notLang
A simple scripting language for GIS and geospatial analysis.
notLang is a domain-specific language (DSL) built on Python, with English and Uzbek messages. It brings vector/raster GIS analysis, satellite data (Google Earth Engine, Copernicus/Sentinel Hub, NASA GIBS, WorldClim), spatial statistics (Moran's I, LISA, DBSCAN), road-network analysis (OpenStreetMap/osmnx) and 160+ other commands together in one plain syntax.
| 🌐 Website | notgis.top |
| 📚 Documentation | docs.notgis.top |
| ▶️ Playground (in your browser) | play.notgis.top |
| 💻 GitHub | github.com/notGIS/notlang |
| 🧩 VS Code extension | Marketplace: notlang.notlang |
load "area.geojson" as a
buffer a 500 -> buf
connect gee
get "COPERNICUS/S2_SR_HARMONIZED" date "2026-06-01" "2026-06-30" bbox buf cloud_mask index "NDVI(B8,B4)" -> ndvi
zonal_stats ndvi with a stats "mean,max" -> result
table result to "report.html"
Installation
pip install notlang[full]
For a minimal install: pip install notlang. Optional features (GEE, 3D,
hydrology, ...) are added through extras when you need them, for example
pip install notlang[gee,cartography].
Requires Python 3.9 or newer.
| Extra | Adds |
|---|---|
full |
Almost everything below (recommended) |
colab |
A light bundle for Colab (does not re-download packages Colab already has) |
gee |
Google Earth Engine |
jupyter |
%load_ext notlang, ipyleaflet widgets |
webmap, threed |
Interactive maps (folium); 3D (plotly, pyvista) |
spectral, hydrology, interpolation |
Spectral indices; hydrology; kriging |
spatial-stats, timeseries, network |
Moran/LISA/DBSCAN; time series; OSM networks |
cartography, cartography-extra, parquet, vector-io |
Map decorations; zonal stats, static basemaps (incl. OpenFreeMap); Parquet; Fiona |
Windows (CMD / PowerShell)
py -m venv .venv
.venv\Scripts\activate
pip install notlang[full]
notlang --version
If the notlang command is "not found" (the Scripts folder is not on PATH), run
it through Python — the result is identical:
python -m notlang script.ntl
Google Colab
The easiest way in Colab is the colab extra. It adds nearly all features
(GEE, hydrology, interpolation, network analysis, ...) without re-downloading the
heavy packages Colab already ships (ipython, plotly, scikit-learn, xarray,
folium) or the 3D stack (pyvista/vtk/trame):
!pip install -q notlang[colab]
If you need 3D visualization (show3d), add it separately:
!pip install -q notlang[threed].
Usage
notlang # interactive mode (REPL)
notlang script.ntl # run a script file
notlang init project # create a new project template
notlang --version # version
notlang --about # about the author/project
notlang --lang uz ... # message language: en (default) or uz
python -m notlang ... # all of the above, works without PATH too
In Jupyter / Google Colab:
%load_ext notlang
%%notlang
load "area.geojson" as a
info a
webmap a
A Python variable from the notebook can be used inside a string as "$name"
(only in %%notlang):
path = "area.geojson"
%%notlang
load "$path" as a
You can also use it directly from Python:
from notlang import Interpreter
it = Interpreter()
for message in it.run_iter('load "area.geojson" as a\nbuffer a 500 -> b'):
print(message)
gdf = it.variables["b"].data # geopandas.GeoDataFrame
Earth Engine exports to Drive → local → show
connect gee "my-project" # regular account (a service account has no Drive storage)
load "area.geojson" as a
export_gee "USGS/SRTMGL1_003" bbox a to "drive" filename "DEM" scale 30 bands "elevation"
gee_status wait 600 # wait until the task is COMPLETED
download_drive "notlang_exports/DEM.tif" to "dem.tif" -> dem
show dem
Everyday vector tools
sjoin points polygons within left -> j # spatial join (keep all points)
sjoin polygons points contains stats "count,sum:pop" -> s # points per polygon: join_count, pop_sum
calc s density "pop_sum / area_ha" -> s2 # field calculator
update s2 set size "large" where "density > 5" -> s3 # edit values by condition
grid area 500 hex -> cells # square grid, or hexagons with `hex`
voronoi points clip area -> zones # Thiessen polygons
random_points area 100 seed 7 -> samples # random points inside polygons
unique s3 size # unique values of a column
split s3 size to "parts/" format "gpkg" # one file per value
set_crs layer 4326 -> layer2 # assign a CRS (no reprojection)
envelope layer -> frame # bounding rectangle (`each` = per feature)
dedupe layer by id -> clean # remove duplicates
polygonize lines -> blocks # polygons from closed lines
Everyday raster tools
rastercalc "log(dem) * where(slope > 5, 1, 0)" -> o # formula over several rasters
rastercalc "img_b4 - img_b3" -> d # pick a band of a multi-band raster
contour dem 50 -> lines # contour lines every 50 units (column `level`)
focal dem 3 mean -> smooth # moving window: mean sum min max std median range count
kde points 500 resolution 50 field pop -> density # kernel density (points per km², UTM raster)
rastercalc uses raster layer names directly in the formula. Functions: abs sqrt log log10 log2 exp sin cos tan arcsin arccos arctan floor ceil round sign minimum maximum clip where isnan power;
constants nan pi e; conditions with > < == & | ~ (use parentheses: (a > 1) & (b < 2)).
