Compare commits

..

No commits in common. "fdb5deaa8bb8afa15f579744b0df4075022e7fd6" and "1078367d6a06f6d50502742d3f178c92c6c5c2ca" have entirely different histories.

3 changed files with 0 additions and 198 deletions

View file

@ -1,96 +0,0 @@
# text.extract
Reine Textextraktion aus PDF- und DOCX-Dokumenten. Pure-Rust,
in-WASM, ohne deklarierte Berechtigungen. Der kanonische erste
Schritt in Flows, die Rohtext brauchen — anonymisieren,
zusammenfassen, übersetzen, per LLM klassifizieren,
Audit-redigieren.
## Capability
- `text.extract@0.1.0`
## Eingaben
| Name | Typ | Beschreibung |
| ---------- | ----- | ------------------------------------------------------- |
| `document` | bytes | Das zu extrahierende Dokument. MIME-Typen siehe unten. |
Akzeptierte MIME-Typen (deklariert in
`module.yaml#accepts_mime`, von Studio als File-Picker-Filter
gelesen):
- `application/pdf`
- `application/x-pdf`
- `application/vnd.openxmlformats-officedocument.wordprocessingml.document`
Andere MIME-Typen werden mit `invalid_input` abgelehnt — vor
jedem Parsen.
## Ausgabe
| Name | Typ | Beschreibung |
| ----------- | ---- | --------------------------------------------- |
| `extracted` | json | `ExtractedDocument`-Schema siehe unten. |
```json
{
"engine": "pdf-extract",
"engine_version": "0.7",
"pages": [
{ "number": 1, "text": "…" }
]
}
```
`confidence` und `bbox` pro Seite sind im Schema für eine
spätere Service-Mode-Variante (pdfium / mupdf via
Host-Services) reserviert. v0.1.0 setzt diese Felder nicht;
das Schema akzeptiert sie als optional.
## Berechtigungen
Keine. Reine Rust-Pipeline (`pdf-extract`, `zip`, `quick-xml`).
Kein Dateisystem, kein Netz. `permissions: []` in
`module.yaml` macht das explizit; der Hub erzwingt es.
## Grenzen in v0.1.0
- **Layout-naives PDF.** Mehrspaltige oder tabellenlastige
Dokumente können umsortierten Text liefern. Ein-Spalten-
Fließtext ist solide.
- **Kein OCR.** Gescannte (nur-Bild-)PDFs liefern pro Seite
leeren Text. Künftiges `text.ocr@^0` als Vor-Schritt.
- **Nur DOCX bei Office.** `.doc` (Altformat), `.rtf`, `.odt`
noch nicht unterstützt.
`1.0.0` setzt entweder die Service-Mode-Variante
(pdfium-Qualität + Layout) voraus oder einen
nachgewiesen-guten Extraktions-Score auf einem
repräsentativen Korpus.
## Beispiel-Flow
```yaml
name: extract-anonymise
inputs:
document: bytes
steps:
- id: extract
use: text.extract@^0
with:
document: $inputs.document
- id: redact
use: text.anonymize@^0
with:
text: $extract.extracted.pages[*].text
outputs:
redacted: $redact.anonymized
```
## Build
```bash
cargo build --release --target wasm32-wasip2
# Ausgabe: target/wasm32-wasip2/release/text_extract.wasm
```

View file

@ -1,92 +0,0 @@
# text.extract
Plain-text extraction from PDF and DOCX documents. Pure-Rust,
in-WASM, no permissions required. The canonical first step in
any flow that needs raw text — anonymise, summarise, translate,
LLM-classify, audit-redact.
## Capability
- `text.extract@0.1.0`
## Inputs
| Name | Type | Description |
| ---------- | ----- | ------------------------------------------------------ |
| `document` | bytes | The document to extract. Accepted MIME types below. |
Accepted MIME types (declared in `module.yaml#accepts_mime`,
read by Studio as the file-picker filter):
- `application/pdf`
- `application/x-pdf`
- `application/vnd.openxmlformats-officedocument.wordprocessingml.document`
Other MIME types are rejected with a `invalid_input` error
before any parsing happens.
## Output
| Name | Type | Description |
| ----------- | ---- | ------------------------------------------------- |
| `extracted` | json | `ExtractedDocument` schema below. |
```json
{
"engine": "pdf-extract",
"engine_version": "0.7",
"pages": [
{ "number": 1, "text": "…" }
]
}
```
Per-page `confidence` and `bbox` fields are reserved in the
schema for a future service-mode variant (pdfium / mupdf via
host services) that surfaces layout information. v0.1.0 does
not populate them; the schema accepts them as optional.
## Permissions
None. Pure-Rust pipeline (`pdf-extract`, `zip`, `quick-xml`).
No filesystem, no network. The empty `permissions: []` in
`module.yaml` makes that explicit; the hub enforces it.
## Limits in v0.1.0
- **Layout-naive PDF.** Multi-column or table-heavy documents
may produce reordered text. Single-column body text is solid.
- **No OCR.** Scanned (image-only) PDFs return empty text per
page. Pair with a future `text.ocr@^0` step for those.
- **DOCX-only Office.** `.doc` (legacy binary), `.rtf`, `.odt`
are not yet supported.
Bumping to `1.0.0` requires either the service-mode variant
(pdfium-quality output + layout) or a confirmed-good extraction
score on a representative corpus.
## Example flow
```yaml
name: extract-anonymise
inputs:
document: bytes
steps:
- id: extract
use: text.extract@^0
with:
document: $inputs.document
- id: redact
use: text.anonymize@^0
with:
text: $extract.extracted.pages[*].text
outputs:
redacted: $redact.anonymized
```
## Build
```bash
cargo build --release --target wasm32-wasip2
# Output: target/wasm32-wasip2/release/text_extract.wasm
```

View file

@ -29,13 +29,3 @@ outputs:
# zip, quick-xml). No filesystem, no network. The empty list makes # zip, quick-xml). No filesystem, no network. The empty list makes
# this explicit; the hub enforces it. # this explicit; the hub enforces it.
permissions: [] permissions: []
# MIME-type allow-list for the `document` input. Studio reads this
# through ModuleInfoResponse.accepts_mime and uses it as the OS
# file-picker filter so operators don't accidentally pick a `.png`
# for a PDF-extract step. Same list documented inline next to
# `inputs.document` above; kept synchronised by hand for now.
accepts_mime:
- application/pdf
- application/x-pdf
- application/vnd.openxmlformats-officedocument.wordprocessingml.document