I'm using the FTS methods in sqlite-utils for this website: drwn.io. I wanted to get pagination to have some kind of infinite scrolling in the landing page, and I ended up using that.

I tried deployment in WSL. It is working fine https://jmeter.vercel.app/

rs1333049 associated with coronary artery disease
https://www.snpedia.com/index.php/Rs1333049

select rsid, genotype, case genotype
  when 'CC' then '1.9x increased risk for coronary artery disease'
  when 'CG' then '1.5x increased risk for CAD'
  when 'GG' then 'normal'
end as interpretation from genome where rsid = 'rs1333049'

rs53576: the oxytocin receptor (OXTR) gene

select rsid, genotype, case genotype
  when 'AA' then 'Lack of empathy?'
  when 'AG' then 'Lack of empathy?'
  when 'GG' then 'Optimistic and empathetic; handle stress well'
end as interpretation from genome where rsid = 'rs53576'

Both rs10757274 and rs2383206 can both indicate higher risks of heart disease
https://www.snpedia.com/index.php/Rs2383206

select rsid, genotype, case genotype
  when 'AA' then 'Normal'
  when 'AG' then '~1.2x increased risk for heart disease'
  when 'GG' then '~1.3x increased risk for heart disease'
end as interpretation from genome where rsid = 'rs10757274'

select rsid, genotype, case genotype
  when 'AA' then 'Normal'
  when 'AG' then '1.4x increased risk for heart disease'
  when 'GG' then '1.7x increased risk for heart disease'
end as interpretation from genome where rsid = 'rs2383206'

rs7903146 Influences risk of Type-2 diabetes
https://www.snpedia.com/index.php/Rs7903146

select rsid, genotype, case genotype
  when 'CC' then 'Normal (lower) risk of Type 2 Diabetes and Gestational Diabetes.'
  when 'CT' then '1.4x increased risk for diabetes (and perhaps colon cancer).'
  when 'TT' then '2x increased risk for Type-2 diabetes'
end as interpretation from genome where rsid = 'rs7903146'

The "Warrior Gene" https://www.snpedia.com/index.php/Rs4680

select rsid, genotype, case genotype
  when 'AA' then '(worrier) advantage in memory and attention tasks'
  when 'AG' then 'Intermediate dopamine levels, other effects'
  when 'GG' then '(warrior) multiple associations, see details'
end as interpretation from genome where rsid = 'rs4680'

Risk of autoimmune disorders: https://www.snpedia.com/index.php/Genotype

select rsid, genotype, case genotype
  when 'AA' then '2x risk of rheumatoid arthritis and other autoimmune diseases'
  when 'GG' then 'Normal risk for autoimmune disorders'
end as interpretation from genome where rsid = 'rs2476601'

I encountered your issue when trying to find a solution and came up with the following, maybe it can help.

import logging.config
from typing import Tuple

import structlog
import uvicorn

from example.config import settings

shared_processors: Tuple[structlog.types.Processor, ...] = (
    structlog.contextvars.merge_contextvars,
    structlog.stdlib.add_logger_name,
    structlog.stdlib.add_log_level,
    structlog.processors.TimeStamper(fmt="iso"),
)

logging_config = {
    "version": 1,
    "disable_existing_loggers": False,
    "formatters": {
        "json": {
            "()": structlog.stdlib.ProcessorFormatter,
            "processor": structlog.processors.JSONRenderer(),
            "foreign_pre_chain": shared_processors,
        },
        "console": {
            "()": structlog.stdlib.ProcessorFormatter,
            "processor": structlog.dev.ConsoleRenderer(),
            "foreign_pre_chain": shared_processors,
        },
        **uvicorn.config.LOGGING_CONFIG["formatters"],
    },
    "handlers": {
        "default": {
            "level": "DEBUG",
            "class": "logging.StreamHandler",
            "formatter": "json" if not settings.debug else "console",
        },
        "uvicorn.access": {
            "level": "INFO",
            "class": "logging.StreamHandler",
            "formatter": "access",
        },
        "uvicorn.default": {
            "level": "INFO",
            "class": "logging.StreamHandler",
            "formatter": "default",
        },
    },
    "loggers": {
        "": {"handlers": ["default"], "level": "INFO"},
        "uvicorn.error": {
            "handlers": ["default" if not settings.debug else "uvicorn.default"],
            "level": "INFO",
            "propagate": False,
        },
        "uvicorn.access": {
            "handlers": ["default" if not settings.debug else "uvicorn.access"],
            "level": "INFO",
            "propagate": False,
        },
    },
}


def setup_logging() -> None:
    structlog.configure(
        processors=[
            structlog.stdlib.filter_by_level,
            *shared_processors,
            structlog.stdlib.PositionalArgumentsFormatter(),
            structlog.processors.StackInfoRenderer(),
            structlog.processors.format_exc_info,
            structlog.stdlib.ProcessorFormatter.wrap_for_formatter,
        ],
        context_class=dict,
        logger_factory=structlog.stdlib.LoggerFactory(),
        wrapper_class=structlog.stdlib.AsyncBoundLogger,
        cache_logger_on_first_use=True,
    )
    logging.config.dictConfig(logging_config)

