0.31 (2019-11-11)

This version adds compatibility with Python 3.8 and breaks compatibility with Python 3.5.

If you are still running Python 3.5 you should stick with 0.30.2, which you can install like this:

pip install datasette==0.30.2
  • Format SQL button now works with read-only SQL queries - thanks, Tobias Kunze (#602)
  • New ?column__notin=x,y,z filter for table views (#614)
  • Table view now uses select col1, col2, col3 instead of select *
  • Database filenames can now contain spaces - thanks, Tobias Kunze (#590)
  • Removed obsolete ?_group_count=col feature (#504)
  • Improved user interface and documentation for datasette publish cloudrun (#608)
  • Tables with indexes now show the `` CREATE INDEX`` statements on the table page (#618)
  • Current version of uvicorn is now shown on /-/versions
  • Python 3.8 is now supported! (#622)
  • Python 3.5 is no longer supported.

0.30.2 (2019-11-02)

  • /-/plugins page now uses distribution name e.g. datasette-cluster-map instead of the name of the underlying Python package (datasette_cluster_map) (#606)
  • Array faceting is now only suggested for columns that contain arrays of strings (#562)
  • Better documentation for the --host argument (#574)
  • Don’t show None with a broken link for the label on a nullable foreign key (#406)

0.30.1 (2019-10-30)

  • Fixed bug where ?_where= parameter was not persisted in hidden form fields (#604)
  • Fixed bug with .JSON representation of row pages - thanks, Chris Shaw (#603)

0.30 (2019-10-18)

  • Added /-/threads debugging page
  • Allow EXPLAIN WITH... (#583)
  • Button to format SQL - thanks, Tobias Kunze (#136)
  • Sort databases on homepage by argument order - thanks, Tobias Kunze (#585)
  • Display metadata footer on custom SQL queries - thanks, Tobias Kunze (#589)
  • Use --platform=managed for publish cloudrun (#587)
  • Fixed bug returning non-ASCII characters in CSV (#584)
  • Fix for /foo v.s. /foo-bar bug (#601)

0.29.3 (2019-09-02)

  • Fixed implementation of CodeMirror on database page (#560)
  • Documentation typo fixes - thanks, Min ho Kim (#561)
  • Mechanism for detecting if a table has FTS enabled now works if the table name used alternative escaping mechanisms (#570) - for compatibility with a recent change to sqlite-utils.

0.29.2 (2019-07-13)

  • Bumped Uvicorn to 0.8.4, fixing a bug where the querystring was not included in the server logs. (#559)
  • Fixed bug where the navigation breadcrumbs were not displayed correctly on the page for a custom query. (#558)
  • Fixed bug where custom query names containing unicode characters caused errors.

0.29.1 (2019-07-11)

  • Fixed bug with static mounts using relative paths which could lead to traversal exploits (#555) - thanks Abdussamet Kocak!
  • Datasette can now be run as a module: python -m datasette (#556) - thanks, Abdussamet Kocak!

0.29 (2019-07-07)

ASGI, new plugin hooks, facet by date and much, much more…


ASGI is the Asynchronous Server Gateway Interface standard. I’ve been wanting to convert Datasette into an ASGI application for over a year - Port Datasette to ASGI #272 tracks thirteen months of intermittent development - but with Datasette 0.29 the change is finally released. This also means Datasette now runs on top of Uvicorn and no longer depends on Sanic.

I wrote about the significance of this change in Porting Datasette to ASGI, and Turtles all the way down.

The most exciting consequence of this change is that Datasette plugins can now take advantage of the ASGI standard.

New plugin hook: asgi_wrapper

The asgi_wrapper(datasette) plugin hook allows plugins to entirely wrap the Datasette ASGI application in their own ASGI middleware. (#520)

Two new plugins take advantage of this hook:

  • datasette-auth-github adds a authentication layer: users will have to sign in using their GitHub account before they can view data or interact with Datasette. You can also use it to restrict access to specific GitHub users, or to members of specified GitHub organizations or teams.
  • datasette-cors allows you to configure CORS headers for your Datasette instance. You can use this to enable JavaScript running on a whitelisted set of domains to make fetch() calls to the JSON API provided by your Datasette instance.

New plugin hook: extra_template_vars

The extra_template_vars(template, database, table, view_name, request, datasette) plugin hook allows plugins to inject their own additional variables into the Datasette template context. This can be used in conjunction with custom templates to customize the Datasette interface. datasette-auth-github uses this hook to add custom HTML to the new top navigation bar (which is designed to be modified by plugins, see #540).

Secret plugin configuration options

Plugins like datasette-auth-github need a safe way to set secret configuration options. Since the default mechanism for configuring plugins exposes those settings in /-/metadata a new mechanism was needed. Secret configuration values describes how plugins can now specify that their settings should be read from a file or an environment variable:

    "plugins": {
        "datasette-auth-github": {
            "client_secret": {
                "$env": "GITHUB_CLIENT_SECRET"

These plugin secrets can be set directly using datasette publish. See Custom metadata and plugins for details. (#538 and #543)

Facet by date

If a column contains datetime values, Datasette can now facet that column by date. (#481)

Easier custom templates for table rows

If you want to customize the display of individual table rows, you can do so using a _table.html template include that looks something like this:

{% for row in display_rows %}
        <h2>{{ row["title"] }}</h2>
        <p>{{ row["description"] }}<lp>
        <p>Category: {{ row.display("category_id") }}</p>
{% endfor %}

This is a backwards incompatible change. If you previously had a custom template called _rows_and_columns.html you need to rename it to _table.html.

