ArticleManager Presentor

A user-friendly, survey-first research manager with a database-backed GUI, structured article understanding, captures, notes, comparison table maker, and LaTeX render preview.

Key features

• Project workspace with taxonomy + writing outline.
• Dashboard intelligence: article counts, read/included stats, cluster distribution, missing BibTeX/analysis.
• Management tab dedicated to article CRUD + bulk file-driven article add.
• Separate Analysis tab for structured article analysis and capture/source memory.
• Article decision system fields:
  • role, task, method tags,
  • evidence strength,
  • relevance score,
  • read status,
  • decision flag.
• Structured article analysis form:
  • problem, setting, dataset, algorithm,
  • baselines, metrics, findings,
  • limitations, assumptions, future work,
  • commentary, extracted figures/equations/tables.
• Capture system with provenance:
  • source URL, page title, selected text, screenshot, tag, comment, timestamp, linked article.
• Noter tab for side-by-side note taking with pinning and anchor metadata.
• Comparison matrix builder:
  • preview rows,
  • LaTeX table export,
  • CSV output.
• BibTeX workflow:
  • article link,
  • validation warnings (missing author/year/title/venue, malformed).
• Export:
  • project CSV,
  • project .bib,
  • full database snapshot (.db) save/load from UI.
• LaTeX render tab using MathJax (live preview).
• Presenter upgrades:
  • full-screen toggle for center preview pane (Esc to exit),
  • zoom enable/disable with wheel + keyboard shortcuts,
  • fit-width zoom preset,
  • file tree search/filter,
  • direct main.tex render from Presenter,
  • bulk multi-file add into SourceArticles/.
• Project bootstrap defaults:
  • SourceArticles/ and .bib/ directories,
  • main.tex + .bib/references.bib created automatically,
  • each new article gets its own folder under SourceArticles/.
• NLP auto-tagging now combines stemming-aware keywords, keyphrase extraction, and weighted topic clustering.

Run

python app.py

Open: http://localhost:5000

LaTeX PDF rendering

/api/latex/render uses pdflatex from PATH by default. If the server process cannot see PATH (common on Windows services/IDEs), set PDFLATEX_PATH to an absolute executable path before starting the app.

Example (Windows):

$env:PDFLATEX_PATH = "C:\Program Files\MiKTeX\miktex\bin\x64\pdflatex.exe"
python app.py

Google Chrome extension (Open + RunT helper)

This repo now includes a Chrome extension in chrome_extension/.

What it does:

• Displays the default RunT command.
• Open the Thing button opens http://localhost:5000.
• ReRun button copies the RunT command so you can paste and run it in PowerShell.

Load it in Chrome:

1. Open chrome://extensions.
2. Turn on Developer mode.
3. Click Load unpacked and choose the chrome_extension folder.

Default RunT command shown by the extension:

$env:PDFLATEX_PATH = "C:\Program Files\MiKTeX\miktex\bin\x64\pdflatex.exe"
& C:/Users/parib/AppData/Local/Programs/Python/Python310/python.exe c:/Users/parib/Desktop/ArticleManager/app.py

API highlights

• POST /api/projects
• GET /api/dashboard?project=...
• GET/POST /api/articles
• GET/POST /api/analysis
• GET/POST /api/bib
• GET/POST /api/notes
• GET/POST /api/checklist + PUT /api/checklist
• POST /api/extension/capture
• GET /api/comparison?project=...&article_ids=1,2,3
• GET /api/export/articles.csv?project=...
• GET /api/export/project.bib?project=...
• GET /api/database/export
• POST /api/database/import

Data

• SQLite database: article_manager.db