And then it'll be run on the startup event:

@app.on_event("startup")
async def startup_event() -> None:
    setup_logging()

It depends on a setting called debug which controls if it should output the regular uvicorn logging or json.

It might be possible with this library: https://docs.python.org/3/library/imghdr.html

quick test of the downloaded blob:

>>> import imghdr
>>> imghdr.what('material_culture-1-image.blob')
'jpeg'

The output plugin would be cool. I'll look into making my first datasette plugin. I'm also imagining displaying the image in the browser -- but that would be a step 2.

Hi Simon

Just finding this old issue regarding downloading blobs. Nice work!

https://user-images.githubusercontent.com/4997607/104946741-df4bad80-59ba-11eb-96e3-727c85cc4dc6.png">

As a feature request, maybe it would be possible to assign a blob column as a certain data type (e.g. .jpg) and then each blob could be downloaded as that type of file (perhaps if the file types were constrained to normal blobs that people store in sqlite databases, this could avoid the execution stuff mentioned above).

I guess the column blob-type definition could fit into this dropdown selection:

https://user-images.githubusercontent.com/4997607/104947000-479a8f00-59bb-11eb-87d9-1644e5940894.png">

Let me know if I should open a new issue with a feature request. (This could slowly go in the direction of displaying image blob-types in the browser.)

Thanks for the great tool!

edit: just reading the rest of the twitter thread: https://twitter.com/simonw/status/1318685933256855552

perhaps this is already possible in some form with the plugin datasette-media: https://github.com/simonw/datasette-media

Make sense. If you're coming from the sqlite3 side of things, rather than the datasette side, wanting the fts methods to work for views makes more sense. sqlite3 allows fts5 tables on views, so I was looking for CLI functionality to build the fts virtual tables. Ultimately, though, sharing fts virtual tables across tables and derivative views is likely more efficient.

Maybe an explicit error message like, "fts is not supported for views" rather than just throwing an exception that the method doesn't exist" might be helpful. Not critical though.

Thanks.

I found and fixed another bug, this one around importing the tweets table. @simonw let me know if you'd prefer this broken out into multiple PRs, happy to do that if it makes review/merging easier.

@simonw : Did you have the time to take a look at this ?

Y, thank you to both of you!

Great, really happy I could help! Reverse proxies get tricky.

+1

Yep! Fixes it. If I navigate to https://corpora.tika.apache.org/datasette, I get a 404 (database not found: datasette), but if I navigate to https://corpora.tika.apache.org/datasette/file_profiles/, everything WORKS!

Thank you!

I found this issue while troubleshooting the same behavior with an nginx reverse proxy. The solution was to make sure I set:

proxy_pass http://server:8001/baseurl/
instead of just:

proxy_pass http://server:8001
The custom SQL query and header links are now correct.

We did this for Sno under macOS — it's a PyInstaller binary/setup which uses Packages for packaging.

• Building & Signing
• Packaging & Notarizing
• Github Workflow has the CI side of it

FYI (if you ever get to it) for Windows you need to get a code signing certificate. And if you want automated CI, you'll want to get an "EV CodeSigning for HSM" certificate from GlobalSign, which then lets you put the certificate into Azure Key Vault. Which you can use with azuresigntool to sign your code & installer. (Non-EV certificates are a waste of time, the user still gets big warnings at install time).

I was able to fix this, at least enough to get my archive to import. Not sure if there's more work to be done here or not.

This now works for me, though I'm entirely ensure if it's a just-my-export thing or a wider issue. Also, this doesn't contain any tests. So I'm not sure if there's more work to be done here, or if this is good enough.

Correction: the failure is on lists-member.js (I was thrown by the block variable name, but that's just a coincidence)

For reasons I've written about elsewhere, I'm in favor of modules. It has several beneficial effects. One, old browsers just ignore it all together. Two, if you include the same plain script on the page more than once, it will be executed twice, but if you include the same module script on a page twice, it will only execute once. Three, you get a module local namespace, instead of having to use the global window namespace or a function private namespace.

OTOH, if you are going to use an old style script, the code from before isn't ideal, because you wipe out your registry if the script it included more than once. Also you may as well use object methods and splat arguments.