See Custom templates for full details.

?_through= for joins through many-to-many tables

The new ?_through={json} argument to the Table view allows records to be filtered based on a many-to-many relationship. See Special table arguments for full documentation - here’s an example. (#355)

This feature was added to help support facet by many-to-many, which isn’t quite ready yet but will be coming in the next Datasette release.

Small changes

  • Databases published using datasette publish now open in Immutable mode. (#469)
  • ?col__date= now works for columns containing spaces
  • Automatic label detection (for deciding which column to show when linking to a foreign key) has been improved. (#485)
  • Fixed bug where pagination broke when combined with an expanded foreign key. (#489)
  • Contributors can now run pip install -e .[docs] to get all of the dependencies needed to build the documentation, including cd docs && make livehtml support.
  • Datasette’s dependencies are now all specified using the ~= match operator. (#532)
  • white-space: pre-wrap now used for table creation SQL. (#505)

Full list of commits between 0.28 and 0.29.

0.28 (2019-05-19)

A salmagundi of new features!

Supporting databases that change

From the beginning of the project, Datasette has been designed with read-only databases in mind. If a database is guaranteed not to change it opens up all kinds of interesting opportunities - from taking advantage of SQLite immutable mode and HTTP caching to bundling static copies of the database directly in a Docker container. The interesting ideas in Datasette explores this idea in detail.

As my goals for the project have developed, I realized that read-only databases are no longer the right default. SQLite actually supports concurrent access very well provided only one thread attempts to write to a database at a time, and I keep encountering sensible use-cases for running Datasette on top of a database that is processing inserts and updates.

So, as-of version 0.28 Datasette no longer assumes that a database file will not change. It is now safe to point Datasette at a SQLite database which is being updated by another process.

Making this change was a lot of work - see tracking tickets #418, #419 and #420. It required new thinking around how Datasette should calculate table counts (an expensive operation against a large, changing database) and also meant reconsidering the “content hash” URLs Datasette has used in the past to optimize the performance of HTTP caches.

Datasette can still run against immutable files and gains numerous performance benefits from doing so, but this is no longer the default behaviour. Take a look at the new Performance and caching documentation section for details on how to make the most of Datasette against data that you know will be staying read-only and immutable.

Faceting improvements, and faceting plugins

Datasette Facets provide an intuitive way to quickly summarize and interact with data. Previously the only supported faceting technique was column faceting, but 0.28 introduces two powerful new capabilities: facet-by-JSON-array and the ability to define further facet types using plugins.

Facet by array (#359) is only available if your SQLite installation provides the json1 extension. Datasette will automatically detect columns that contain JSON arrays of values and offer a faceting interface against those columns - useful for modelling things like tags without needing to break them out into a new table. See Facet by JSON array for more.

The new register_facet_classes() plugin hook (#445) can be used to register additional custom facet classes. Each facet class should provide two methods: suggest() which suggests facet selections that might be appropriate for a provided SQL query, and facet_results() which executes a facet operation and returns results. Datasette’s own faceting implementations have been refactored to use the same API as these plugins.

datasette publish cloudrun

Google Cloud Run is a brand new serverless hosting platform from Google, which allows you to build a Docker container which will run only when HTTP traffic is received and will shut down (and hence cost you nothing) the rest of the time. It’s similar to Zeit’s Now v1 Docker hosting platform which sadly is no longer accepting signups from new users.

The new datasette publish cloudrun command was contributed by Romain Primet (#434) and publishes selected databases to a new Datasette instance running on Google Cloud Run.

See Publishing to Google Cloud Run for full documentation.

register_output_renderer plugins

Russ Garrett implemented a new Datasette plugin hook called register_output_renderer (#441) which allows plugins to create additional output renderers in addition to Datasette’s default .json and .csv.

Russ’s in-development datasette-geo plugin includes an example of this hook being used to output .geojson automatically converted from SpatiaLite.

Medium changes

  • Datasette now conforms to the Black coding style (#449) - and has a unit test to enforce this in the future
  • New Special table arguments:
    • ?columnname__in=value1,value2,value3 filter for executing SQL IN queries against a table, see Table arguments (#433)
    • ?columnname__date=yyyy-mm-dd filter which returns rows where the spoecified datetime column falls on the specified date (583b22a)
    • ?tags__arraycontains=tag filter which acts against a JSON array contained in a column (78e45ea)
    • ?_where=sql-fragment filter for the table view (#429)
    • ?_fts_table=mytable and ?_fts_pk=mycolumn querystring options can be used to specify which FTS table to use for a search query - see Configuring full-text search for a table or view (#428)
  • You can now pass the same table filter multiple times - for example, ?content__not=world&content__not=hello will return all rows where the content column is neither hello or world (#288)
  • You can now specify about and about_url metadata (in addition to source and license) linking to further information about a project - see Source, license and about
  • New ?_trace=1 parameter now adds debug information showing every SQL query that was executed while constructing the page (#435)
  • datasette inspect now just calculates table counts, and does not introspect other database metadata (#462)
  • Removed /-/inspect page entirely - this will be replaced by something similar in the future, see #465
  • Datasette can now run against an in-memory SQLite database. You can do this by starting it without passing any files or by using the new --memory option to datasette serve. This can be useful for experimenting with SQLite queries that do not access any data, such as SELECT 1+1 or SELECT sqlite_version().

