jung/AGENTS.md

# Jung Reading List → Russian Editions: Research Notes

## Goal

For every book in the ISAP Zurich reading list (https://isapzurich.com/en/library/reading-list),
find the **Russian edition(s)**: correct Russian title, publisher, year, translator, ISBN,
libgen download page (if exists), and РГБ (RSL) bibliographic record (if exists).

We work **section by section**. Current: **Section 04 — Pictures** (`sections/04-pictures/`).
Sections 1–3, 5–7 are skeletons for later.

## Repository layout

```
AGENTS.md            # this file — plan, conventions, source status
CONTEXT.md           # misc findings that fit no other category
tools/               # search helpers (lg.py, rsl.py, sx.sh)
authors/             # one dossier per author (created on demand)
sections/0N-<name>/  # one .md per BOOK (chapter-level list items are merged into the book file)
```

## Conventions

- **One file per book**, named `NN-<slug>.md` (NN = order in the reading list).
- **Status marker** at top of each book file:
  - `⬜ not searched yet`
  - `🔎 searching in progress`
  - `✅ RU edition(s) found`
  - `🔶 partial` (e.g. only an essay/chapter available in RU)
  - `❌ no RU edition found (verified)`
- **Book file contents:** human-readable editions table + raw **RSL JSON** block +
  libgen edition URLs (`https://libgen.vg/edition.php?id=...` — contains download links;
  direct file links rotate, don't store them) + notes (identity check, RU↔EN title pairing source).
- **Author dossiers** (`authors/<lastname>-<firstname>.md`) hold: EN identity
  (3–5 flagship EN titles, dates), RU surname candidates (all plausible transliterations,
  German names often have several: Нойманн/Нойман/Нейманн), verified RU works with EN mapping,
  homonym warnings.
- **Commit per completed book** (or when meaningful author info is gained), message like:
  `sec04: <book> — <finding>`.
- Update `CONTEXT.md` for anything useful that fits no category.

## Sources and their status (verified 2026)

### 1. РГБ (Russian State Library) via libgen.vg biblioservice — PRIMARY for existence + editions
- Tool: `tools/rsl.py "Full Name"` (JSON, polite, 3s between pages)
- Raw URL: `https://libgen.vg/biblioservice.php?value=<query>&type=rsl&format=json`
- Rich records: title parts, **author with life dates** `[1905-1960]` (great identity anchor),
  translator, publisher, city, year, pages, ISBN, series, UDC tags, even TOC field.
- **Limitations:** no pagination (fixed first page) → always query by **full first+last name**
  for people, or 2–3 distinctive words. Loose matching (searches author AND title fields).
- Other `type=` values on this mirror are **dead**: worldcat, googlebooks, isbndb, udc
  (all return "Nothing found" even for control queries). `type=isbn` decodes ISBN structure:
  checksum error + **registered publisher group name** (e.g. 5-519 → «Клуб Касталия»,
  5-98712 → «Петроглиф») — cheap, unlimited; used by `tools/isbnval.py` as the structure layer.

### 2. libgen.vg full-text search — PRIMARY for readable editions
- Tool: `tools/lg.py "query"` (whole words only, res=100, prints pagination hint)
- Raw URL: `https://libgen.vg/index.php?req=<query>&res=100`
- **Requires a browser User-Agent** (otherwise nginx default page / robot block).
- **Search semantics (verified):** whole-word, case-insensitive. Stems do NOT work reliably
  (`происхожден` → 0 hits; full `Происхождение` → many). Multi-word = AND.
  → Use **full inflected words** (nominative, as stored in titles) and 2–3 distinctive words.
- Author search works (`req=Нойманн`) but pulls homonyms — filter by first name in results.
- Pagination: `&page=N` works.
- Edition page: `https://libgen.vg/edition.php?id=<id>`.

### 3. cogito-shop.com — current Russian trade (Jungian/psychoanalysis specialists)
- URL: `https://cogito-shop.com/search/?q=<query>` (param is **q**, not query)
- Matches author AND title. Good for "is it sellable now in RU".

### 4. SearXNG (local) — cross-verification
- Tool: `tools/sx.sh "query"` (JSON API at http://localhost:8888; the web_search tool blocks localhost)
- Use to verify RU↔EN title pairing in a third party's words, resolve transliterations.
- Keep queries ≤ 3–4 words; «…» quotes only on distinctive RU titles. Unstable relevance —
  drop junk queries, don't retry blindly.

### 5. ru.wikipedia / en.wikipedia — author dossiers
- ru-wiki articles often have «…на русском» sections listing actual RU titles.
- en-wiki for the EN bibliography (identity anchor: 3–5 flagship titles).