Other rasters are resampled to the grid of the first one; NoData and division by zero stay NoData.
Measurement units
area a in "ha,km2" -> b # columns area_ha, area_km2
area a in "mill" -> b # square miles (area_mi2)
length a in "km,mi,ft" -> b # columns length_km, length_mi, length_ft
Area units: m2, km2, ha, acre, mi2 (mile/mill), ft2, yd2, in2, cm2, dm2, mm2, sotix.
Length units: m, km, mi (mile/mill), ft, yd, in, nmi, cm, dm, mm. Without a unit the
columns stay area_m2 and length_m.
Buffers
Distances are in metres (geographic CRS such as EPSG:4326 are buffered in an automatically chosen UTM zone and returned in the original CRS).
buffer a 500 -> b # feature + outer buffer (default)
buffer a 500 outer -> b # outer ring only
buffer a 500 inner -> b # inner ring only (polygons)
buffer a 500 both -> b # both sides: 500 m inside and 500 m outside
buffer a 500 both to "b.geojson" -> b # buffer and save to GeoJSON in one step
show a, b
Basemaps
show and webmap accept basemap "<name>":
| Name | Source |
|---|---|
street / osm, light, dark, topo, satellite |
Raster tiles (OpenStreetMap, CARTO, OpenTopoMap, Esri) |
bright, liberty, positron, ofm_dark |
OpenFreeMap vector tiles — free, no API key |
OpenFreeMap needs pip install mapbox-vector-tile for static show images
(included in the cartography-extra and full extras).
Language
Messages are English by default. Use language "uz" inside a script, --lang uz
on the command line, or the NOTLANG_LANG=uz environment variable to switch to
Uzbek.
Features
- Vector/raster GIS: buffer (outer / inner / both), clip, zonal statistics, reclassification, hydrology, interpolation (IDW/kriging)
- Satellite platforms: Google Earth Engine (server-side cloud masking, index calculation, Drive/Asset export), Copernicus/Sentinel Hub (6 collections, Statistical API), NASA GIBS, STAC
- Climate data: WorldClim (averages, historical, CMIP6 future projections)
- Spatial statistics: Moran's I, LISA, DBSCAN, K-Means, hotspots (Getis-Ord Gi*)
- Road networks: via OpenStreetMap (osmnx) — download, routing, no API key
- Other open sources: GADM, Natural Earth, GBIF, USGS earthquakes, NASA POWER, SoilGrids, OpenAerialMap, Microsoft Building Footprints, and more
- Visualization: static maps, interactive web maps (widget in Jupyter, standalone HTML elsewhere), 3D views, charts, OpenFreeMap / OSM / CARTO basemaps
- Bilingual: English and Uzbek messages
- Safe formulas:
bandmathandfilter ... whereexpressions are restricted (arithmetic and comparisons only) — arbitrary Python code is never executed - Versioning: with
versioning "on", saved files are never overwritten; each one is saved with a timestamp
Documentation
Full guides (Colab, Windows CMD/PowerShell, GEE, Copernicus, open APIs and new commands) are at docs.notgis.top. You can also try notLang right in your browser at play.notgis.top. The source code is on GitHub.
Editor support
VS Code — the official notLang extension
(v0.9.28): syntax highlighting, IntelliSense (180+ commands), 50+ snippets, run a
script with Ctrl+Shift+Enter, a REPL, and results (webmap, 3D, charts) opened
automatically inside VS Code. It finds notLang by itself (venv, notlang or
python -m notlang), on Windows too. Search for "notLang" in the Marketplace or
install the .vsix file (works in Cursor as well). The notlang.language
setting (en / uz) switches the extension's messages.
Syntax highlighting, completions and snippets are also available for Sublime Text and Notepad++.
Author
Xushnudbek Rahimberganov Website: https://notgis.top Email: notgis@proton.me
License
MIT
Metadata
Release files for notlang 0.9.28
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| notlang-0.9.28.tar.gz | 246.6 kB | Without provenance | Details for notlang-0.9.28.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| notlang-0.9.28-py3-none-any.whl | Python 3 | none | any | Details for notlang-0.9.28-py3-none-any.whl |
Total release size: 464.4 kB
Release files / notlang-0.9.28.tar.gz
| Download URL | notlang-0.9.28.tar.gz |
|---|---|
| Size | 246.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
af42d08f861421422cfdc16d4a11f88927d395209993a94fc0848e574e29a59e
|
|
BLAKE2b-256 checksum How to use checksums |
482af48624f08dce410de1a901948f8881d652968ec6a5f863107ae358f50c7f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.18
|
Release files / notlang-0.9.28-py3-none-any.whl
| Download URL | notlang-0.9.28-py3-none-any.whl |
|---|---|
| Size | 217.8 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
c8172ca78de1f2e42e0c7d72d5bf08acd3111c7e6484e3e1d0e2d0ede93512c5
|
|
BLAKE2b-256 checksum How to use checksums |
e2c2180fe1ab89acb521a934a32a8a8539f84b6d4de28b75933ee708fff6a562
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.18
|
Security for notlang 0.9.28
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.