Small changes

  • We now show the size of the database file next to the download link (#172)
  • New /-/databases introspection page shows currently connected databases (#470)
  • Binary data is no longer displayed on the table and row pages (#442 - thanks, Russ Garrett)
  • New show/hide SQL links on custom query pages (#415)
  • The extra_body_script plugin hook now accepts an optional view_name argument (#443 - thanks, Russ Garrett)
  • Bumped Jinja2 dependency to 2.10.1 (#426)
  • All table filters are now documented, and documentation is enforced via unit tests (2c19a27)
  • New project guideline: master should stay shippable at all times! (31f36e1)
  • Fixed a bug where sqlite_timelimit() occasionally failed to clean up after itself (bac4e01)
  • We no longer load additional plugins when executing pytest (#438)
  • Homepage now links to database views if there are less than five tables in a database (#373)
  • The --cors option is now respected by error pages (#453)
  • datasette publish heroku now uses the --include-vcs-ignore option, which means it works under Travis CI (#407)
  • datasette publish heroku now publishes using Python 3.6.8 (666c374)
  • Renamed datasette publish now to datasette publish nowv1 (#472)
  • datasette publish nowv1 now accepts multiple --alias parameters (09ef305)
  • Removed the datasette skeleton command (#476)
  • The documentation on how to build the documentation now recommends sphinx-autobuild

0.27.1 (2019-05-09)

  • Tiny bugfix release: don’t install tests/ in the wrong place. Thanks, Veit Heller.

0.27 (2019-01-31)

0.26.1 (2019-01-10)

  • /-/versions now includes SQLite compile_options (#396)
  • datasetteproject/datasette Docker image now uses SQLite 3.26.0 (#397)
  • Cleaned up some deprecation warnings under Python 3.7

0.26 (2019-01-02)

  • datasette serve --reload now restarts Datasette if a database file changes on disk.
  • datasette publish now now takes an optional --alias argument. This will attempt to set an alias after the deploy completes.
  • Fixed a bug where the advanced CSV export form failed to include the currently selected filters (#393)

0.25.2 (2018-12-16)

0.25.1 (2018-11-04)

Documentation improvements plus a fix for publishing to Zeit Now.

  • datasette publish now now uses Zeit’s v1 platform, to work around the new 100MB image limit. Thanks, @slygent - closes #366.

0.25 (2018-09-19)

New plugin hooks, improved database view support and an easier way to use more recent versions of SQLite.

  • New publish_subcommand plugin hook. A plugin can now add additional datasette publish publishers in addition to the default now and heroku, both of which have been refactored into default plugins. publish_subcommand documentation. Closes #349
  • New render_cell plugin hook. Plugins can now customize how values are displayed in the HTML tables produced by Datasette’s browseable interface. datasette-json-html and datasette-render-images are two new plugins that use this hook. render_cell documentation. Closes #352
  • New extra_body_script plugin hook, enabling plugins to provide additional JavaScript that should be added to the page footer. extra_body_script documentation.
  • extra_css_urls and extra_js_urls hooks now take additional optional parameters, allowing them to be more selective about which pages they apply to. Documentation.
  • You can now use the sortable_columns metadata setting to explicitly enable sort-by-column in the interface for database views, as well as for specific tables.
  • The new fts_table and fts_pk metadata settings can now be used to explicitly configure full-text search for a table or a view, even if that table is not directly coupled to the SQLite FTS feature in the database schema itself.
  • Datasette will now use pysqlite3 in place of the standard library sqlite3 module if it has been installed in the current environment. This makes it much easier to run Datasette against a more recent version of SQLite, including the just-released SQLite 3.25.0 which adds window function support. More details on how to use this in #360
  • New mechanism that allows plugin configuration options to be set using metadata.json.

0.24 (2018-07-23)

A number of small new features:

  • datasette publish heroku now supports --extra-options, fixes #334
  • Custom error message if SpatiaLite is needed for specified database, closes #331
  • New config option: truncate_cells_html for truncating long cell values in HTML view - closes #330
  • Documentation for datasette publish and datasette package, closes #337
  • Fixed compatibility with Python 3.7
  • datasette publish heroku now supports app names via the -n option, which can also be used to overwrite an existing application [Russ Garrett]
  • Title and description metadata can now be set for canned SQL queries, closes #342
  • New force_https_on config option, fixes https:// API URLs when deploying to Zeit Now - closes #333
  • ?_json_infinity=1 querystring argument for handling Infinity/-Infinity values in JSON, closes #332
  • URLs displayed in the results of custom SQL queries are now URLified, closes #298

0.23.2 (2018-07-07)

Minor bugfix and documentation release.