### 6. livelib.ru author pages — all RU works of a known author
- e.g. `livelib.ru/author/<id>-<slug>` lists every RU edition (server-rendered, scrapable).
- Author ID: find via SearXNG `site:livelib.ru <author>`. The /search page is JS-only (nope).
- **Publisher pages** `/publisher/<id>-<slug>` = the publisher's full book list (server-rendered,
  book links `/book/<id>-<slug>` carry title + author slugs) — cheap reverse publisher sweep.
  Example: /publisher/2076-medkov-s-b (Медков С. Б. — small esoteric press w/ Jung line).

### 7. cogito-shop person pages — shop stock per author
- `cogito-shop.com/person/<first>_<last>/` (transliterated, first-name first,
  e.g. `/person/mariya_luiza_fon_frants/`, `/person/yaffe_aniela/`).
- Slug discoverable from the shop's search page (results contain person links) or guessable.
- Lists every edition the shop carries (title variants!). Product pages lack full
  metadata in HTML (JS-rendered) — use RSL/livelib for publisher/year/ISBN.

### 8. GNB SPb (State Public Library of St. Petersburg) catalog — SECONDARY bibliographic source
- URL: `https://www.gbs.spb.ru/ru/search/detail/?id=<hash>` (rich GOST records: title,
  responsible parties incl. translators, publisher w/ full legal names, year, pages, ISBN,
  notes like «Др. кн. авт.»)
- **Complements RSL:** RSL (RGB) misses some small-press editions (e.g. Pattis-Zhoya
  «Аборты…» Т8/ЦГИ 2017 is in GBS but NOT in RSL). Query via SearXNG `site:gbs.spb.ru <author>`.

### 9. chitai-gorod.ru / ozon / labirint — trade retail pages
- chitai-gorod author pages list RU works (some noise); product pages give publisher/year/ISBN.
- labirint.ru/books/<id> server-rendered w/ full specs (publisher, year, pages, ISBN, translator).
- Ozon product pages: blocked from curl (redirect loop) — fetch_content also fails; use browser.

#### 10. flibusta.su — big e-book library, server-rendered search
- Search: `https://flibusta.su/booksearch/?ask=<query>` (param is **ask**, not query!)
  HTML parsing: links `/book/<id>-<slug>` and `/author/<id>-<slug>`.
- API endpoints are blocked on .su (api/search.php → 403, /jsonrpc → 404); flibusta.is unreachable.
- Good for negative verification: author search returns all same-name authors (e.g. «Нойманн»
  → 5 namesakes, no Эрих; «Амманн» → Кит/Даниэль, no Рут; «Паттис» → only С. Паттисон).
- Verified 2026-07-09: no «Выявляющий образ», no «Число и время» (Franz), no «Сэндплей» titles,
  no «Тайный мир рисунка» (Furth), no «Тест Дерево», no «Искусство и творческое бессознательное».

### 11. Open Library — the "English RSL" (author works lists, full names, ISBNs)
- Tool: `tools/ol.py "author:Lastname, First"` or `tools/ol.py 'title:"Exact EN Title"'`
- `openlibrary.org/search.json?q=...&fields=title,author_name,publish_year,language,publisher,isbn`
- No key, no rate limits observed (1s between calls is polite). Aggregates all editions of a
  work into one doc. **Best use: resolve full EN author names from initial-only list entries**
  (`title:"..."` → author_name) and get the EN works list + ISBNs to feed Google Books.
- **RU coverage is sparse but REAL** (romanized titles, e.g. «Chelovek i ego simvoly»); the
  `search.json?isbn=` facet works for RU ISBNs too. Don't use a `language=rus` filter — it
  under-reports. RU hunting still mainly RSL/libgen/flibusta/GBS SPb/Google Books.

### 12. Wikidata — identity anchors (life dates + OFFICIAL RU name)
- Tool: `tools/wd.py "Full Name"`
- `wbsearchentities` + `wbgetentities`: QID, P569/P570 (birth/death), EN/DE/RU labels, RU aliases.
  The RU label is the single most reliable spelling for RSL/libgen queries.
- Only useful for Wikipedia-scale names (Jung, Neumann, Kandinsky, Jaffé, …); niche authors
  are absent — fall back to OL title search or publisher pages.

### 13. archive.org — EN original texts (content verification)
- `archive.org/advancedsearch.php?q=title:("...")&output=json` (metadata); direct file downloads
  work (`/download/<item>/<file>`). OCR `_text.pdf` variants → pdftotext.
- Used to verify the 4-essay match for Neumann «Сборник статей» (item in.ernet.dli.2015.139225).

