Compare commits

..

3 Commits

Author SHA1 Message Date
deme
1e27410f50 [examples/openspending,home][m]: add min date and max date facets 2023-05-17 20:50:45 -03:00
deme
8cb3cd4ddb [examples/openspending,home][xl]: removes datasets table, implement dataset cards grid, implement country facet 2023-05-17 18:43:19 -03:00
Luccas Mateus de Medeiros Gomes
902e5e07a0 [examples/openspending][m] - added loader + fetching from datapackage
- Also added an indexing example
2023-05-17 14:57:50 -03:00
37 changed files with 572 additions and 3405 deletions

View File

@@ -1,5 +1,8 @@
{
"extends": ["next", "next/core-web-vitals"],
"extends": [
"next",
"next/core-web-vitals"
],
"ignorePatterns": ["!**/*", ".next/**/*"],
"overrides": [
{

View File

@@ -1,7 +0,0 @@
node_modules
**/.next/**
**/_next/**
**/dist/**
**/__tmp__/**
lerna.json
.github

View File

@@ -1 +0,0 @@
{}

View File

@@ -99,3 +99,4 @@ And run the production build with:
```
npm run start
```

View File

@@ -1,28 +1,20 @@
import Link from "next/link";
import HomeIcon from "../icons/HomeIcon";
export default function Breadcrumbs({
links,
}: {
links: { title: string; href?: string; target?: string }[];
}) {
export default function Breadcrumbs({ links }: { links: { title: string, href?: string, target?: string }[] }) {
const current = links.at(-1);
return (
<div className="flex items-center uppercase font-black text-xs">
<Link className="flex items-center" href="/">
<HomeIcon />
</Link>
return <div className="flex items-center uppercase font-black text-xs">
<Link className="flex items-center" href='/'><HomeIcon /></Link>
{/* {links.length > 1 && links.slice(0, -1).map((link) => {
{/* {links.length > 1 && links.slice(0, -1).map((link) => {
return <>
<span className="mx-4">/</span>
<Link href={link.href}>{link.title}</Link>
</>
})} */}
<span className="mx-4">/</span>
<span>{current.title}</span>
</div>
);
}
<span className="mx-4">/</span>
<span>{current.title}</span>
</div >
}

View File

@@ -1,13 +1,3 @@
export default function ExternalLinkIcon({ className = "" }) {
return (
<div className={`inline-block w-4 ${className}`}>
<svg
xmlns="http://www.w3.org/2000/svg"
viewBox="0 0 64 64"
fill="currentColor"
>
<path d="M 40 10 C 38.896 10 38 10.896 38 12 C 38 13.104 38.896 14 40 14 L 47.171875 14 L 30.585938 30.585938 C 29.804938 31.366938 29.804938 32.633063 30.585938 33.414062 C 30.976938 33.805063 31.488 34 32 34 C 32.512 34 33.023063 33.805062 33.414062 33.414062 L 50 16.828125 L 50 24 C 50 25.104 50.896 26 52 26 C 53.104 26 54 25.104 54 24 L 54 12 C 54 10.896 53.104 10 52 10 L 40 10 z M 18 12 C 14.691 12 12 14.691 12 18 L 12 46 C 12 49.309 14.691 52 18 52 L 46 52 C 49.309 52 52 49.309 52 46 L 52 34 C 52 32.896 51.104 32 50 32 C 48.896 32 48 32.896 48 34 L 48 46 C 48 47.103 47.103 48 46 48 L 18 48 C 16.897 48 16 47.103 16 46 L 16 18 C 16 16.897 16.897 16 18 16 L 30 16 C 31.104 16 32 15.104 32 14 C 32 12.896 31.104 12 30 12 L 18 12 z" />
</svg>
</div>
);
}
return <div className={`inline-block w-4 ${className}`}><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 64 64" fill="currentColor"><path d="M 40 10 C 38.896 10 38 10.896 38 12 C 38 13.104 38.896 14 40 14 L 47.171875 14 L 30.585938 30.585938 C 29.804938 31.366938 29.804938 32.633063 30.585938 33.414062 C 30.976938 33.805063 31.488 34 32 34 C 32.512 34 33.023063 33.805062 33.414062 33.414062 L 50 16.828125 L 50 24 C 50 25.104 50.896 26 52 26 C 53.104 26 54 25.104 54 24 L 54 12 C 54 10.896 53.104 10 52 10 L 40 10 z M 18 12 C 14.691 12 12 14.691 12 18 L 12 46 C 12 49.309 14.691 52 18 52 L 46 52 C 49.309 52 52 49.309 52 46 L 52 34 C 52 32.896 51.104 32 50 32 C 48.896 32 48 32.896 48 34 L 48 46 C 48 47.103 47.103 48 46 48 L 18 48 C 16.897 48 16 47.103 16 46 L 16 18 C 16 16.897 16.897 16 18 16 L 30 16 C 31.104 16 32 15.104 32 14 C 32 12.896 31.104 12 30 12 L 18 12 z"/></svg></div>
}