0.23.1 (2018-06-21)

Minor bugfix release.

  • Correctly display empty strings in HTML table, closes #314
  • Allow “.” in database filenames, closes #302
  • 404s ending in slash redirect to remove that slash, closes #309
  • Fixed incorrect display of compound primary keys with foreign key references. Closes #319
  • Docs + example of canned SQL query using || concatenation. Closes #321
  • Correctly display facets with value of 0 - closes #318
  • Default ‘expand labels’ to checked in CSV advanced export

0.23 (2018-06-18)

This release features CSV export, improved options for foreign key expansions, new configuration settings and improved support for SpatiaLite.

See datasette/compare/0.22.1…0.23 for a full list of commits added since the last release.

CSV export

Any Datasette table, view or custom SQL query can now be exported as CSV.


Check out the CSV export documentation for more details, or try the feature out on

If your table has more than max_returned_rows (default 1,000) Datasette provides the option to stream all rows. This option takes advantage of async Python and Datasette’s efficient pagination to iterate through the entire matching result set and stream it back as a downloadable CSV file.

Foreign key expansions

When Datasette detects a foreign key reference it attempts to resolve a label for that reference (automatically or using the Specifying the label column for a table metadata option) so it can display a link to the associated row.

This expansion is now also available for JSON and CSV representations of the table, using the new _labels=on querystring option. See Expanding foreign key references for more details.

New configuration settings

Datasette’s Configuration now also supports boolean settings. A number of new configuration options have been added:

  • num_sql_threads - the number of threads used to execute SQLite queries. Defaults to 3.
  • allow_facet - enable or disable custom Facets using the _facet= parameter. Defaults to on.
  • suggest_facets - should Datasette suggest facets? Defaults to on.
  • allow_download - should users be allowed to download the entire SQLite database? Defaults to on.
  • allow_sql - should users be allowed to execute custom SQL queries? Defaults to on.
  • default_cache_ttl - Default HTTP caching max-age header in seconds. Defaults to 365 days - caching can be disabled entirely by settings this to 0.
  • cache_size_kb - Set the amount of memory SQLite uses for its per-connection cache, in KB.
  • allow_csv_stream - allow users to stream entire result sets as a single CSV file. Defaults to on.
  • max_csv_mb - maximum size of a returned CSV file in MB. Defaults to 100MB, set to 0 to disable this limit.

Control HTTP caching with ?_ttl=

You can now customize the HTTP max-age header that is sent on a per-URL basis, using the new ?_ttl= querystring parameter.

You can set this to any value in seconds, or you can set it to 0 to disable HTTP caching entirely.

Consider for example this query which returns a randomly selected member of the Avengers:

select * from [avengers/avengers] order by random() limit 1

If you hit the following page repeatedly you will get the same result, due to HTTP caching:


By adding ?_ttl=0 to the zero you can ensure the page will not be cached and get back a different super hero every time:


Improved support for SpatiaLite

The SpatiaLite module for SQLite adds robust geospatial features to the database.

Getting SpatiaLite working can be tricky, especially if you want to use the most recent alpha version (with support for K-nearest neighbor).

Datasette now includes extensive documentation on SpatiaLite, and thanks to Ravi Kotecha our GitHub repo includes a Dockerfile that can build the latest SpatiaLite and configure it for use with Datasette.

The datasette publish and datasette package commands now accept a new --spatialite argument which causes them to install and configure SpatiaLite as part of the container they deploy.

Every commit to Datasette master is now automatically deployed by Travis CI to - ensuring there is always a live demo of the latest version of the software.

The demo uses the fixtures from our unit tests, ensuring it demonstrates the same range of functionality that is covered by the tests.

You can see how the deployment mechanism works in our .travis.yml file.


  • Got JSON data in one of your columns? Use the new ?_json=COLNAME argument to tell Datasette to return that JSON value directly rather than encoding it as a string.
  • If you just want an array of the first value of each row, use the new ?_shape=arrayfirst option - example.

0.22.1 (2018-05-23)

Bugfix release, plus we now use versioneer for our version numbers.

  • Faceting no longer breaks pagination, fixes #282

  • Add __version_info__ derived from __version__ [Robert Gieseke]

    This might be tuple of more than two values (major and minor version) if commits have been made after a release.

  • Add version number support with Versioneer. [Robert Gieseke]

    Versioneer Licence: Public Domain (CC0-1.0)

    Closes #273

  • Refactor inspect logic [Russ Garrett]

0.22 (2018-05-20)

The big new feature in this release is Facets. Datasette can now apply faceted browse to any column in any table. It will also suggest possible facets. See the Datasette Facets announcement post for more details.