### 14. Google Books — book page via curl OR fetch_content (API is 429 without key)
- `books.google.com/books?vid=ISBN<13digits>`: **plain curl works** — 200 + `<title>"T - A -
  Google Книги"` if indexed, 404 if not. Good ISBN validator incl. RU (АСТ/Эксмо/БукСМарт hit;
  small-press 404). googleapis.com/books API stays 429 without a key.
- **Keep volume LOW**: Google may block agents on sustained querying (user warning 2026-07-15).
  Budget: a few dozen requests per session, 2–5s random pauses, never in tight loops; treat as
  an auxiliary cross-check, not a bulk source. If 429/403-redirect walls appear — stop and fall
  back to OL/libgen/RSL.
- `fetch_content` on the same URL → richer metadata: full author/editor names, edition, series.
  Resolved Ottmann=Klaus, Goldstein=Ralph, Brutsche=Paul, Elder=George, Moon=Beverly,
  Killick=Katherine, Pennington/Staples, Rowland=Susan, Bolander=Karen, Acton=Mary via this or OL.

### 15. Shop directory (HSE bookshelf34 pattern) — where RU psychoanalytic books are sold
- HSE «Книжный шкаф» (hse.ru/ma/therapy/bookshelfNN) = a DIRECTORY of shops, not a book list:
  NikBook (nikbook.ru — WBS shop, `?q=` search is fuzzy/works), Когито-Центр, **Скифия**
  (skifiabook.ru — publisher/shop, JS catalog, has a full price list download), marketplaces.
- NikBook Jungian finds (2026-07-09): Тоцци «Активное воображение в теории, практике и обучении»,
  Конгер «Юнг и Райх. Тело как Тень», Эдингер «Библия и психе».
- beta2alpha.ru — RU publisher (Tilda site; sells via Ozon seller beta-2-alpha); Jungian stock TBC.

### 16. Publisher sites — RU full texts and bibliographies
- hi-human.org (Living Human Heritage RU) — publishes book forewords/sections as pages (source
  of the Neumann EN-works list with RU titles).
- castalia.ru / castaliasilvasacra.ru (Клуб Касталия) — author articles + shop.
- daimon-verlag.ch / chironpublications.com / spring-publications.com — EN Jungian presses
  (spring-publications unreachable from this box; use Google Books ISBN instead).

### 17. alib.ru — RU/KZ book marketplace, 3.2M listings (used + new, incl. small press)
- Tool: `tools/alib.py "query"` (stem search! matches all inflections — unlike libgen).
- Endpoint: `https://www.alib.ru/find3.php4?tfind=<query>`; **query and pages are CP1251**
  (utf8 → mojibake). Phrase mode: double quotes. Also: author-first, year range, ISBN digits,
  price range. `>Купить<` links = listing count.
- Value: catches small-press / out-of-print editions shops don't carry (found: Furth 2nd ed.
  2014, Turner handbook 2015, Neumann КДУ/Маниф/Питер editions). Seller titles can be wrong —
  verify against RSL before trusting (e.g. «Челокес и миф» Kastaalia 2018 — not in castalia.ru
  catalog → seller error).
- Also: alib.top (Ukraine), 33ob.ru (vinyl), amarka.ru (stamps) — sister sites.

### 18. Local CRW renderer (fastcrw.com, Firecrawl-compatible, localhost:3000)
- Tool: `tools/crw.py <url> [--html] [--wait N ms]` — POST /v1/scrape {"url","formats":["markdown"]}
- Use for JS-heavy sites curl can't read (soznanie.ast-academy.ru festival site rendered fine).
- Limits: Ozon and chitai-gorod search are API/anti-bot protected — CRW returns nothing.
- Insales shops (castalia.ru) don't need it: product data is server-side JSON-LD (curl OK).

### 19. NLR / РНБ (National Library of Russia) via Primo (primo.nlr.ru) + CRW
- Tool: `tools/nlr.py "query"` — free-text search (all fields, words ANDed), page 1 (20 rows).
- Search is client-rendered → the tool renders the `search.do?fn=search&ct=search&
  vl(freeText0)=QUERY&vid=07NLR_VU1&mode=Basic&initialSearch=true` URL via CRW and parses
  title/author/year/holding (+ doc IDs when the linked layout renders).
- Full record (GOST description, incl. translators): CRW on
  `display.do?tabs=detailsTab&ct=display&fn=search&doc=<07NLR_LMS#########>&displayMode=full&vid=07NLR_VU1`
  (plain curl gives only the shell — details tab is XHR).
- Value: complements RSL — own St. Petersburg holdings + GOST records; author rows carry
  life dates and NLR auth IDs. Count line parsing is flaky — trust the parsed row list.
- Verified 2026-07-09: all 35 section-04 surnames swept (see sections/04-pictures/SWEEP-R7.md).