The event based architecture probably makes more sense though. Just make up some event names prefixed with datasette: and listen for them on the root. The only concern with that approach is it can sometimes be tricky to make sure your plugins are run after datasette has run. Maybe

function mycallback(){
  // whatever
}

if (window.datasette) {
  window.datasette.init(mycallback);
} else {
  document.addEventListener('datasette:init', mycallback);
}

I was thinking JavaScript plugins going with server side template extensions custom HTML. Attach my own widgets on there and listen for Datasette events to refresh when user interacts with main UI. Like a map view or table that updates according to selected column. There's certainly other ways to look at this. Perhaps you could list possible hooks or high level design doc on what would be possible with the plugin system?

Re: modules. I would like to see modules supported at least in development. The developer experience is so much better than what JavaScript coding has been in the past. With large parts of NPM at your disposal I’d imagine even less experienced coder can whisk a custom plugin in no time. Proper production build system (like one you get with Pika or Parcel) could package everything up into bundles that older browsers can understand. Though that does come with performance and size penalties alongside the added complexity.

Codecov Report

Merging #1170 (a5761cc) into main (1e8fa3a) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1170   +/-   ##
=======================================
  Coverage   91.55%   91.55%           
=======================================
  Files          32       32           
  Lines        3932     3932           
=======================================
  Hits         3600     3600           
  Misses        332      332           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1e8fa3a...a5761cc. Read the comment docs.

With regards to JS/Browser events, given your example of menu items that plugins could add, I could imagine this code to work:

// as part of datasette
datasette.events.AddMenuItem = 'DatasetteAddMenuItemEvent';
document.addEventListener(datasette.events.AddMenuItem, (e) => {
  // do whatever is needed to add the menu item. Data comes from `e`
  alert(e.title + ' ' + e.link);
});

// as part of a plugin
const event = new Event(datasette.events.AddMenuItem, {link: '/foo/bar', title: 'Go somewhere'});
Document.dispatchEvent(event)

but I'm already commited to requiring support for () => {} arrow functions

Don't think you are :) (e.g. gzipped, using arrow functions in my example saves 2 bytes over spelling out function). On FMS, past month, looking at popular browsers, looks like we'd have 95.41% arrow support, 94.19% module support, and 4.58% (mostly IE9/IE11/Safari 9) supporting neither.

I have yet to build Datasette plugin and am unfamiliar with Pluggy. Since browsers have event handling builtin Datasette could communicate with plugins through it. Handlers register as listeners for custom Datasette events and Datasette's JS can then trigger said events.

I was also wondering if you had looked at Javascript Modules for JS plugins? With services like Skypack (https://www.skypack.dev) NPM libraries can be loaded directly into browser, no build step needed. Same goes for local JS if you adhere to ES Module spec.

If minification is required then tools such as Snowpack (https://www.snowpack.dev) could fit better. It uses https://github.com/evanw/esbuild for bundling and minification.

On plugins you'd simply:

import {register} from '/assets/js/datasette'
register.on({'click' : my_func})

In Datasette HTML pages' head you'd merely import these files as modules one by one.

Amazing work! And you've put in far more work than I'd expect to reduce the payload (which is admirable).

So, to add a plugin with the current design, it goes in (a) the template or (b) a bookmarklet, right?

Sorry to go on about it, but it's my only example ;) And thought it might be of interest/use. Here is FixMyStreet's Cypress workflow https://github.com/mysociety/fixmystreet/blob/master/.github/workflows/cypress.yml with the master script that sets up server etc at https://github.com/mysociety/fixmystreet/blob/master/bin/browser-tests (that has features such as working inside/outside Vagrant, and can do JS code coverage) and then the tests are at https://github.com/mysociety/fixmystreet/tree/master/.cypress/cypress/integration

If you're using arrow functions, you can presumably use default parameters, not much difference in support. That would save you 9 bytes. But OTOH you need "use strict"; to use arrow functions etc, and that's 13 bytes.