In addition to the work on facets:

  • Added docs for introspection endpoints

  • New --config option, added --help-config, closes #274

    Removed the --page_size= argument to datasette serve in favour of:

    datasette serve --config default_page_size:50 mydb.db

    Added new help section:

    $ datasette --help-config
    Config options:
      default_page_size            Default page size for the table view
      max_returned_rows            Maximum rows that can be returned from a table
                                   or custom query (default=1000)
      sql_time_limit_ms            Time limit for a SQL query in milliseconds
      default_facet_size           Number of values to return for requested facets
      facet_time_limit_ms          Time limit for calculating a requested facet
      facet_suggest_time_limit_ms  Time limit for calculating a suggested facet
  • Only apply responsive table styles to .rows-and-column

    Otherwise they interfere with tables in the description, e.g. on

  • Refactored views into new views/ modules, refs #256

  • Documentation for SQLite full-text search support, closes #253

  • /-/versions now includes SQLite fts_versions, closes #252

0.21 (2018-05-05)

New JSON _shape= options, the ability to set table _size= and a mechanism for searching within specific columns.

  • Default tests to using a longer timelimit

    Every now and then a test will fail in Travis CI on Python 3.5 because it hit the default 20ms SQL time limit.

    Test fixtures now default to a 200ms time limit, and we only use the 20ms time limit for the specific test that tests query interruption. This should make our tests on Python 3.5 in Travis much more stable.

  • Support _search_COLUMN=text searches, closes #237

  • Show version on /-/plugins page, closes #248

  • ?_size=max option, closes #249

  • Added /-/versions and /-/versions.json, closes #244

    Sample output:

      "python": {
        "version": "3.6.3",
        "full": "3.6.3 (default, Oct  4 2017, 06:09:38) \n[GCC 4.2.1 Compatible Apple LLVM 9.0.0 (clang-900.0.37)]"
      "datasette": {
        "version": "0.20"
      "sqlite": {
        "version": "3.23.1",
        "extensions": {
          "json1": null,
          "spatialite": "4.3.0a"
  • Renamed ?_sql_time_limit_ms= to ?_timelimit, closes #242

  • New ?_shape=array option + tweaks to _shape, closes #245

    • Default is now ?_shape=arrays (renamed from lists)
    • New ?_shape=array returns an array of objects as the root object
    • Changed ?_shape=object to return the object as the root
    • Updated docs
  • FTS tables now detected by inspect(), closes #240

  • New ?_size=XXX querystring parameter for table view, closes #229

    Also added documentation for all of the _special arguments.

    Plus deleted some duplicate logic implementing _group_count.

  • If max_returned_rows==page_size, increment max_returned_rows - fixes #230

  • New hidden: True option for table metadata, closes #239

  • Hide idx_* tables if spatialite detected, closes #228

  • Added class=rows-and-columns to custom query results table

  • Added CSS class rows-and-columns to main table

  • label_column option in metadata.json - closes #234

0.20 (2018-04-20)

Mostly new work on the Plugins mechanism: plugins can now bundle static assets and custom templates, and datasette publish has a new --install=name-of-plugin option.

  • Add col-X classes to HTML table on custom query page

  • Fixed out-dated template in documentation

  • Plugins can now bundle custom templates, #224

  • Added /-/metadata /-/plugins /-/inspect, #225

  • Documentation for –install option, refs #223

  • Datasette publish/package –install option, #223

  • Fix for plugins in Python 3.5, #222

  • New plugin hooks: extra_css_urls() and extra_js_urls(), #214

  • /-/static-plugins/PLUGIN_NAME/ now serves static/ from plugins

  • <th> now gets class=”col-X” - plus added col-X documentation

  • Use to_css_class for table cell column classes

    This ensures that columns with spaces in the name will still generate usable CSS class names. Refs #209

  • Add column name classes to <td>s, make PK bold [Russ Garrett]

  • Don’t duplicate simple primary keys in the link column [Russ Garrett]

    When there’s a simple (single-column) primary key, it looks weird to duplicate it in the link column.

    This change removes the second PK column and treats the link column as if it were the PK column from a header/sorting perspective.

  • Correct escaping for HTML display of row links [Russ Garrett]

  • Longer time limit for test_paginate_compound_keys

    It was failing intermittently in Travis - see #209

  • Use application/octet-stream for downloadable databases

  • Updated PyPI classifiers

  • Updated PyPI link to

0.19 (2018-04-16)

This is the first preview of the new Datasette plugins mechanism. Only two plugin hooks are available so far - for custom SQL functions and custom template filters. There’s plenty more to come - read the documentation and get involved in the tracking ticket if you have feedback on the direction so far.

  • Fix for _sort_desc=sortable_with_nulls test, refs #216

  • Fixed #216 - paginate correctly when sorting by nullable column

  • Initial documentation for plugins, closes #213

  • New --plugins-dir=plugins/ option (#212)

    New option causing Datasette to load and evaluate all of the Python files in the specified directory and register any plugins that are defined in those files.

    This new option is available for the following commands:

    datasette serve mydb.db --plugins-dir=plugins/
    datasette publish now/heroku mydb.db --plugins-dir=plugins/
    datasette package mydb.db --plugins-dir=plugins/
  • Start of the plugin system, based on pluggy (#210)

    Uses originally created for the py.test project

    We’re starting with two plugin hooks:


    This is called when a new SQLite connection is created. It can be used to register custom SQL functions.