### 20. RSL direct (aleph.rsl.ru, Ex Libris Aleph) — PARTIAL
- find-a flow works: GET /F/-?func=file&file_name=find-a → parse action URL (session token)
  → POST func=find-b&find_code=WAU&request=<name> → results 10/page with pagination.
- Full record: follow the result row's `full-set-set` link (session-dependent); gives
  translator, series, ISBN, UDC (verified: Нойманн «Амур и Психея» МИФ 2024, ISBN
  978-5-00214-519-5, пер. М. Виноградова).
- Quirk: find_code WTI (title) and SYS (ISBN) return EMPTY 0-byte responses; only WAU
  (author) reliably works so far. libgen proxy (tools/rsl.py) still covers title/ISBN queries.
- search.rsl.ru — same catalog, Yii app, no JSON API found (2026-07-09).

### 21. ISBN validation pipeline — `tools/isbnval.py` (2026-07-15)
- Per ISBN: (1) libgen.vg biblioservice `type=isbn` — checksum + registered publisher group;
  (2) Open Library `search.json?isbn=` — title/author/year; (3) Google Books `?vid=ISBN` curl
  — title tag, 200/404; (4) NLR free-text for RU misses. Random pauses 2–5s, resumable.
- Verdicts: VALID (found w/ matching title) / STRUCT-ONLY (real ISBN, prefix registered,
  not indexed — small-press RU) / BAD-STRUCT. Output TSV: `data/isbn-validation.tsv`.
- Found in sec04: 2 bad-check-digit typos (5-89613-003-6, 5-263-00368-2), 2 publisher-prefix
  mismatches (Ленанд vs URSS; Академический проект vs Азбука-Аттикус), 2 co-publishing
  prefixes (Петроглиф for ЦГИ 2020; Пальмира for Касталия 2025).
- **triumph.ru** (РГБ ISBN search, user-provided): `https://www.triumph.ru/html/serv/find-isbn.php?isbn=<13digits>`
  (relative to the poisk-isbn.html page!) → returns links to РГБ/МГУ/РНБ/БЕН РАН/ГПНТБ catalogs
  pre-filled with the ISBN. Itself no data; МГУ (nbmgu.ru) results are JS-rendered, БЕН РАН
  (Koha) is scrapable but small collection. NLR absence ≠ ISBN invalid.
- **isbnsearch.org**: nice per-ISBN pages (author/publisher/year) but rate-walls after ~10
  requests ("Please Verify to Continue", no recovery in 7 min) — not for batches.

### 22. MGU Scientific Library (nbmgu.ru) — server-rendered, curl-able (2026-07-15)
- Tool: `tools/mgu.py "query" [FIELD] [pages] [method]` or `--multi "AUT:X" "TIT:Y"`.
  The "JS" advanced search is a plain GET: `/search/?adv=1&q1=<q>&f1=<F>&v1=<M>&cat=BOOK[&p=N]`.
- Fields: ANY/AUT/COA/TIT/KEY/RUB/YEA/PLA/PUB/SER/ISB/ISS/NBM; rows AND between q1..q3.
  Method v: 0=Слова (stemming, default) · 1=Словосочетание · 2=Начинается с · 3=Дословно.
- "Всего: N" + GOST row (title/authors/notes) + "City : Publisher, Year" + shelf code + uid link.
  20 rows/page, p=0-based; **out-of-range p silently falls back to page 0** (dedupe by uid!).
- **ISB field NOT populated** (0 hits, 3 formats × control ISBNs) — no ISBN search; use
  AUT/TIT/SER/PUB. `storing.aspx?uid=` page = holdings/order only (no full record).
- Value: second RU catalog (complements RSL/РНБ), strong **SER** sweep (e.g. «Библиотека
  аналитической психологии» → full series list in one query), GOST rows w/ translators.
- Anchor titles in result rows contain a raw `>` (title="Хранение<br/>Заказ") — regexes must
  not use [^>]+ across the anchor attrs.