Your latest 250-byte one, with use strict, gzips to 199 bytes. The following might be 292 bytes, but compresses to 204, basically the same, and works in any browser (well, IE9+) at all:

var datasette=datasette||{};datasette.plugins=function(){var d={};return{register:function(b,c,e){d[b]||(d[b]=[]);d[b].push([c,e])},call:function(b,c){c=c||{};var e=[];(d[b]||[]).forEach(function(a){a=a[0].apply(a[0],a[1].map(function(a){return c[a]}));void 0!==a&&e.push(a)});return e}}}();

Source for that is below; I replaced the [fn,parameters] because closure-compiler includes a polyfill for that, and I ran closure-compiler --language_out ECMASCRIPT3:

var datasette = datasette || {};
datasette.plugins = (() => {
    var registry = {};
    return {
        register: (hook, fn, parameters) => {
            if (!registry[hook]) {
                registry[hook] = [];
            }
            registry[hook].push([fn, parameters]);
        },
        call: (hook, args) => {
            args = args || {};
            var results = [];
            (registry[hook] || []).forEach((data) => {
                /* Call with the correct arguments */
                var result = data[0].apply(data[0], data[1].map(parameter => args[parameter]));
                if (result !== undefined) {
                    results.push(result);
                }
            });
            return results;
        }
    };
})();

If you could say that all hook functions had to accept one options parameter (and could use object destructuring if they wished to only see a subset), you could have this, which minifies (to all-browser-JS) to 200 bytes, gzips to 146, and works practically the same:

var datasette = datasette || {};
datasette.plugins = (() => {
    var registry = {};
    return {
        register: (hook, fn) => {
            registry[hook] = registry[hook] || [];
            registry[hook].push(fn);
        },
        call: (hook, args) => {
            var results = (registry[hook] || []).map(fn => fn(args||{}));
            return results;
        }
    };
})();

var datasette=datasette||{};datasette.plugins=function(){var b={};return{register:function(a,c){b[a]=b[a]||[];b[a].push(c)},call:function(a,c){return(b[a]||[]).map(function(a){return a(c||{})})}}}();

Called the same, definitions tiny bit different:

datasette.plugins.register('numbers', ({a, b}) => a + b)
datasette.plugins.register('numbers', o => o.a * o.b)
datasette.plugins.call('numbers', {a: 4, b: 6})

Very much looking forward to seeing this functionality come together. This is probably out-of-scope for an initial release, but in the future it could be useful to also think of how to run this is a container'ized context. For example, an immutable datasette container that points to an S3 bucket of SQLite DBs or CSVs. Or an immutable datasette container pointing to a NFS volume elsewhere on a Kubernetes cluster.

I like the idea of _internal, it's a nice way to get a data catalog quickly. I wonder if this trick applies to db other than SQLite.

Hi @simonw, do I need to do anything else for this PR to be considered to be included? I've tried using this project and it is quite nice to be able to explore a repository, but noticed that a couple commands don't allow you to use authorization from the environment variable.

Instead of scanning the directory every 10s, have you considered listening for the native system events to notify you of updates?

I think python has a nice module to do this for you called watchdog

This comes around if you’ve run the photo export without running an s3 upload.

Codecov Report

Merging #1159 (c820abd) into main (a882d67) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1159   +/-   ##
=======================================
  Coverage   91.55%   91.55%           
=======================================
  Files          32       32           
  Lines        3930     3930           
=======================================
  Hits         3598     3598           
  Misses        332      332           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a882d67...c820abd. Read the comment docs.

Codecov Report

Merging #1158 (37ce72f) into main (90eba4c) will not change coverage.
The diff coverage is 87.50%.

@@           Coverage Diff           @@
##             main    #1158   +/-   ##
=======================================
  Coverage   91.55%   91.55%           
=======================================
  Files          32       32           
  Lines        3930     3930           
=======================================
  Hits         3598     3598           
  Misses        332      332           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 90eba4c...37ce72f. Read the comment docs.

Hey Simon! I really enjoy datasette so far, just started trying it out today following your iPhone photos example.

I am not sure if you had run into this or not, but it seems like they might have changed one of the column names from
ZGENERICASSET to ZASSET. Should I open a PR?

Would change:
- here
- here
- here

I think this should probably be closed as won't fix.

Attempting to make a patch for this I realized that the since_id would limit to tweets posted since that since_id, not when it was favorited. So favoriting something in the older would be missed if you used --since with a cron script

Better to just use --stop_after set to a couple hundred

I have also run into this a bit, would it be possible to post your requirements.txt so I can try and reproduce your blog post?

it looks like almost all of the memory consumption is coming from json.load().

another direction here may be to use the new "Semantic Location History" data which is already broken down by year and month.

it also provides much more interesting data, such as estimated address, form of travel, etc.

You're right, probably more straightforward to have the links for JSON. I was imagining to toggle the href for the 'Export JSON' link (button) to the selected shape, but it'll probably be needlessly complex in the end.

From a quick look at the README, it does seem to do everything I need, thanks!

I think the argument for inclusion in core is to lower the chances of unwanted data access. A local server can be accessed by anybody who can make an HTTP request to your computer regardless of CORS rules, but the default * rule additionally opens up access to the local instance to any website you visit while it is running.

