Build the website from public source
The website source and a synthetic fixture are public. Developers can inspect the Arabic and English interface, test keyword search, and follow the separate service distribution.
Changes to Lebanese Legislation, recorded retrospectively under the dates they happened. Read what changed and find the sources and interfaces you can use.
The website source and a synthetic fixture are public. Developers can inspect the Arabic and English interface, test keyword search, and follow the separate service distribution.
The court-scan release contains 44,766 PDFs and indexes for 274,936 pages. It preserves source links and hashes, with separate rights and privacy guidance.
A public handwriting demo pairs three original scan pages with a local editor. Downloads remain unreviewed drafts and nothing is automatically submitted.
The Hugging Face collection brings structured legal data, court scans and the browser transcription demo into one public resource list.
Version 0.1.4 service images are publicly available for Linux AMD64 and ARM64. The first database release manifest remains pending. This update explains what researchers and developers can use today.
A public MIT-licensed starter documents a local database, API and MCP setup with configurable services.
An optional vector index broadens discovery while rereading candidate passages from the canonical collection.
One search combines ordinary results with a later source-excerpt section, keeping the quoted wording tied to the library.
Source-linked Council notes add another part of the context needed to read a law’s status.
Ruling pages gain direct links to the legal articles identified in the citation graph.
When an exact citation is absent, search can show same-number texts from other dates with a clear notice.
Hard-to-read judgment pages can be published with a low-confidence warning and the original PDF link.
Article lookup accepts spelling variants while preserving distinct legal provision numbers.
API and MCP listings add date filters, pagination and article-level amendment relationships.
Scanned texts gain an issue-level publication route as well as their individual document pages.
Reading judgment records adds citations to the legal graph, extending the material available for article-to-ruling research.
The main website now opens in Arabic, with the English interface under /en/.
Judgment pages and API responses can mark words where machine readings disagree.
On 21 October 2026, unofficial English titles added another route into the collection. The Arabic source remains the reference for checking the wording.
Topic guides offer another entry point into the library, with links back to the provisions discussed.
Ruling pages gained access to full judgment scans, machine readings and search beyond catalogue summaries.
Notes and version records explain corrections without presenting every wording difference as a new law.
The update workflow links newly collected texts and refreshes the dataset alongside the library.
Number, year and date recognition helps a legal citation reach the intended text.
Date corrections record the original value, the replacement and the evidence used to settle the difference.
Disputed words remain visible on scanned legislation and in machine-readable responses.
Gazette page scans provide another source when a recent text is missing from the main collection.
A glossary fallback helps readers reach Arabic legal records when an English or French query has no direct match.
The official registry lists com.lebaneselegislation/mcp with the hosted endpoint. Its publication record dates to 13 October 2026.
Changes that multiply sums across a code now have links to provisions containing those amounts.
Arabic navigation brought the catalogue, source information and document pages into the language of the legal texts.
The public MCP server gives compatible assistants tools to search the library and retrieve its records.
On 13 October 2026, Atom feeds made changes to Lebanese law easier to follow without an account. The feeds point back to the amending texts.
The public dataset makes the library available as structured tables for analysis and reuse.
A legal graph connects texts, articles, rulings and Gazette issues while preserving the kind of relationship.
A read-only HTTP interface lets applications search and retrieve records without parsing website pages.
The library documents attribution and rights separately from the software used to serve it.
On 8 October 2026, Lebanese Legislation launched as a public library of Lebanese legal texts. Read the source, follow its references, and check the official publication.
An amendment layer separates a legal change from the wording originally displayed by the source.
Article navigation and stable code pages make long legal texts easier to read and cite.
The first archive established the source trail behind the library: original pages, stable record identities and structured exports.
The first project step was to preserve raw legal material and plan repeatable updates before building a reading and search interface.