    This is called with the Jinja2 environment. It can be used to register custom template tags and filters.

    An example plugin which uses these two hooks can be found at or installed using pip install datasette-plugin-demos

    Refs #14

  • Return HTTP 405 on InvalidUsage rather than 500. [Russ Garrett]

    This also stops it filling up the logs. This happens for HEAD requests at the moment - which perhaps should be handled better, but that’s a different issue.

0.18 (2018-04-14)

This release introduces support for units, contributed by Russ Garrett (#203). You can now optionally specify the units for specific columns using metadata.json. Once specified, units will be displayed in the HTML view of your table. They also become available for use in filters - if a column is configured with a unit of distance, you can request all rows where that column is less than 50 meters or more than 20 feet for example.

  • Link foreign keys which don’t have labels. [Russ Garrett]

    This renders unlabeled FKs as simple links.

    Also includes bonus fixes for two minor issues:

    • In foreign key link hrefs the primary key was escaped using HTML escaping rather than URL escaping. This broke some non-integer PKs.
    • Print tracebacks to console when handling 500 errors.
  • Fix SQLite error when loading rows with no incoming FKs. [Russ Garrett]

    This fixes ERROR: conn=<sqlite3.Connection object at 0x10bbb9f10>, sql = 'select ', params = {'id': '1'} caused by an invalid query when loading incoming FKs.

    The error was ignored due to async but it still got printed to the console.

  • Allow custom units to be registered with Pint. [Russ Garrett]

  • Support units in filters. [Russ Garrett]

  • Tidy up units support. [Russ Garrett]

    • Add units to exported JSON
    • Units key in metadata skeleton
    • Docs
  • Initial units support. [Russ Garrett]

    Add support for specifying units for a column in metadata.json and rendering them on display using pint

0.17 (2018-04-13)

  • Release 0.17 to fix issues with PyPI

0.16 (2018-04-13)

  • Better mechanism for handling errors; 404s for missing table/database

    New error mechanism closes #193

    404s for missing tables/databases closes #184

  • long_description in markdown for the new PyPI

  • Hide Spatialite system tables. [Russ Garrett]

  • Allow explain select / explain query plan select #201

  • Datasette inspect now finds primary_keys #195

  • Ability to sort using form fields (for mobile portrait mode) #199

    We now display sort options as a select box plus a descending checkbox, which means you can apply sort orders even in portrait mode on a mobile phone where the column headers are hidden.

0.15 (2018-04-09)

The biggest new feature in this release is the ability to sort by column. On the table page the column headers can now be clicked to apply sort (or descending sort), or you can specify ?_sort=column or ?_sort_desc=column directly in the URL.

  • table_rows => table_rows_count, filtered_table_rows => filtered_table_rows_count

    Renamed properties. Closes #194

  • New sortable_columns option in metadata.json to control sort options.

    You can now explicitly set which columns in a table can be used for sorting using the _sort and _sort_desc arguments using metadata.json:

        "databases": {
            "database1": {
                "tables": {
                    "example_table": {
                        "sortable_columns": [

    Refs #189

  • Column headers now link to sort/desc sort - refs #189

  • _sort and _sort_desc parameters for table views

    Allows for paginated sorted results based on a specified column.

    Refs #189

  • Total row count now correct even if _next applied

  • Use .custom_sql() for _group_count implementation (refs #150)

  • Make HTML title more readable in query template (#180) [Ryan Pitts]

  • New ?_shape=objects/object/lists param for JSON API (#192)

    New _shape= parameter replacing old .jsono extension

    Now instead of this:


    We use the _shape parameter like this:


    Also introduced a new _shape called object which looks like this:


    Returning an object for the rows key:

    "rows": {
        "pk1": {
        "pk2": {

    Refs #122

  • Utility for writing test database fixtures to a .db file

    python tests/ /tmp/hello.db

    This is useful for making a SQLite database of the test fixtures for interactive exploration.

  • Compound primary key _next= now plays well with extra filters

    Closes #190

  • Fixed bug with keyset pagination over compound primary keys

    Refs #190

  • Database/Table views inherit source/license/source_url/license_url metadata

    If you set the source_url/license_url/source/license fields in your root metadata those values will now be inherited all the way down to the database and table templates.

    The title/description are NOT inherited.

    Also added unit tests for the HTML generated by the metadata.

    Refs #185

  • Add metadata, if it exists, to heroku temp dir (#178) [Tony Hirst]

  • Initial documentation for pagination

  • Broke up test_app into test_api and test_html

  • Fixed bug with .json path regular expression

    I had a table called geojson and it caused an exception because the regex was matching .json and not \.json

  • Deploy to Heroku with Python 3.6.3

0.14 (2017-12-09)

The theme of this release is customization: Datasette now allows every aspect of its presentation to be customized either using additional CSS or by providing entirely new templates.

Datasette’s metadata.json format has also been expanded, to allow per-database and per-table metadata. A new datasette skeleton command can be used to generate a skeleton JSON file ready to be filled in with per-database and per-table details.

The metadata.json file can also be used to define canned queries, as a more powerful alternative to SQL views.

