citationer 2.1.2

One-click bibliometric analysis CLI tool

Citationer

One-click bibliometric analysis CLI tool — scan, import, analyze, and visualize your literature collection in the terminal.

Citationer is a lightweight, local-first, zero-config CLI tool for researchers. Drop into a directory with bibliographic export files, run a single command, and get a complete literature analysis — from descriptive statistics with terminal charts to knowledge graphs and AI-powered topic labeling.

---

Features

Category	Capability
🔍 Format Detection	Auto-detect CNKI, Web of Science exports (.xlsx, .txt, .ciw)
📊 Descriptive Stats	Yearly trends, top journals/authors/institutions, h-index — with terminal charts
📈 Terminal Charts	Braille line charts + Unicode bar charts rendered directly in terminal
🔗 Network Analysis	Keyword co-occurrence, author/institution collaboration, co-citation, bibliographic coupling
📝 Text Mining	Tokenization (jieba + spaCy), keyword frequency, LDA/NMF topic modeling, TF-IDF summarization, clustering
🤖 LLM-Powered AI	Topic labeling, literature review generation, trend identification, multi-dimensional classification — supports DeepSeek, OpenAI, Ollama
⚙ Configurable	CLI-driven config (config show/set/init), env-var support, multi-provider LLM
🎨 Rich Terminal	Color tables, progress bars, interactive HTML network graphs (Plotly)
📦 Pipe-friendly	JSON/CSV/GEXF/GraphML export — works with grep, jq, Gephi, Cytoscape

---

Installation

# Recommended: isolated install via pipx
pipx install citationer

# Or via pip
pip install citationer

# With all optional dependencies (NLP, network, AI, viz)
pip install "citationer[all]"

# From source
git clone https://github.com/JasonCENG/citationer.git
cd citationer
pip install -e ".[all,dev]"

---

Quick Start

# 1. Navigate to your literature directory
cd /path/to/literature

# 2. Scan for bibliographic files
citationer scan

# 3. Import into the local database (auto-clears old data)
citationer import

# 4. Clean & deduplicate
citationer clean

# 5. View the overview dashboard
citationer stats overview

---

Command Reference

Data Management

citationer scan                  # Scan directory for bibliographic files
citationer status                # Quick status check
citationer import                # Import files (clears old data by default)
citationer import --keep         # Append to existing data
citationer clean                 # Validate & deduplicate records

Descriptive Statistics (stats)

citationer stats overview             # Dashboard: totals, years, h-index, languages
citationer stats yearly               # Braille line chart + table
citationer stats yearly --cumulative  # Dual bar+line chart
citationer stats yearly --no-chart    # Table only (pipe-friendly)
citationer stats journals --top 20    # Horizontal bar chart
citationer stats authors --top 20     # Bar chart + Price's Law core authors
citationer stats institutions --top 20 # Bar chart

Text Mining (text)

citationer text preprocess       # Tokenize + language detection
citationer text keywords --top 30     # Keyword frequency
citationer text keywords --per-year   # Keyword × year heatmap
citationer text topics --method lda   # LDA topic modeling
citationer text topics --method nmf   # NMF topic modeling
citationer text summarize              # TF-IDF extractive summary
citationer text cluster --method kmeans  # Document clustering

Network Analysis (network)

citationer network keywords --top 50 --threshold 3   # Co-occurrence network
citationer network coauthors --min-papers 2           # Author collaboration
citationer network coauthors --type institutions       # Institution collaboration
citationer network cocitation --top 30                 # Co-citation analysis
citationer network coupling --top 30                   # Bibliographic coupling

# Export formats: csv, gexf, graphml, html (interactive)
citationer network keywords --output-format gexf --output graph.gexf
citationer network coauthors --viz --output network.html

LLM-Powered Analysis (ai)

# Configure your LLM first
citationer config set llm.api_key sk-your-key
citationer config set llm.model deepseek-chat

citationer ai topics --auto-label     # Auto-label LDA topics
citationer ai summarize               # Generate literature review (200-500 words)
citationer ai trends                  # Identify research trends & gaps
citationer ai classify                # Multi-dimensional classification
citationer ai info                    # View LLM config & cache stats

# Preview without API call
citationer ai summarize --dry-run

Configuration (config)

citationer config show                # View all settings
citationer config set llm.api_key sk-xxx  # Set API key
citationer config set llm.model gpt-4o    # Change model
citationer config set llm.base_url https://api.openai.com/v1  # Change provider
citationer config init                # Initialize config file with defaults

---

LLM Provider Configuration

Citationer supports any OpenAI-compatible API. Edit .citationer/config.yaml or use env vars:

# .citationer/config.yaml
llm:
  api_key: "sk-xxx"
  model: "deepseek-chat"
  base_url: "https://api.deepseek.com"
  temperature: 0.3
  max_tokens: 4096

Provider	base_url
DeepSeek	https://api.deepseek.com
OpenAI	https://api.openai.com/v1
Ollama (local)	http://localhost:11434/v1

Environment variables override the config file: CITATIONER_LLM_API_KEY, CITATIONER_LLM_MODEL, etc.

---

Supported Bibliographic Formats

Source	Format	Extensions	Status
Web of Science	Plain text / Tab-delimited / Excel	.txt, .ciw, .xlsx, .xls	✅ Stable
CNKI (知网)	Excel export	.xlsx	✅ Stable
Scopus	CSV/Excel	.csv, .xlsx	🔜 Phase 3
PubMed	XML/MEDLINE	.xml, .nbib	🔜 Phase 3
BibTeX / RIS	Generic	.bib, .ris	🔜 Phase 3

---

Development

# Install with dev dependencies
pip install -e ".[all,dev]"

# Run tests
pytest tests/ -v

# Lint & type check
ruff check src/ tests/
mypy src/ --ignore-missing-imports

# Coverage report
pytest tests/ --cov=src/citationer --cov-report=term-missing

---

Documentation

• User Manual — Full user manual (English)
• Handbook (中文) — Chinese user manual
• PRD v2.0 — Product requirements document
• PRD v1.0 — Original PRD (historical)

---

License

MIT © Jason Yu