OCR + Parsing
March 6, 2026 · View on GitHub
🏥 Quick Return to Emergency Room
You are in a specialist desk.
For full triage and doctors on duty, return here:
- WFGY Global Fix Map — main Emergency Room, 300+ structured fixes
- WFGY Problem Map 1.0 — 16 reproducible failure modes
Think of this page as a sub-room.
If you want full consultation and prescriptions, go back to the Emergency Room lobby.
A hub to triage and repair noisy text inputs from scanned PDFs, images, HTML scraping, or parser drift.
Use this folder when the document looks fine to the eye but retrieval or reasoning keeps failing.
Orientation: what each page does
| Page | What it solves | Typical symptom |
|---|---|---|
| Layout, Headers, Footers | Remove noise from margins and repeated text | Answers reference “page 3 footer” instead of body |
| Tokenization & Casing | Normalize Unicode, case, and hyphens | E-mail ≠ Email, half-width/full-width mismatch |
| Tables & Columns | Preserve table schema and cell order | Numbers drift across columns |
| Images & Figures | OCR and align captions | Figure text missing or attached to wrong section |
| Scanned PDFs & Quality | Handle skewed/blurred pages | Whole sections unreadable to OCR |
| Multi-language & Fonts | Normalize mixed scripts | Chinese/English tokens split or duplicated |
When to use
- OCR tables or citations look visually correct but answers miss the right section.
- Code blocks or math collapse after parsing.
- Mixed-language documents behave inconsistently.
- Special characters or hyphen splits break tokens.
- Headers or section anchors disappear during export.
FAQ
Why does OCR “look fine” but retrieval fails?
Because tokenization and indexing see hidden breaks (Unicode variants, line merges, wrong anchors) that humans overlook.
What is the most common root cause?
Headers/footers leaking into the body and breaking ΔS alignment.
Do I need to retrain embeddings after fixing?
No — most fixes are structural (schema/normalization). Re-indexing with the same embeddings is enough.
Acceptance targets
- ΔS(question, retrieved) ≤ 0.45 for three paraphrases.
- Coverage ≥ 0.70 for the target section.
- λ_observe convergent across two seeds.
- Human audit shows no missing headers, captions, or broken tables.
Fix in 60 seconds
-
Ground-truth one page
Pick one Q/A pair and keep a screenshot baseline. -
Measure ΔS
Log ΔS(question, retrieved) and ΔS(retrieved, anchor). -
Probe λ_observe
Ask for cite-first. If citation fails but free explanation works, drift confirmed. -
Patch minimally
- Re-run OCR with line/table fences
- Normalize casing and Unicode
- Preserve anchors, math, captions
- Drop low-confidence spans and export with
{section_id, page_no, char_span}
🔗 Quick-Start Downloads (60 sec)
| Tool | Link | 3-Step Setup |
|---|---|---|
| WFGY 1.0 PDF | Engine Paper | 1️⃣ Download · 2️⃣ Upload · 3️⃣ Ask “Answer using WFGY + |
| TXT OS | TXTOS.txt | 1️⃣ Download · 2️⃣ Paste into LLM · 3️⃣ Type “hello world” — OS boots instantly |
Explore More
| Layer | Page | What it’s for |
|---|---|---|
| ⭐ Proof | WFGY Recognition Map | External citations, integrations, and ecosystem proof |
| ⚙️ Engine | WFGY 1.0 | Original PDF tension engine and early logic sketch (legacy reference) |
| ⚙️ Engine | WFGY 2.0 | Production tension kernel for RAG and agent systems |
| ⚙️ Engine | WFGY 3.0 | TXT based Singularity tension engine (131 S class set) |
| 🗺️ Map | Problem Map 1.0 | Flagship 16 problem RAG failure taxonomy and fix map |
| 🗺️ Map | Problem Map 2.0 | Global Debug Card for RAG and agent pipeline diagnosis |
| 🗺️ Map | Problem Map 3.0 | Global AI troubleshooting atlas and failure pattern map |
| 🧰 App | TXT OS | .txt semantic OS with fast bootstrap |
| 🧰 App | Blah Blah Blah | Abstract and paradox Q&A built on TXT OS |
| 🧰 App | Blur Blur Blur | Text to image generation with semantic control |
| 🏡 Onboarding | Starter Village | Guided entry point for new users |
If this repository helped, starring it improves discovery so more builders can find the docs and tools.
