chneas is built on other people's work. Here is all of it, with licences.
These are loaded from public CDNs at runtime and are not modified or redistributed by us.
| Library | Licence | Used by | For |
|---|---|---|---|
| tesseract.js | Apache-2.0 | Shade | Optical character recognition in a Web Worker |
| tesseract.js-core | Apache-2.0 | Shade | WebAssembly build of the Tesseract engine |
| Tesseract OCR | Apache-2.0 | Shade | The underlying OCR engine, and its eng trained data |
| jsQR | Apache-2.0 | Shade | QR-code detection so barcodes get covered too |
| pdf.js | Apache-2.0 | Mimir, Argus | PDF text extraction and page rendering |
| fflate | MIT | Argus, Selkie, Gryphon | Reading ZIP containers: .docx, .xlsx, .zip |
Apache-2.0 and MIT both permit commercial use. Apache-2.0 additionally grants an express patent licence from its contributors, which is why it is a safer base for a product than a bare MIT dependency. Both require that copyright and licence notices are preserved: this page is that notice. No library has been modified.
Mimir queries these directly from your browser. Each has its own terms; we do not cache, store or redistribute their data.
| Source | Terms | Notes |
|---|---|---|
| Crossref | Open metadata, no key required | Requests identify themselves via the optional polite-pool email. |
| OpenAlex | CC0 public domain | Free, no key. Polite pool honoured. |
| Semantic Scholar | Free tier, no key | Rate-limited without an API key; failures degrade to other sources rather than to a false verdict. |
| NCBI E-utilities | Public API | Usage guidelines apply at high volume. |
| DataCite | Open REST API | Covers datasets, theses and reports the academic indexes miss. |
| Google Books API | Google APIs Terms of Service | Used unauthenticated for public volume search. Results link back to Google Books. |
| OpenLibrary | Open, Internet Archive | ISBN resolution. |
Figures quoted on the tool pages are facts drawn from public reporting, credited at the point of use. They are referenced, not reproduced.
No competitor's code, UI, copy or assets. Every page here was written from scratch. The detection techniques: pattern matching with checksum validation, contrast sampling of rendered PDF pages, Unicode class scanning, ZIP part inspection: are standard, long-published methods, not anyone's proprietary process. Tool names are ordinary English words used descriptively.
If you believe anything here infringes your rights, email legal@chneas.com and we will look at it properly and quickly.
Released under the MIT licence. Source: see the repository link in the README.