That's probably not what people typically intend, particularly when the data is of a sensitive nature. A default of requiring the user to specify the origin (allowing * but encouraging a narrower scope) would solve this problem entirely, I think.

Alright I could give it a try! This might be a stupid question, can you tell me how to run the server from my fork? So that I can test the changes?

Thanks for opening the issue, @simonw. Let me elaborate on my Tweets.

datasette-chartjs provides drop down lists to pick the chart visualization (e.g. bar, line, doughnut, pie, ...) as well as the column used for the "x axis" (e.g. time).

A user can change the values on-demand. The chart will be redrawn w/o querying the database again.

However, if a user wants to change the underlying query, they will use the SQL field provided by datasette or any of the other datasette built-in features to amend a query. In order to maintain a user's selections for the plugin, datasette-chartjs copies some parts of datasette-vega which persist the chosen visualization and column in the hash part of a URL (the stuff behind the #). The plugin load the config from the hash upon initialization on the next page and use it accordingly.

Additionally, datasette-vega and datasette-chartjs need to make sure to include the hash in all links and forms that cause a reload of the page. This is, such that the config persists between clicks.

This ticket is about moving thes parts into datasette that provide the functionality to do so. This includes:

1. a way to load config options with a given prefix from the current URL hash
2. a way to update the current URL hash with a new config value or a bunch of config options
3. updating all necessary links and forms on the current page to include the URL hash whenever its updated
4. to prevent leaking config options to external pages, only "internal" links should be updated

There's another, optional, feature that we might want to think about during the design phase: the scope of the config. Links within a datasette instance have 1 of 3 scopes:

1. global, for the whole datasette project
2. database, for all tables in a database
3. table, only for a table within a database

When updating the links and forms as pointed out in 3. above, it might be worth considering which links need to be updated. I could imagine a plugin that wants to persist some setting across all tables within a database but another setting only within a table.

Codecov Report

Merging #1145 (a8588f9) into main (0c616f7) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1145   +/-   ##
=======================================
  Coverage   91.41%   91.41%           
=======================================
  Files          31       31           
  Lines        3881     3881           
=======================================
  Hits         3548     3548           
  Misses        333      333           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0c616f7...a8588f9. Read the comment docs.

I'm looking into using datasette with a database with spatialite geometry columns, and came across this issue. Has there been any progress on this since 2018?

In one of my tables I'm just storing lat/lon points in a spatialite point geometry, and I've managed to make datasette-cluster-map display the points by extracting the lat and lon in SQL - using something like select ... ST_X(location) as longitude, ST_Y(location) as latitude from Blah. Something more 'built-in' would be great though - particularly for the tables I have that store more complex geometries.

Please let me know if there's anything I can do to help get this merged.

This is causing problems for me because it means when I build my Docker image my databases aren't considered immutable, which I would like them to be so that a download link is produced.

Agreed, it would definitely provide better controls. However, I do feel it makes for a bit of inconsistent UX for the 'Advanced export' section, with links to download for JSON, checkboxes and radio buttons + button to download for CSV. Do you think this example makes the UX a bit nicer/consistent?

I could give it a try if you'd like but I've never contributed to an actual project!

'Include all rows' seem like a fairly obvious alternative

Hello Simon,
I have a similar problem with horizontal scrollbar display with Datasette version 0.51 and superior for a table with more than 30 rows. With Datasette 0.50, the horizontal scrollbar is displayed, if I upgrade Datasette to 0.51 and superior, the horizontal scrollbar disappears.

Datasette 0.50: horizontal scrollbar

Datasette 0.51 and superior: no horizontal scrollbar

Thanks,

Bumping to ubuntu-20.04 appears to have solved my syntax error. 🤷

Hi Simon

Thank you for the quick fix! And glad you like our use of Datasette (launches 1. january 2021). It's a site that currently (more to come) makes all minutes and their annexes from Aarhus City Council and the major committees (1997-2019) available to the public. So we're putting Datasette to good use :)

I can't imagine this helps (esp. given your point about potential rewrites), but you can see that /datasette/ was correctly added to the sql form, but not to the "export-links"

https://user-images.githubusercontent.com/6739646/101680055-234baa00-3a2e-11eb-8650-2b369bc6f031.png">

I don't think this fixes it:

grep -R datasette .
./sites-available/000-default.conf:        ProxyPass /datasette http://127.0.0.1:8001/
./sites-available/000-default.conf:        #ProxyPassReverse /datasette http://127.0.0.1:8001/
./sites-available/corpora-le-ssl.conf:  ProxyPass /datasette http://0.0.0.0:8001
./sites-available/corpora-le-ssl.conf:  #ProxyPassReverse /datasette http://0.0.0.0:8001
./sites-enabled/corpora-le-ssl.conf:  ProxyPass /datasette http://0.0.0.0:8001
./sites-enabled/corpora-le-ssl.conf:  #ProxyPassReverse /datasette http://0.0.0.0:8001

And I confirmed that I actually restarted the server. :rofl:

https://corpora.tika.apache.org/datasette/file_profiles

Anything we can do to help debug this? Thank you, again!

