- version 0.1.2 in pyproject.toml and __init__.py
- CHANGELOG: [Unreleased] -> [0.1.2] with "Beim Update zu tun"
- install examples and version notes in README, BEDIENUNGSANLEITUNG,
FUNKTIONSWEISE point to v0.1.2
- no code changes; widget files unchanged (v0.1.0)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Forgejo scales SVG diagrams down to the content width (~900 px); the
left-to-right overview/deployment diagrams were 1650-2960 px wide, i.e.
5-9 px font. Redrawn top-to-bottom with multi-line labels (now 450-730 px,
no scaling); replaced an overlapping self-loop label by an explicit step.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Architecture and data flow (Mermaid), modules, query language internals
(tokenizer rules, grammar, term compilation), evaluation and timeouts,
highlighting and snippets, Markdown splitting and dialects, search engine
pipeline, Web API and widget, design decisions with rejected alternatives,
limits and extension ideas.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- CHANGELOG.md (Keep a Changelog, German) for 0.1.0, 0.1.1 and unreleased
documentation work, incl. "Beim Update zu tun"
- README: documentation table (manual, how it works, changelog, license)
- manual: release steps now include updating CHANGELOG/FUNKTIONSWEISE
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New section 8: requirements for searching a collection (encoding, size,
split mode vs. structure, metadata via naming scheme, limits) and for
jump anchors (predictable HTML paths, three ways to ids, same Markdown
dialect, keeping source and HTML in sync, consistency check script),
plus a checklist. Following sections renumbered 9-12.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Installation, query syntax, CLI, Python API, Web API, widget, server
operation (gunicorn, systemd, nginx), security, recipes, error codes,
troubleshooting and release workflow. Linked from README.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- split_markdown/load_markdown: dialect="commonmark"|"python-markdown" so list
blocks match the HTML structure of the renderer (list after paragraph needs a
blank line; nesting needs 4 spaces in Python-Markdown)
- SearchEngine caches the searched text per block on load
- regex timeout only for /regex/ terms (words and phrases are linear);
typical queries over 26k blocks: ~30-140 ms instead of 250-700 ms
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
