# Suktas — sources & provenance (internal)

> Internal provenance record for the `suktas/` module (Phase 15 — Vedic
> Suktas). **Not** shipped in any user-facing field. The only user-facing
> "source" allowed is the scriptural locus (`vedicSource`, e.g. "Ṛgveda Khila")
> — the canonical origin of the hymn, which is part of the content, not a
> digitization source. Digitization sources live only here.

## Conventions for this module

- **Accented (sasvara) Devanagari is the primary surface.** Every sūkta is
  shipped with the svara marks (udātta `॑` U+0951, anudātta `॒` U+0952, the
  vedic tone marks `᳚`/`ꣳ`) carried inline in `devanagari`, verbatim from a
  published accented edition. This is what makes the text usable for accurate
  Vedic recitation, the explicit goal of Phase 15.
- **IAST is best-effort, NOT round-trip-asserted.** sanscript cannot round-trip
  the combining accent marks, so `iast` is generated by **stripping the svara
  marks first** and transliterating the base text — a clean pronunciation aid
  that carries no accent notation. This is the same posture the Awadhi stotras
  and the Sundarkand romanization are held to. The audit asserts IAST
  *presence* (and the `ṃ`-anusvāra / no-`ṁ` convention), **not** a Dev→IAST→Dev
  round-trip, for sūktas.
- **Meaning glosses are researched sense-renderings, shipped `draft` /
  `reviewPending`.** Vedic Sanskrit interpretation is genuinely contested
  (Sāyaṇa vs. modern philology), so the en/hi gloss states the established,
  widely-accepted sense and is flagged for a human Sanskritist pass. Never a
  copied translation.
- **≥2 independent sources.** The accented edition (primary) is cross-checked
  against an independent digitization with an accent-insensitive token-
  containment scan (the same skeleton check the stotras use).

---

## Śrī Sūktam (`sri-suktam`) — shipped 2026-06-01 (`draft`)

- **Recension / locus.** The sixteen ṛcas of the Śrī Sūkta, a *khila*
  (apocryphal appendix) to the fifth maṇḍala of the Ṛgveda. Devatā: Śrī /
  Lakṣmī; the invocation is addressed to Agni (Jātavedas) to *bring* her.
- **Primary source (accented):** Sanskrit Documents `doc_veda` accented
  Śrī Sūktam edition (`s-sukta-accent.html`, "shrIsūktam … sasvaram"),
  proofread Devanagari with full svara marks.
- **Independent 2nd source:** stotranidhi.com Śrī Sūktam (Sanskrit). Accent-
  insensitive token-containment scan = **97%** across the 16 ṛcas (≥5-char
  words; v9 88%, v16 67% — the low spots are sandhi/anusvāra digitization
  differences in the phalaśruti `यः शुचिः…`, not missing text; every ṛca is
  present in both). Well above the 91–96% bar the stotras shipped at.
- **Scope / editorial.** Shipped = the **16 numbered ṛcas** (॥१॥–॥१६॥), the
  hymn proper. The accented source page also carries a large appendix of
  commonly-recited but non-khila material — additional dhyāna ślokas
  (`पद्मानने…`, `अश्वदायी…`, `या सा पद्मासनस्था…`, `लक्ष्मीं क्षीरसमुद्र…`),
  the `सर्वमङ्गलमाङ्गल्ये` and `नारायणि नमोऽस्तु ते` verses, the Mahālakṣmī /
  Lakṣmī Gāyatrī (`महादेव्यै च विद्महे…`), and a closing śānti. These are
  **intentionally not shipped** (the same "ship the canonical core, document
  the appendix" posture as the VSN nyāsa-drop and the stotra appendix-drops).
  The `about` notes their existence.
- **ṛca count note.** v16 (the phalaśruti) refers to reciting *pañcadaśarcam*
  ("the fifteen ṛcas") — the well-known 15-vs-16 counting question (whether the
  phalaśruti itself is counted). Shipped as 16 verses (15 invocatory + the
  phalaśruti), the standard pārāyaṇa form.
- **IAST.** Accent-stripped sanscript output; anusvāra normalised to `ṃ`
  (no `ṁ`). The Vedic nasal `ꣳ` in v9 (`ईश्वरीꣳ`) is rendered `ṃ`.
- **Open debt.** Human Sanskritist / printed accented-edition (e.g. a Gita
  Press *Ṛgvedokta Devī Sūktam* or a Taittirīya pārāyaṇa book) verification of
  both the accent placement and the meaning glosses. Rows ship `draft`.