## Not usable / low value
- libgen biblioservice worldcat/googlebooks/isbndb/udc (broken on this mirror)
- imaton.com (МААП educational publisher, not translations)
- bookmate.com (403 bot protection), gtmarket.ru (JS app, no HTML content)
- other libgen mirrors (libgen.lol → error page; .vg works with browser UA)

## Identity check protocol (MANDATORY before trusting a hit)

1. **Life dates** in RSL author field (e.g. `Нойманн, Эрих [1905-1960]`).
2. **EN↔RU cross-mapping:** the found person's *other* RU books must map to known EN titles
   (e.g. found «Страх феминного» + «Любовь и Душа» → EN *Feminine* + *Love and its Opposites*
   → same Erich Neumann).
3. **Co-publisher signature:** known translation pipelines (Vakler/Рефл-бук 1990s,
   Касталия, Прайм-Еврознак, АСТ/ЭКСМО, Питер «Мастера психологии»).
4. **Content spot-check** of the actual target title (open edition page / TOC) when equating
   an RU title with an EN title that sounds different
   (e.g. *The Archetypal World of Henry Moore* ⇄ «Искусство и творческое бессознательное» — verify it is about Moore).
5. **Homonym rejection:** different first name / field / era → reject or mark ambiguous
   (trap found: Верена Каст ≠ Вернер Каст; «Каст» search returns Verena Kast books).

## Workflow (per section) — v2 (2026-07-09)

**Phase 0 — Full names + identity** (`authors/`):
1. Resolve every initial-only author to a full EN name: `tools/ol.py 'title:"EN Title"'`
   (author_name) → else `tools/wd.py` (QID + RU label) → else Google Books ISBN page →
   else publisher/review page. Record in `sections/<n>/AUTHORS-EN.md` (EN full name, RU name
   to query, name source).
2. For big names: `wd.py` gives life dates + the OFFICIAL RU label (best RSL spelling).
3. Build the transliteration matrix per author (2–3 variants: Нойманн/Нойман, Кох/Коч,
   Калфф/Кальфф, Ферс/Фёрт) — run every query with ALL variants.

**Phase 1 — Author sweep (EN→RU):** per author: `rsl.py "<First> <Last>"` → `lg.py "<Last>"`
→ `cogito q=<Last>` → `flibusta ?ask=<Last>` → livelib/chitai-gorod author page. Apply identity
check. Fill author dossier.

**Phase 2 — Title-gap sweep** for still-unmatched books: `rsl.py "<RU title words>"` (check
author field!), `lg.py` 2–3 whole-word title combos, cogito, flibusta by RU title, GBS SPb via
SearXNG `site:gbs.spb.ru`.

**Phase 2b — Publisher sweep (RU→EN, reverse direction):** per Jungian RU publisher/imprint
(Касталия, hi-human/Living Human Heritage, ЦГИ/Т8, Когито, Фантом Пресс, Азбука, Серебряные
нити, Питер, Ваклер, Добросвет, Деметра, НикБук, Скифия, beta2alpha, Эксмо/АСТ психология):
enumerate the full Jungian catalog (site search or RSL **series** queries: «Юнгианская
психология», «Библиотека аналитической психологии», «Классики зарубежной психологии») → map
every title to an EN original → backfill both missing items and future sections. This catches
books author queries can never hit (generic titles like «Сборник статей», reworked titles).

**Phase 2c — Community reading lists (oracles):** ISAP Zurich, HSE (hse.ru/ma/therapy),
МИСП, Jung Institute Moscow, education-psy.ru sandplay list, carljung.ru, koob.ru — these
explicitly state which books have RU editions (or lack them). Scrape each, diff vs our ❌ list.

**Phase 3 — Content verification (for suspicious candidates):** when an RU title's mapping to
an EN title is not obvious, verify 1:1 against the EN original: archive.org OCR text (TOC /
editorial note) vs RU essay/chapter list (koob/livelib/preview). Example: Neumann «Сборник
статей» = Art and the Creative Unconscious: Four Essays (4/4 essay match).

**Phase 4 — Edition collection:** ALL editions from ALL sources (title variants!), full metadata
(publisher/year/pages/translator/ISBN) — prefer RSL/GBS SPb records (GOST) over shop data.

**Phase 5 — Book notes + commits** (one commit per completed book or meaningful author gain).

Negatives gate (before ❌): author sweep (RSL + libgen + cogito + flibusta) + one title-level
SearXNG + livelib author page + one shop/retail check. Log every RU query in MISSING.md.

Special cases:
- **Jung's CW volumes:** translated and widely available — don't deep-search; note the RU
  volume for the item (CW15 = «Дух в человеке, искусстве и литературе», Харвест 2003;
  Red Book = «Красная книга (Liber Novus)», Касталия 2025, trans. О. Комков).
- **Essays/chapters** (e.g. Jaffé in *Man and his Symbols*): mark 🔶 if only the parent
  book/anthology is available in RU.

## Politeness

- ~4s between libgen requests, ~3s between RSL pages, ~2s between shop requests.
- All tools cache nothing by default; reruns are fine but avoid duplicate queries in a batch.