Hi Simon. Any news on using wildcard-searches with datasette? Thanks!

hey! I'd like to take a look at this if you're open to a PR for it

Codecov Report

Merging #1130 (8d4c69c) into main (49d8fc0) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1130   +/-   ##
=======================================
  Coverage   91.42%   91.42%           
=======================================
  Files          31       31           
  Lines        3873     3873           
=======================================
  Hits         3541     3541           
  Misses        332      332           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 49d8fc0...8d4c69c. Read the comment docs.

Codecov Report

Merging #1128 (7004c3b) into main (49d8fc0) will decrease coverage by 0.00%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main    #1128      +/-   ##
==========================================
- Coverage   91.42%   91.42%   -0.01%     
==========================================
  Files          31       31              
  Lines        3873     3872       -1     
==========================================
- Hits         3541     3540       -1     
  Misses        332      332              

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 49d8fc0...7004c3b. Read the comment docs.

Are you thinking that those metadata tables would be added to the SQLite DB by Datasette, when you tell it to wrap up the database, with the metadata coming from the metadata.json? Would it be easy to allow the prepopulation of those tables in the database itself? We've been struggling with the best way to make sure that the data is always accompanied by metadata, and baking it all into the database itself would be nice, since then we wouldn't need to worry about separately distributing different files in different contexts.

Codecov Report

Merging #1122 (94ea22f) into main (a970276) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1122   +/-   ##
=======================================
  Coverage   91.49%   91.49%           
=======================================
  Files          31       31           
  Lines        3856     3856           
=======================================
  Hits         3528     3528           
  Misses        328      328           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a970276...94ea22f. Read the comment docs.

Are there common patterns for storing column-based metadata inside SQLite itself? I know Postgres allows "comment" fields, which this is kind of trying to replicate. Should the units and description and possibly other per-column metadata fields be combined into a single (tabular?) structure, that would be displayed above the data on the table / query results page?

Codecov Report

Merging #1120 (ddad8db) into main (461670a) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1120   +/-   ##
=======================================
  Coverage   91.48%   91.49%           
=======================================
  Files          31       31           
  Lines        3852     3856    +4     
=======================================
+ Hits         3524     3528    +4     
  Misses        328      328           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 461670a...ddad8db. Read the comment docs.

I just deployed this and its working great.

~In a very unscientific benchmark my response times went from around 22-25ms to 33-36ms, but I didn't even dig enough to confirm the latency is related to the change. It's on a VPS, so maybe the load changed.~ I don't see any difference in performance.

Codecov Report

Merging #1117 (ccdf2c6) into main (dea3c50) will decrease coverage by 0.00%.
The diff coverage is 95.23%.

@@            Coverage Diff             @@
##             main    #1117      +/-   ##
==========================================
- Coverage   91.48%   91.48%   -0.01%     
==========================================
  Files          30       31       +1     
  Lines        3841     3852      +11     
==========================================
+ Hits         3514     3524      +10     
- Misses        327      328       +1     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update dea3c50...ccdf2c6. Read the comment docs.

@simonw thanks for investigating so quickly. If it is undesirable to change that hidden behavior, maybe something like this is a suitable workaround:

SELECT * FROM pragma_table_xinfo('deeds') where hidden in (0,2);
0|body|TEXT|0||0|0
1|id|INT GENERATED ALWAYS|0||0|2
2|consideration|INT GENERATED ALWAYS|0||0|2

datasette-connectors provides an API for making connectors for any file based database. For example, datasette-pytables is a connector for HDF5 files, so now is possible to use this type of files with Datasette.

It'd be nice if Datasette coud provide that API directly, for other file formats and for urls too.

Thank you!

Codecov Report

Merging #1112 (1a30fc2) into main (37d18a5) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1112   +/-   ##
=======================================
  Coverage   91.44%   91.44%           
=======================================
  Files          30       30           
  Lines        3833     3833           
=======================================
  Hits         3505     3505           
  Misses        328      328           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 37d18a5...1a30fc2. Read the comment docs.

I can confirm this issue, running version 0.51.1 under Windows.

Fixed by commenting out the following line near the top of datasette\utils\asgi.py :

#from os import EX_CANTCREAT

Trying to run on Windows today, I get an error from the utils/asgi.py module.