---

## Batch B — shipped 2026-06-01 (`draft`)

All four authored from the Sanskrit Documents `doc_veda` accented (sasvara)
editions (primary), cross-checked against the **stotranidhi.com** accented
editions (independent 2nd), with the same accent-insensitive token-containment
scan. IAST = accent-stripped sanscript, anusvāra `ṃ`. All `reviewPending`.

- **Puruṣa Sūktam (`purusha-suktam`)** — locus Ṛgveda 10.90 + the uttara-
  nārāyaṇa as recited in the Taittirīya Āraṇyaka. Primary `puruSukta.html`;
  2nd stotranidhi (**90%** token containment). Shipped = **18 pūrva-nārāyaṇa
  mantras + 6 uttara-nārāyaṇa mantras** across 2 sections. `deity: 'vishnu'`
  (the Puruṣa = Nārāyaṇa in the recited form). Editorial: the framing Taittirīya
  śānti (`तच्छं योः…`) and the leading stray `ॐ ।` before mantra 1 were dropped;
  mantra 1 opens `ॐ सहस्रशीर्षा`.
- **Nārāyaṇa Sūktam (`narayana-suktam`)** — Taittirīya Āraṇyaka 10.13
  (Mahānārāyaṇa Upaniṣad). Primary `narayanasukta.html`; 2nd stotranidhi
  (**95%**). `deity: 'vishnu'`. **No per-mantra `॥N॥` numbering** in either
  source — both lay the hymn out in the *same* pada-line units, so it is
  shipped **line-by-line (29 segments)**, the segmentation both sources agree
  on (not an imposed division). Closes with the Nārāyaṇa Gāyatrī.
- **Durgā Sūktam (`durga-suktam`)** — Taittirīya Āraṇyaka 4.10.2 (the Agni–
  Durgā "jātavedase sunavāma" mantras). Primary `durgAsUkta.html` (which also
  carries a swararahit copy, an in-source sanity check); 2nd stotranidhi
  (**89%**). `deity: 'durga'`. Shipped = **7 mantras + the closing Kātyāyanī
  (Durgā) Gāyatrī** (8 verses). Framing śānti dropped.
- **Mantra Puṣpam (`mantra-pushpam`)** — Taittirīya Āraṇyaka 1.22. Primary
  `mantrapushpa.html`; 2nd stotranidhi (**100%**). **Universal — no deity
  association** (the "flower of mantras" recited at the end of any pūjā).
  Shipped = the **8 āyatana mantras + the Rājādhirāja (Kubera) salutation**
  (9 verses). Framing śānti (`भद्रं कर्णेभिः…`) dropped.

## Batch C — shipped 2026-06-01 (`draft`) — Śrī Rudram

**Śrī Rudram (`sri-rudram`)** — the Rudrapraśna / Śatarudrīya, the central
Kṛṣṇa-Yajurvedīya litany to Rudra-Śiva. `deity: 'shiva'`. Locus: **Taittirīya
Saṃhitā 4.5** (Namakam) **& 4.7** (Chamakam) — the 4.7 locus is confirmed by
the source colophon (`इति श्री कृष्णयजुर्वेदीय तैत्तिरीय संहितायां चतुर्थकाण्डे
सप्तमः प्रपाठकः`).

