Vahan Scraper

A scraper for the Vahan Dashboard: India's national vehicle registration database.

Two scraping methods are available:

	scraper.py	api.py
Method	Playwright browser automation	Direct AJAX/HTTP requests
Output	.xlsx (downloaded from site)	.csv (parsed from HTML table)
Speed	Slower (browser startup + download per file)	~10× faster (plain HTTP, no browser)
Use when	You need the original XLSX format	You need speed or can't install Playwright

Both tools use identical CLI flags and write to the same --out directory layout, so you can switch between them without changing anything else.

---

Requirements

pip install -r requirements.txt
playwright install chromium   # only needed for scraper.py

---

scraper.py - Browser-based (XLSX output)

python3 scripts/scraper.py --yaxis <value> --xaxis <value> [options]

Arguments

Argument	Required	Default	Description
--yaxis	Yes*	—	Y-Axis variable
--xaxis	Yes*	—	X-Axis variable
--year	No	current year	Single year to scrape
--start-year	No	—	Start of year range (inclusive)
--end-year	No	current year	End of year range (inclusive)
--state	No	all-states aggregate	State name, partial match (e.g. Kerala)
--all-rtos	No	—	Loop through every RTO in the selected state (requires --state)
--rto	No	—	One or more specific RTO names (requires --state)
--out	No	vahan_data/	Output directory
--list-options	No	—	Print all available dropdown options and exit
--no-headless	No	—	Show the browser window (useful for debugging)

*Required unless --list-options is used.

Examples

# See all available states, axes, and years from the live site
python3 scripts/scraper.py --list-options

# All-states aggregate, current year
python3 scripts/scraper.py --yaxis "Vehicle Category" --xaxis "Fuel"

# Specific state and year
python3 scripts/scraper.py --yaxis "Vehicle Category" --xaxis "Fuel" --state "Kerala" --year 2025

# Year range for a state
python3 scripts/scraper.py --yaxis "Vehicle Category" --xaxis "Fuel" \
  --state "Kerala" --start-year 2020 --end-year 2026

# All RTOs in a state, year range
python3 scripts/scraper.py --yaxis "Vehicle Category" --xaxis "Fuel" \
  --state "Kerala" --all-rtos --start-year 2020 --end-year 2026

# Specific RTOs
python3 scripts/scraper.py --yaxis "Maker" --xaxis "Fuel" \
  --state "Kerala" --rto "TRIVANDRUM RTO - KL1" "KOLLAM RTO - KL2"

# Debug (show browser window)
python3 scripts/scraper.py --yaxis "Vehicle Category" --xaxis "Fuel" \
  --state "Delhi" --no-headless

---

api.py - HTTP-based (CSV output, no browser)

Replays the PrimeFaces AJAX protocol the dashboard uses internally. Significantly faster than the browser-based approach.

python3 scripts/api.py --yaxis <value> --xaxis <value> [options]

Same flags as scraper.py except there is no --no-headless (no browser involved). Output files are .csv instead of .xlsx.

Examples

# See all available options (reads from live site, no browser)
python3 scripts/api.py --list-options

# All-states aggregate, current year
python3 scripts/api.py --yaxis "Vehicle Category" --xaxis "Fuel"

# Specific state and year
python3 scripts/api.py --yaxis "Vehicle Category" --xaxis "Fuel" --state "Kerala" --year 2025

# Year range for a state
python3 scripts/api.py --yaxis "Vehicle Category" --xaxis "Fuel" \
  --state "Kerala" --start-year 2020 --end-year 2026

# All RTOs in a state, year range
python3 scripts/api.py --yaxis "Vehicle Category" --xaxis "Fuel" \
  --state "Kerala" --all-rtos --start-year 2020 --end-year 2026

# Specific RTOs (partial name match)
python3 scripts/api.py --yaxis "Maker" --xaxis "Fuel" \
  --state "Kerala" --rto "TRIVANDRUM" "KOLLAM"

---

Axis options

Y-Axis

Vehicle Category, Vehicle Class, Norms, Fuel, Maker, State

X-Axis

Vehicle Category, Norms, Fuel, Vehicle Category Group, Financial Year, Calendar Year, Month Wise

---

Output structure

Both tools write to the same layout under --out (default: vahan_data/):

vahan_data/
└── <state>/                      # e.g. Kerala_87_ or all_states
    └── <rto>/                    # e.g. TRIVANDRUM_RTO___KL1 or all_rtos
        └── <yaxis>__<xaxis>/
            ├── 2020.xlsx          # scraper.py
            ├── 2021.csv           # api.py
            └── ...

Already-downloaded files are skipped automatically, so interrupted runs resume safely.