Part of the Abstract Intelligence Platform
This module is part of a unified system for transforming raw media into structured, searchable, and SEO-optimized data.
abstract_pdfs handles document ingestion and publishing:
- PDF → structured pages (text + images)
- metadata + manifest generation
- static HTML output (viewer + gallery)
Full system: https://github.com/AbstractEndeavors/abstract-intelligence
abstract_pdfs — Document Processing & SEO Pipeline for PDF-Based Content
A structured pipeline for transforming PDFs into searchable, metadata-rich, web-ready content, combining OCR, page-level analysis, metadata generation, and static site scaffolding.
Designed for:
- large PDF collections
- SEO-driven content indexing
- document-to-web publishing pipelines
- structured ingestion of unstructured media
🔹 What This System Is
abstract_pdfs is not a PDF utility — it is a full document processing pipeline:
- ingests raw PDFs
- decomposes them into pages, images, and text
- extracts and generates metadata
- enriches content via NLP APIs
- builds structured outputs (JSON + HTML)
- generates navigable web content (galleries + viewers)
The result is a fully browsable, searchable document corpus.
🔹 Pipeline Overview
PDF Input
↓
Slice / Decompose (images + text per page)
↓
OCR + Text Extraction (layout-aware engines)
↓
Metadata Generation
├─ summaries
├─ keywords
├─ descriptions
↓
Manifest Creation (per-page + per-document)
↓
HTML Generation
├─ PDF viewer pages
├─ gallery index pages
↓
Static Site Output (SEO-ready)
flowchart TD
A[PDF Input]
B[DocumentPipeline]
C[SliceManager\nPage Images + Text + OCR]
D[Per-Page Assets\nThumbnails / Text / Info JSON]
E[Manifest Generation\nPage + Document Metadata]
F[NLP Enrichment\nSummaries + Keywords + Descriptions]
G[HTML Generation\nViewer Pages + Gallery Indexes]
H[Static Output\nSearchable / SEO-ready PDF Corpus]
A --> B --> C --> D --> E --> F --> G --> H
🔹 Core Capabilities
Document Decomposition
-
Splits PDFs into:
- page images
- extracted text
- structured page directories
-
Maintains consistent directory structure for downstream processing
Metadata & SEO Enrichment
-
Generates:
- summaries
- keywords
- descriptions
-
Integrates with NLP endpoints for:
- text analysis
- keyword refinement
- summarization
Example: page-level analysis via API calls
Manifest Generation
-
Produces structured JSON per page:
- metadata
- text
- image references
- SEO fields
-
Aggregates into document-level manifests
Static Site Generation
-
Generates:
- PDF viewer pages (page-by-page navigation)
- gallery index pages (directory browsing)
-
Automatically builds:
- thumbnails
- descriptions
- keyword tags
Example: dynamic card generation for directories
Path ↔ URL Mapping
-
Converts filesystem structure into web-accessible URLs
-
Maintains consistency between:
- local storage (
/srv/media/...) - public endpoints (
/pdfs/...)
- local storage (
Content Structuring
-
Page-level:
- text
- summary
- keywords
-
Document-level:
- aggregated metadata
- full-text indexing
🔹 Architecture
The system is composed of modular components:
-
DocumentPipeline
- orchestrates ingestion → processing → output
-
SliceManager
- handles PDF decomposition and OCR
-
Manifest Generators
- build structured JSON representations
-
HTML Generators
- render viewer and gallery pages
-
Metadata Utilities
- enrich content via external NLP services
Each stage is:
- independent
- composable
- replaceable
🔹 Key Design Decisions
Page-Level First
All processing happens per-page, enabling:
- granular indexing
- targeted metadata
- scalable processing
Structured Over Raw
Outputs are always:
- JSON manifests
- structured metadata
- normalized fields
Not just raw text dumps.
SEO as a First-Class Concern
Every page includes:
- meta tags
- OpenGraph / social metadata
- keyword tagging
- canonical URLs
Filesystem as Source of Truth
- directory structure = content hierarchy
- no database required
- easily deployable as static site
🔹 Why This Exists
Traditional PDF workflows:
- store documents as opaque blobs
- lack searchability
- lack metadata
- are not web-native
abstract_pdfs transforms PDFs into:
- structured, indexable content
- web-ready assets
- searchable knowledge bases
🔹 Example Use Cases
- PDF → website publishing pipelines
- document archives (research, legal, media)
- SEO-driven content platforms
- knowledge base generation
- preprocessing for LLM / search systems
🔹 Integration Context
This system integrates with:
- OCR pipelines (layout_ocr / abstract_ocr)
- NLP systems (abstract_hugpy)
- static hosting (Nginx / CDN)
- search indexing systems
🔹 Design Philosophy
- Documents are data, not files
- Structure before presentation
- Metadata is as important as content
- Static outputs scale better than dynamic systems
Metadata
Release files for abstract-pdfs 0.0.39
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| abstract_pdfs-0.0.39.tar.gz | 118.4 kB | Without provenance | Details for abstract_pdfs-0.0.39.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| abstract_pdfs-0.0.39-py3-none-any.whl | Python 3 | none | any | Details for abstract_pdfs-0.0.39-py3-none-any.whl |
Total release size: 284.6 kB
Release files / abstract_pdfs-0.0.39.tar.gz
| Download URL | abstract_pdfs-0.0.39.tar.gz |
|---|---|
| Size | 118.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
97fe9cd6c95331fa185c24a2ca78657ba85797555b125297e8c88520da87162d
|
|
BLAKE2b-256 checksum How to use checksums |
376591d04f62cb4320f7ba29b48f8f315570ea4e3525a38e3a730ea147380a37
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.11
|
Release files / abstract_pdfs-0.0.39-py3-none-any.whl
| Download URL | abstract_pdfs-0.0.39-py3-none-any.whl |
|---|---|
| Size | 166.2 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f732a55fb77c30ef52986328431d3718fc3f6a7cb978c7ccf79e23c241c9d211
|
|
BLAKE2b-256 checksum How to use checksums |
e5fb5150260a451ede8fe51cffa8bb74a2e5df4a9e56e7e2b7aa06c4d60ae8df
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.11
|
Security for abstract-pdfs 0.0.39
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.