- **Primary source (accented):** Sanskrit Documents `doc_veda` sasvara editions
  — `rudra.html` (rudraprashnaH) for the Namakam, `chamaka.html`
  (chamakaprashnaH) for the Chamakam. Full svara marks throughout.
- **Independent 2nd source:** stotranidhi.com **Devanagari** pages
  (`/hi/sri-rudra-prashna-namakam-in-sanskrit/`, `/hi/sri-rudra-prashna-chamakam-in-sanskrit/`
  — the site's default Namakam/Chamakam slugs serve Telugu/Kannada; the
  Devanagari is under the `/hi/…-in-sanskrit/` path). Accent+visarga-insensitive
  token-containment scan (≥5-char words), per anuvāka:
  - **Namakam:** 92, 96, 90, 84, 100, 100, 100, 100, 94, 97, 91 % (avg ~95%);
    closing 94%.
  - **Chamakam:** 100, 93, 100, 97, 100, 100, 100, 100, 100, 100, 100 % (avg
    ~99%); closing 80% (a short 20-token block — `इडा देवहूः…` — mostly
    sandhi-joined long words, not missing text).
  All at/above the 89–100% bar Batches A/B shipped at.
- **Structure (Ishank's decision).** **One `sukta`, 24 sections**, one anuvāka
  per section: Namakam Anuvāka 1–11 (sections 1–11), **Namakam concluding
  mantras** (section 12), Chamakam Anuvāka 1–11 (sections 13–23), **Chamakam
  concluding mantra** (section 24). Each section holds the full accented anuvāka
  as a single verse with a per-anuvāka en/hi gloss. The two sources do **not**
  agree on within-anuvāka mantra sub-segmentation (SD lays each anuvāka out as a
  flowing paragraph; stotranidhi breaks only the ṛc anuvākas 1/10/11 into
  lines), so the anuvāka — the unit both sources delimit unambiguously via the
  `॥N॥` markers — is the verse grain (the same "ship the units both sources
  agree on" posture as Nārāyaṇa Sūktam's line-segments).
- **Scope / editorial.**
  - **Preamble dropped.** The long ritual preamble on the Namakam page
    (laghunyāsa, the gaṇapati invocation, the śivopāsana mantras, the
    karanyāsa / aṅganyāsa, the dhyāna ślokas) is **not shipped** — the text
    begins at the Namakam proper (`ॐ नमो भगवते रुद्राय ॥ ॐ नमस्ते रुद्र मन्यवे…`),
    the same nyāsa-drop posture as the VSN.
  - **Concluding mantras shipped (Ishank's decision).** The recitation portions
    after each `॥11॥` **are** shipped: for the Namakam, the Tryambaka
    (Mahāmṛtyuñjaya) + `यो रुद्रो अग्नौ…` + `तमु ष्टुहि…` + `अयं मे हस्तो…` +
    `ये ते सहस्रम्…` + `मृत्युर्मे पाहि` + śānti (section 12); for the Chamakam,
    the `इडा देवहूः…` prayer + śānti (section 24).
  - **A1/A2 boundary.** SD (primary) places the transitional namaskāra
    `नमस्ते अस्तु भगवन् विश्वेश्वराय…सदाशिवाय नमः` at the **start of Namakam
    Anuvāka 2**; stotranidhi groups it as the **close of Anuvāka 1** (marks
    `॥१॥` after it). Identical content — only the anuvāka label differs; we
    follow the SD (primary) grouping.
  - **`शम्भवे नमः ।` (Namakam A2 opener).** Present **verbatim in the primary
    SD source** at the head of Anuvāka 2, but **absent in stotranidhi** and not
    part of the canonical A2 opening — a single-source reading. **Kept verbatim
    from the primary source** (Ishank's decision), flagged here for the human
    pass.
  - **Stray Latin comma** in Namakam A1 (`तयाऽस्मान्,`) — a digitization
    artifact in the primary source — was stripped.
- **IAST.** Accent-stripped (svara + Vedic-extension marks removed, `ꣳ`/`ग्ं`
  guttural-nasal → anusvāra before transliteration) sanscript output; anusvāra
  normalised to `ṃ` (no `ṁ`). Presence-asserted, not round-tripped.
- **Gloss posture & open debt.** Per-anuvāka researched sense-renderings of the
  established meaning (`draft`, `reviewPending`). The **dense litany anuvākas
  are glossed at the thematic grain**, not phrase-by-phrase, and these are the
  soft spots for the human Sanskritist pass:
  - **Namakam Anuvāka 3, 4, 9** — long `namaḥ`-litanies of fierce/marginal
    beings, artisans-and-tools, and host/spirit names; many of the names are
    obscure and variously interpreted.
  - **Chamakam Anuvāka 7, 8** — the soma-ritual implements and the yajña
    furnishings; the technical ritual vocabulary is specialised.
  Human Sanskritist / printed accented-edition verification of the accent
  placement, the `शम्भवे नमः` and A1/A2-boundary calls, and all 24 glosses is
  owed; rows ship `draft`.