tts-voice-creator-clone-and-design-2/architecture/magazine-ingestion.md

SOP — Magazine ingestion

Status: P0 complete (backend). P1 (listening workflow UI) not started.

Why magazines are not books

The audiobook pipeline assumes one continuous text in reading order. A magazine breaks every part of that assumption: several unrelated articles share a page, columns run in parallel, adverts and sidebars interrupt body copy, and a feature routinely jumps twenty pages to its tail. Narrating a magazine page-by-page produces audio that walks out of an article mid-sentence and into an advert.

The canonical object is therefore the article, not the page. Everything below exists to reconstruct articles from a flat PDF.

Pipeline

Implemented in core/magazine.py; driven by routes/magazine.py.

PDF
 → extract_pages()        PyMuPDF: text blocks with bbox, font size/name, weight
 → ocr_page()             tesseract TSV fallback when a page has <120 native chars
 → classify_blocks()      title / deck / byline / body / caption / pullquote /
                          sidebar / furniture / ad / figure
 → mark_advert_pages()    spread an "ADVERTISEMENT" label across its whole page
 → mark_pullquotes()      demote display text that merely repeats body copy
 → assign_reading_order() column detection, then top-to-bottom within a column
 → score/parse_contents() Stage A: locate and parse the table of contents
 → group_articles()       Stage C: group blocks, link "continued on page N"
 → human review           Stage D, in the UI — nothing is narrated until approved

Why server-side, not in the browser

Text extraction for plain PDFs runs client-side (PDF.js), and that stays. Layout analysis does not, for three reasons: PyMuPDF exposes real font metrics and embedded-image rectangles that PDF.js does not; OCR coordinates from tesseract merge with those bboxes in one coordinate space; and the result is expensive enough (minutes for an issue) that it must be computed once and persisted, not recomputed per browser session.

PDF.js remains responsible for rendering in the PDF view.

Docling/Unstructured were considered and rejected: they pull torch plus layout model weights (multi-GB, ARM64 wheel risk on the DGX Spark, GPU contention with the TTS stack) to improve on heuristics that already work on the cases we hit. Revisit only if real issues show grouping failures the heuristics cannot fix.

Decisions that are easy to get wrong

Font baseline is document-wide, never per page. classify_blocks judges size as a ratio against base_size. A page carrying a full-bleed photograph and one caption has no body copy of its own, so a per-page baseline makes that caption its own reference and classifies it as body text. extract_pages measures the baseline once across every block in the issue.

Reading order is column-major. Sorting blocks by y interleaves parallel columns, splicing the left column's sentence into the right column's. Columns are inferred from vertical whitespace (where blocks are absent), because magazines change column count per page — a fixed grid is wrong. Blocks spanning

60% of the page width sort ahead of both columns, so a full-width headline precedes the body it introduces.

Furniture is identified by repetition, not position. A running head and a headline both sit at the top of a page. What separates them is that the running head recurs on many pages at the same edge; find_repeated_lines needs the whole document, so classification waits until every page has been read.

A page is claimed only if it contributes. An advert between two features starts no article, so without an explicit check it gets swallowed by the preceding one and highlighted as part of it in the PDF view.

Pull quotes are demoted. Narrating one reads the same sentence twice a few seconds apart. A display-sized block whose text already appears in the body copy is a pull quote, not a deck.

Confidence, and why review is mandatory

Signal	Confidence
Anchored by a contents entry	0.90
Started by a headline, no contents entry	0.55
Article whose blocks contain no body copy	×0.4

Nothing is narrated until a human sets status: "approved". This is deliberate: a five-second correction beats a 30-minute recording with a sidebar in the middle. POST …/podcast refuses unapproved articles outright.

Storage

Magazines live in the Read Aloud store and are distinguished by meta.json → kind: "magazine", so the existing library listing, cover, progress and audio endpoints keep working.

<CONFIG_DIR>/reader_library/<doc_id>/
    meta.json           kind, title, issue, ingest status, default_voice,
                        podcast hosts, per-article progress
    source.pdf          the original
    layout.json         every classified block, with bbox + reading order
    articles.json       detected articles; status gates narration
    pages/NNNN.png      rendered pages (PDF view + OCR input)
    images/*.png        extracted figures ≥64px
    podcast/<id>.json   generated host scripts
    audio/<idx>.mp3     synthesised audio (existing reader endpoints)

layout.json is the join table between audio, text, image and PDF coordinates: every block carries page + bbox, so a sentence in the reflowed view can jump to its source region on the rendered page.

Podcast mode

Per article, alongside narration — not instead of it. An approved article's narratable blocks are sent to the configured LLM (llm_url / llm_model, reusing _post_llm_chat_completion from routes/conversation.py) with a system prompt that forbids inventing facts and demands strict JSON. The parser tolerates a markdown fence and surrounding prose, and reassigns unknown host labels rather than dropping the turn — an unrecognised label would otherwise leave a line with no voice. 2–3 hosts, each bound to a voice the user picks; synthesis reuses the multi-speaker path the audiobook casting already uses.

Tests

• tests/test_magazine.py — 34 tests on the pure geometry/classification logic: column reading order, the document-baseline regression, furniture, pull quotes, contents parsing, continuation linking, advert pages.
• tests/test_magazine_routes.py — 10 tests: path traversal rejection, review gating, LLM script parsing including malformed output.

Both run without PyMuPDF: the I/O functions import it lazily so the logic worth testing needs no native dependency or sample PDF.

Verified end-to-end against a synthetic 8-page issue (contents page, two-column feature with a byline and a page-8 jump, full-page advert, second feature, reviews) and against an image-only rendering of the same file to exercise OCR.

OCR engine selection (not yet wired)

Engines → OCR & Document Parsing lists layout-aware parsers (KDL Frontier Parser nano, Baidu Unlimited-OCR) as connection-probed cards. Those are presentation and health-check only — there is no server-side dispatch, so ingestion currently always uses local tesseract via ocr_lang. Routing a page image to one of those vLLM endpoints and mapping its structured output onto Block is real work, not a config change; it is the natural home for TODO P2's per-page fallback selection and benchmark comparison.

Known limitations

• Sidebar detection (BLOCK_SIDEBAR) is defined but not yet inferred; sidebars currently classify as body and must be excluded by hand in review. P2.
• continued from page N is parsed but not used for linking — only the forward continued on page N reference drives it.
• Reviews-style pages with no headline are absorbed by the preceding article when the contents page does not list them.
• Word-level timing is not available; P1 targets sentence highlighting.