  • extra_css_urls/extra_js_urls in metadata

    A mechanism in the metadata.json format for adding custom CSS and JS urls.

    Create a metadata.json file that looks like this:

        "extra_css_urls": [
        "extra_js_urls": [

    Then start datasette like this:

    datasette mydb.db --metadata=metadata.json

    The CSS and JavaScript files will be linked in the <head> of every page.

    You can also specify a SRI (subresource integrity hash) for these assets:

        "extra_css_urls": [
                "url": "",
                "sri": "sha384-9qIZekWUyjCyDIf2YK1FRoKiPJq4PHt6tp/ulnuuyRBvazd0hG7pWbE99zvwSznI"
        "extra_js_urls": [
                "url": "",
                "sri": "sha256-k2WSCIexGzOj3Euiig+TlR8gA0EmPjuc79OEeY5L45g="

    Modern browsers will only execute the stylesheet or JavaScript if the SRI hash matches the content served. You can generate hashes using

  • Auto-link column values that look like URLs (#153)

  • CSS styling hooks as classes on the body (#153)

    Every template now gets CSS classes in the body designed to support custom styling.

    The index template (the top level page at /) gets this:

    <body class="index">

    The database template (/dbname/) gets this:

    <body class="db db-dbname">

    The table template (/dbname/tablename) gets:

    <body class="table db-dbname table-tablename">

    The row template (/dbname/tablename/rowid) gets:

    <body class="row db-dbname table-tablename">

    The db-x and table-x classes use the database or table names themselves IF they are valid CSS identifiers. If they aren’t, we strip any invalid characters out and append a 6 character md5 digest of the original name, in order to ensure that multiple tables which resolve to the same stripped character version still have different CSS classes.

    Some examples (extracted from the unit tests):

    "simple" => "simple"
    "MixedCase" => "MixedCase"
    "-no-leading-hyphens" => "no-leading-hyphens-65bea6"
    "_no-leading-underscores" => "no-leading-underscores-b921bc"
    "no spaces" => "no-spaces-7088d7"
    "-" => "336d5e"
    "no $ characters" => "no--characters-59e024"
  • datasette --template-dir=mytemplates/ argument

    You can now pass an additional argument specifying a directory to look for custom templates in.

    Datasette will fall back on the default templates if a template is not found in that directory.

  • Ability to over-ride templates for individual tables/databases.

    It is now possible to over-ride templates on a per-database / per-row or per- table basis.

    When you access e.g. /mydatabase/mytable Datasette will look for the following:

    - table-mydatabase-mytable.html
    - table.html

    If you provided a --template-dir argument to datasette serve it will look in that directory first.

    The lookup rules are as follows:

    Index page (/):
    Database page (/mydatabase):
    Table page (/mydatabase/mytable):
    Row page (/mydatabase/mytable/id):

    If a table name has spaces or other unexpected characters in it, the template filename will follow the same rules as our custom <body> CSS classes - for example, a table called “Food Trucks” will attempt to load the following templates:


    It is possible to extend the default templates using Jinja template inheritance. If you want to customize EVERY row template with some additional content you can do so by creating a row.html template like this:

    {% extends "default:row.html" %}
    {% block content %}
    <p>This line renders the original block:</p>
    {{ super() }}
    {% endblock %}
  • --static option for datasette serve (#160)

    You can now tell Datasette to serve static files from a specific location at a specific mountpoint.

    For example:

    datasette serve mydb.db --static extra-css:/tmp/static/css

    Now if you visit this URL:


    The following file will be served:

  • Canned query support.

    Named canned queries can now be defined in metadata.json like this:

        "databases": {
            "timezones": {
                "queries": {
                    "timezone_for_point": "select tzid from timezones ..."

    These will be shown in a new “Queries” section beneath “Views” on the database page.

  • New datasette skeleton command for generating metadata.json (#164)

  • metadata.json support for per-table/per-database metadata (#165)

    Also added support for descriptions and HTML descriptions.

    Here’s an example metadata.json file illustrating custom per-database and per- table metadata:

        "title": "Overall datasette title",
        "description_html": "This is a <em>description with HTML</em>.",
        "databases": {
            "db1": {
                "title": "First database",
                "description": "This is a string description & has no HTML",
                "license_url": "",
            "license": "The example license",
                "queries": {
                  "canned_query": "select * from table1 limit 3;"
                "tables": {
                    "table1": {
                        "title": "Custom title for table1",
                        "description": "Tables can have descriptions too",
                        "source": "This has a custom source",
                        "source_url": ""
  • Renamed datasette build command to datasette inspect (#130)

  • Upgrade to Sanic 0.7.0 (#168)

  • Package and publish commands now accept --static and --template-dir

    Example usage:

    datasette package --static css:extra-css/ --static js:extra-js/ \
      sf-trees.db --template-dir templates/ --tag sf-trees --branch master

    This creates a local Docker image that includes copies of the templates/, extra-css/ and extra-js/ directories. You can then run it like this:

    docker run -p 8001:8001 sf-trees

    For publishing to Zeit now:

    datasette publish now --static css:extra-css/ --static js:extra-js/ \
      sf-trees.db --template-dir templates/ --name sf-trees --branch master
  • HTML comment showing which templates were considered for a page (#171)

0.13 (2017-11-24)

  • Search now applies to current filters.

