1 {section title in Eke}
{Eke transliteration of first 3-5 sentences of the section, or a meaningful excerpt}
For long chapters, transliterate the opening passage (establishes voice and key terms) plus any key definitions or examples.
Step 3: Produce the English summary file (*-en.md)
The English file has:
- A book overview paragraph (3–5 sentences)
- A table of contents with anchor links
- For each chapter/section: heading in English + Kannada title + cross-links + 3–6 bullet points
- A "Key Concepts" table (Kannada term → English meaning)
- A "Cross-references" section pointing to related DNS Bhat books
Header format:
# {Full English Title}
## {Subtitle or Translation of Kannada Title}
**Author:** D. N. Shankara Bhat (ಡಿ. ಎನ್. ಶಂಕರ ಬಟ್)
**Published:** {year}, {publisher}
**Language:** Kannada
**Source quality:** {Full text / Blog posts (N parts) / Partial transcript}
---
## Book Overview
{3-5 sentence overview of the book's central argument and significance}
---
## Table of Contents
- [Chapter 1 — {English Title} ({Kannada Title})](#ch1)
- [1.1 {Section English title}](#sec-1-1)
For each section:
## Chapter 1 — {English Title}
*({Kannada title})*
### 1.1 {English section title}
[ಕನ್ನಡ →](./{NN}-{slug}-kn.md#sec-1-1) | [Eke →](./{NN}-{slug}-kn-eke.md#sec-1-1)
- Bullet 1: what this section establishes or argues
- Bullet 2: key evidence or examples used
- Bullet 3: connection to the book's central thesis
- Bullet 4 (if needed): objections considered or counter-arguments
For blog-based books (like Books 02 and 18), use the blog post title as the "chapter" and link to the blog file:
### Part 4 — {English title of blog post}
[Blog (ಕನ್ನಡ) →](./{NN}-{slug}-blog.md#{anchor}) | [Eke →](./{NN}-{slug}-kn-eke.md#{anchor})
Eke Romanisation Rules
Eke is a romanisation system designed by Vishwas, inspired by Hosabaraha and Harvard-Kyoto (HK). Use Eke(ek) mode (ellara kannaDa — simplified, no aspirates).
Vowels
| Kannada | Eke | | Kannada | Eke |
|---|
| ಅ | a | | ಆ | A |
| ಇ | i | | ಈ | I |
| ಉ | u | | ಊ | U |
| ಎ | e | | ಏ | E |
| ಒ | o | | ಓ | O |
| ಐ | ay | | ಔ | av |
| ಋ / ೃ (vocalic r, short) | x | | ೠ / ೄ (vocalic r, long) | X |
| ಂ (anusvara) | nasal+C (always assimilated — never M; see Anusvara section below) | | ಃ visarga | H |
Consonants — case rule: uppercase = retroflex or special
| Kannada | Eke | | Kannada | Eke |
|---|
| ಕ | k | | ಗ | g |
| ಚ | c | | ಜ | j |
| ಟ | T | | ಡ | D |
| ತ | t | | ದ | d |
| ಪ | p | | ಬ | b |
| ಮ | m | | ನ | n |
| ಣ | N | | ಯ | y |
| ರ | r (always lowercase — R is exclusively ಱ, never ರ) | | ಲ | l |
| ಳ | L | | ವ | v |
| ಱ | R (archaic retroflex ṟ — extremely rare in modern Kannada) | | | |
| ಸ | s | | ಶ | S |
| ಹ | h | | ಷ | S (same as ಶ in EK) |
Aspirates (mahapranas) — preserved in Eke(ek)
ಖ→kh, ಘ→gh, ಛ→ch, ಝ→jh, ಠ→Th, ಢ→Dh, ಥ→th, ಧ→dh, ಫ→ph, ಭ→bh
Anusvara before plosives — write as nasal+plosive
ಂಕ→nka, ಂಗ→nga, ಂಚ→nca, ಂಜ→nja, ಂಟ→nTa, ಂಡ→nDa, ಂತ→nta, ಂದ→nda, ಂಪ→mpa, ಂಬ→mba
Inherent vowel
Every consonant has inherent a. Virama ್ suppresses it in clusters.
ಕ = ka, ಕ್ = k (in cluster), ಕ್ಕ = kka
Common word examples
| Kannada | Eke |
|---|
| ಕನ್ನಡ | kannaDa |
| ನುಡಿ | nuDi |
| ಬರಹ | baraha |
| ಹೆಸರು | hesaru |
| ಎಸಕ | esaka |
| ಅರಿಮೆ | arime |
| ಮುನ್ನೋಟ | munnOTa |
| ಇಣುಕುನೋಟ | iNukunOTa |
| ಪದ | pada |
| ಸೊಲ್ಲರಿಮೆ | sollarime |
| ಮಾತು | mAtu |
| ಉಲಿ | uli |
| ಒಳರಚನೆ | oLaracane |
| ಹೊಸ | hosa |
| ಕಟ್ಟುವ | kaTTuva |
| ಮೇಲರಿಮೆ | mElarime |
| ಕೀಳರಿಮೆ | kILarime |
Quality notes
Source quality matters — note it clearly in the English file header:
*-book.md from archive.org: high quality, full text → produce complete section-by-section summary
*-blog.md from Wayback Machine: good quality, part-by-part → produce per-post summary
- YouTube transcripts: variable quality, may be garbled → note limitations, summarize available content
- DjVu OCR: often has OCR errors — note this, cross-check with other sources
For partial content (missing blog parts, corrupted transcripts):
- Note what's missing in the overview
- Don't invent content — only summarize what's actually in the source files
- Use "⚠️ Parts X–Y not available" as section placeholders
Cross-reference footer (always include in -en.md)
At the end of every English summary, add a cross-reference table:
---
## Cross-References to Other DNS Bhat Works
| Related Book | Connection |
|---|---|
| [08 — Kannadakke Mahaprana Yake Beda](../08-kannadakke-mahaprana-yake-beda/) | {how it relates} |
| [14 — Nijakku Halegannada Vyakarana](../14-nijakku-halegannada-vyakarana-entahadu/) | {how it relates} |
See references/book-cross-references.md for the full thematic index to populate this table.
File naming conventions
| File | Naming |
|---|
| Folder README | README.md |
| English summary | {NN}-{slug}-en.md |
| Eke romanisation | {NN}-{slug}-kn-eke.md |
| Kannada structured | {NN}-{slug}-kn.md |
| Blog content | {NN}-{slug}-blog.md |
| Book full text | {NN}-{slug}-book.md |
| AI prompt | {NN}-{slug}-claude-prompt.md |
Where {NN} is the zero-padded book number and {slug} is the folder name without the number prefix.
Reference files
references/book-cross-references.md — Thematic index for cross-linking between DNS Bhat works (read when populating the cross-reference footer)