It's trying from os import EX_CANTCREAT which is Unix-only. I commented this line out, and (so far) it's working.

I'm so sorry that you already have --since_id option and that's enough for the case I've mentioned. Thank you for this excellent tool!

We're using mod_proxy.

I don't think we are, but I'll check with Maruan.

I think this is the relevant part of our config?

  Alias "/base/" "/usr/share/corpora/"
  <Directory "/usr/share/corpora/">
    Options +Indexes -Multiviews
    AllowOverride None
  </Directory>

  ProxyPreserveHost On

  ProxyPass /datasette http://0.0.0.0:8001
  ProxyPassReverse /datasette http://0.0.0.0:8001

</VirtualHost>

Codecov Report

Merging #1097 (e89211d) into main (5eb8e9b) will not change coverage.
The diff coverage is 84.61%.

@@           Coverage Diff           @@
##             main    #1097   +/-   ##
=======================================
  Coverage   91.38%   91.38%           
=======================================
  Files          30       30           
  Lines        3785     3785           
=======================================
  Hits         3459     3459           
  Misses        326      326           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5eb8e9b...e89211d. Read the comment docs.

My headers aren't clickable/sortable with custom sql, but I think that's by design.

In the default view, https://corpora.tika.apache.org/datasette/file_profiles/file_profiles, ah, y, now I see that the headers should be sortable, but you're right the base_url is not applied.

base_url works with "View and Edit SQL" and with "(advanced)"

As you point out, does not work with the export csv, json, other or with the "Next page" navigational button at the bottom.

I'm starting this with docker like so:

docker run --name datasette -d -p 8001:8001 -vpwd:/mnt datasetteproject/datasette datasette -p 8001 -h 0.0.0.0 /mnt/file_profiles.db --config sql_time_limit_ms:120000 --config max_returned_rows:100000 --config base_url:/datasette/ --config cache_size_kb:50000

I'm not doing any templating or anything else custom.

Apropos of nothing, I swapped out a simpler db, so this query should now work:

https://corpora.tika.apache.org/datasette/file_profiles?sql=select%0D%0A++*%0D%0Afrom%0D%0A++file_profiles+fp%0D%0Alimit%0D%0A++10

The same is true if I select advanced export and hit the 'export csv' at the bottom of the page.

Just tested with the latest Docker image, and it works pretty much everywhere! THANK YOU!

I did notice that if I try to export json or csv, the base is not applied. Not sure if I should reopen this issue or open a new one.

To see this, go here: https://corpora.tika.apache.org/datasette/corpora-metadata/REF_PARSE_EXCEPTION_TYPES

Click/hover over json or CSV and you'll see that the 'datasette' base is not included.

Again, many thanks!

Codecov Report

Merging #1085 (51e7651) into main (2a981e2) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1085   +/-   ##
=======================================
  Coverage   91.38%   91.38%           
=======================================
  Files          30       30           
  Lines        3785     3785           
=======================================
  Hits         3459     3459           
  Misses        326      326           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2a981e2...51e7651. Read the comment docs.

I think I tried the same db size on the following scenarios in Digital Ocean:
1. Basic ($5/month) with 512MB RAM
2. Basic ($10/month) with 1GB RAM
3. Pro ($12/month) with 1GB RAM

All such attempts conked out with "out of memory" errors

Hi & thanks for the note @simonw! I wish I had more time to play with (and contribute to) datasette. I know you don't need me to tell you that it's super cool :-)

Codecov Report

Merging #1069 into main will increase coverage by 0.01%.
The diff coverage is 95.83%.

@@            Coverage Diff             @@
##             main    #1069      +/-   ##
==========================================
+ Coverage   91.30%   91.32%   +0.01%     
==========================================
  Files          29       29              
  Lines        3736     3756      +20     
==========================================
+ Hits         3411     3430      +19     
- Misses        325      326       +1     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 222f79b...92f3840. Read the comment docs.

Codecov Report

Merging #1065 into main will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1065      +/-   ##
==========================================
+ Coverage   91.23%   91.27%   +0.03%     
==========================================
  Files          28       29       +1     
  Lines        3710     3724      +14     
==========================================
+ Hits         3385     3399      +14     
  Misses        325      325              

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1a861be...5f118b5. Read the comment docs.

Codecov Report

Merging #1061 into main will increase coverage by 0.07%.
The diff coverage is 96.87%.

@@            Coverage Diff             @@
##             main    #1061      +/-   ##
==========================================
+ Coverage   91.13%   91.20%   +0.07%     
==========================================
  Files          27       28       +1     
  Lines        3677     3697      +20     
==========================================
+ Hits         3351     3372      +21     
+ Misses        326      325       -1     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d6f9ff7...1196d08. Read the comment docs.