    Combined search into the same form as filters.

    Closes #133

  • Much tidier design for table view header.

    Closes #147

  • Added ?column__not=blah filter.

    Closes #148

  • Row page now resolves foreign keys.

    Closes #132

  • Further tweaks to select/input filter styling.

    Refs #86 - thanks for the help, @natbat!

  • Show linked foreign key in table cells.

  • Added UI for editing table filters.

    Refs #86

  • Hide FTS-created tables on index pages.

    Closes #129

  • Add publish to heroku support [Jacob Kaplan-Moss]

    datasette publish heroku mydb.db

    Pull request #104

  • Initial implementation of ?_group_count=column.

    URL shortcut for counting rows grouped by one or more columns.

    ?_group_count=column1&_group_count=column2 works as well.

    SQL generated looks like this:

    select "qSpecies", count(*) as "count"
    from Street_Tree_List
    group by "qSpecies"
    order by "count" desc limit 100

    Or for two columns like this:

    select "qSpecies", "qSiteInfo", count(*) as "count"
    from Street_Tree_List
    group by "qSpecies", "qSiteInfo"
    order by "count" desc limit 100

    Refs #44

  • Added --build=master option to datasette publish and package.

    The datasette publish and datasette package commands both now accept an optional --build argument. If provided, this can be used to specify a branch published to GitHub that should be built into the container.

    This makes it easier to test code that has not yet been officially released to PyPI, e.g.:

    datasette publish now mydb.db --branch=master
  • Implemented ?_search=XXX + UI if a FTS table is detected.

    Closes #131

  • Added datasette --version support.

  • Table views now show expanded foreign key references, if possible.

    If a table has foreign key columns, and those foreign key tables have label_columns, the TableView will now query those other tables for the corresponding values and display those values as links in the corresponding table cells.

    label_columns are currently detected by the inspect() function, which looks for any table that has just two columns - an ID column and one other - and sets the label_column to be that second non-ID column.

  • Don’t prevent tabbing to “Run SQL” button (#117) [Robert Gieseke]

    See comment in #115

  • Add keyboard shortcut to execute SQL query (#115) [Robert Gieseke]

  • Allow --load-extension to be set via environment variable.

  • Add support for ?field__isnull=1 (#107) [Ray N]

  • Add spatialite, switch to debian and local build (#114) [Ariel Núñez]

  • Added --load-extension argument to datasette serve.

    Allows loading of SQLite extensions. Refs #110.

0.12 (2017-11-16)

  • Added __version__, now displayed as tooltip in page footer (#108).

  • Added initial docs, including a changelog (#99).

  • Turned on auto-escaping in Jinja.

  • Added a UI for editing named parameters (#96).

    You can now construct a custom SQL statement using SQLite named parameters (e.g. :name) and datasette will display form fields for editing those parameters. Here’s an example which lets you see the most popular names for dogs of different species registered through various dog registration schemes in Australia.

  • Pin to specific Jinja version. (#100).

  • Default to not (#98).

  • Added extra metadata options to publish and package commands. (#92).

    You can now run these commands like so:

    datasette now publish mydb.db \
        --title="My Title" \
        --source="Source" \
        --source_url="" \
        --license="CC0" \

    This will write those values into the metadata.json that is packaged with the app. If you also pass --metadata=metadata.json that file will be updated with the extra values before being written into the Docker image.

  • Added simple production-ready Dockerfile (#94) [Andrew Cutler]

  • New ?_sql_time_limit_ms=10 argument to database and table page (#95)

  • SQL syntax highlighting with Codemirror (#89) [Tom Dyson]

0.11 (2017-11-14)

  • Added datasette publish now --force option.

    This calls now with --force - useful as it means you get a fresh copy of datasette even if Now has already cached that docker layer.

  • Enable --cors by default when running in a container.

0.10 (2017-11-14)

  • Fixed #83 - 500 error on individual row pages.

  • Stop using sqlite WITH RECURSIVE in our tests.

    The version of Python 3 running in Travis CI doesn’t support this.

0.9 (2017-11-13)

  • Added --sql_time_limit_ms and --extra-options.

    The serve command now accepts --sql_time_limit_ms for customizing the SQL time limit.

    The publish and package commands now accept --extra-options which can be used to specify additional options to be passed to the datasite serve command when it executes inside the resulting Docker containers.

0.8 (2017-11-13)

  • V0.8 - added PyPI metadata, ready to ship.

  • Implemented offset/limit pagination for views (#70).

  • Improved pagination. (#78)

  • Limit on max rows returned, controlled by --max_returned_rows option. (#69)

    If someone executes ‘select * from table’ against a table with a million rows in it, we could run into problems: just serializing that much data as JSON is likely to lock up the server.

    Solution: we now have a hard limit on the maximum number of rows that can be returned by a query. If that limit is exceeded, the server will return a "truncated": true field in the JSON.

    This limit can be optionally controlled by the new --max_returned_rows option. Setting that option to 0 disables the limit entirely.