Unsolicited feedback for an unreleased feature of the current unreleased GitHub version (I casually wanted to access a blob row) – the existing #1036 route doesn't support special characters in database or table names (e.g. @() ). Maybe this is motivation for your new idea here.

Also I got this error/crash with my blob and wasn't able to get the file: https://gist.github.com/thadk/28ac32af0e88747ce9056c90b0b19d34

Codecov Report

Merging #1060 into main will increase coverage by 6.42%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1060      +/-   ##
==========================================
+ Coverage   84.71%   91.13%   +6.42%     
==========================================
  Files          28       27       -1     
  Lines        3957     3677     -280     
==========================================
- Hits         3352     3351       -1     
+ Misses        605      326     -279     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update abcf022...4725d46. Read the comment docs.

Codecov Report

Merging #1059 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1059   +/-   ##
=======================================
  Coverage   84.71%   84.71%           
=======================================
  Files          28       28           
  Lines        3957     3957           
=======================================
  Hits         3352     3352           
  Misses        605      605           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7d9fedc...e46327a. Read the comment docs.

Codecov Report

Merging #1056 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1056   +/-   ##
=======================================
  Coverage   84.70%   84.70%           
=======================================
  Files          28       28           
  Lines        3955     3955           
=======================================
  Hits         3350     3350           
  Misses        605      605           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 26bb4a2...a7b2aab. Read the comment docs.

Codecov Report

Merging #1049 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1049   +/-   ##
=======================================
  Coverage   84.72%   84.72%           
=======================================
  Files          28       28           
  Lines        3942     3942           
=======================================
  Hits         3340     3340           
  Misses        602      602           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 42f4851...50a743a. Read the comment docs.

Codecov Report

Merging #1040 into main will increase coverage by 0.10%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1040      +/-   ##
==========================================
+ Coverage   84.65%   84.76%   +0.10%     
==========================================
  Files          28       28              
  Lines        3924     3938      +14     
==========================================
+ Hits         3322     3338      +16     
+ Misses        602      600       -2     

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bf82b3d...4f3165f. Read the comment docs.

Codecov Report

Merging #1044 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1044   +/-   ##
=======================================
  Coverage   84.65%   84.65%           
=======================================
  Files          28       28           
  Lines        3924     3924           
=======================================
  Hits         3322     3322           
  Misses        602      602           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d0cc6f4...6453ab1. Read the comment docs.

Codecov Report

Merging #1043 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1043   +/-   ##
=======================================
  Coverage   84.65%   84.65%           
=======================================
  Files          28       28           
  Lines        3924     3924           
=======================================
  Hits         3322     3322           
  Misses        602      602           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d0cc6f4...dc4129c. Read the comment docs.

The bug is that currently when there are databases passed in, but no -i flag, e.g. in configuration directory mode, inclusion in inspect-data.json does not automatically cause databases to be considered immutable, as described in the documentation.

The reason is that the -i flag is specified multiple=True, which means when it is not passed in we will get an empty list [], rather than None. So the current code decides that no databases are immutable rather than falling back to inspect-data.json -- as is presumably intended.

Thanks. As I said, I think the result (being able to query tree structures like ancestors and descendants) is more important than the implementation, and I agree that this particular sqlite extension is too obscure. Just providing an sqlite utility to build or rebuild a transitive closure table might be more generically useful. I find that hierarchical data shows up pretty frequently in some data science problems.

Codecov Report

Merging #1038 into main will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1038   +/-   ##
=======================================
  Coverage   84.65%   84.65%           
=======================================
  Files          28       28           
  Lines        3924     3924           
=======================================
  Hits         3322     3322           
  Misses        602      602           

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 66120a7...7fc0cce. Read the comment docs.

Here is one quick sketch (done in Figma :P) for an idea: a possible filter to switch between showing all tables from all databases, or grouping tables by database.

(the switch is interactive)

All tables: https://www.figma.com/proto/BjFrMroEtmVx6EeRjvSrox/Datasette-test?node-id=1%3A2&viewport=536%2C348%2C0.5&scaling=min-zoom

Grouped: https://www.figma.com/proto/BjFrMroEtmVx6EeRjvSrox/Datasette-test?node-id=3%3A974&viewport=536%2C348%2C0.5&scaling=min-zoom

When only 1 database: https://www.figma.com/proto/BjFrMroEtmVx6EeRjvSrox/Datasette-test?node-id=1%3A162&viewport=536%2C348%2C0.5&scaling=min-zoom

Is this is useful, I can send some more suggestions/sketches.