View File

@@ -1,10 +1,3 @@
export default function HomeIcon({ className = "" }) {
return (
<div className={`inline-block w-4 ${className}`}>
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24">
{" "}
<path d="M 12 2 A 1 1 0 0 0 11.289062 2.296875 L 1.203125 11.097656 A 0.5 0.5 0 0 0 1 11.5 A 0.5 0.5 0 0 0 1.5 12 L 4 12 L 4 20 C 4 20.552 4.448 21 5 21 L 9 21 C 9.552 21 10 20.552 10 20 L 10 14 L 14 14 L 14 20 C 14 20.552 14.448 21 15 21 L 19 21 C 19.552 21 20 20.552 20 20 L 20 12 L 22.5 12 A 0.5 0.5 0 0 0 23 11.5 A 0.5 0.5 0 0 0 22.796875 11.097656 L 12.716797 2.3027344 A 1 1 0 0 0 12.710938 2.296875 A 1 1 0 0 0 12 2 z" />
</svg>
</div>
);
}
return <div className={`inline-block w-4 ${className}`}><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"> <path d="M 12 2 A 1 1 0 0 0 11.289062 2.296875 L 1.203125 11.097656 A 0.5 0.5 0 0 0 1 11.5 A 0.5 0.5 0 0 0 1.5 12 L 4 12 L 4 20 C 4 20.552 4.448 21 5 21 L 9 21 C 9.552 21 10 20.552 10 20 L 10 14 L 14 14 L 14 20 C 14 20.552 14.448 21 15 21 L 19 21 C 19.552 21 20 20.552 20 20 L 20 12 L 22.5 12 A 0.5 0.5 0 0 0 23 11.5 A 0.5 0.5 0 0 0 22.796875 11.097656 L 12.716797 2.3027344 A 1 1 0 0 0 12.710938 2.296875 A 1 1 0 0 0 12 2 z"/></svg></div>
}

View File

@@ -19,7 +19,10 @@
"owner": "datasets",
"branch": "main",
"repo": "investor-flow-of-funds-us",
"files": ["data/monthly.csv", "data/weekly.csv"],
"files": [
"data/monthly.csv",
"data/weekly.csv"
],
"readme": "README.md"
},
{
@@ -35,10 +38,7 @@
"owner": "fivethirtyeight",
"repo": "data",
"branch": "master",
"files": [
"nba-raptor/historical_RAPTOR_by_player.csv",
"nba-raptor/historical_RAPTOR_by_team.csv"
],
"files": ["nba-raptor/historical_RAPTOR_by_player.csv", "nba-raptor/historical_RAPTOR_by_team.csv"],
"readme": "nba-raptor/README.md"
}
]

View File

@@ -1,5 +1,5 @@
/* eslint-disable @typescript-eslint/no-explicit-any */
declare module "*.svg" {
declare module '*.svg' {
const content: any;
export const ReactComponent: any;
export default content;

View File

@@ -1,4 +1,4 @@
import { Octokit } from "octokit";
import { Octokit } from 'octokit';
export interface GithubProject {
owner: string;
@@ -26,16 +26,15 @@ export async function getProjectReadme(
ref: branch,
});
const data = response.data as { content?: string };
const fileContent = data.content ? data.content : "";
if (fileContent === "") {
const fileContent = data.content ? data.content : '';
if (fileContent === '') {
return null;
}
const decodedContent = Buffer.from(fileContent, "base64").toString();
const decodedContent = Buffer.from(fileContent, 'base64').toString();
return decodedContent;
} catch (error) {
throw new Error(
"Couldn't get project readme please make sure that you are pointing to a valid repo and that the repo in question contains a README.md"
);
console.log(error);
return null;
}
}
@@ -51,13 +50,13 @@ export async function getLastUpdated(
const response = await octokit.rest.repos.listCommits({
owner,
repo,
path: readme,
ref: branch,
});
return response.data[0].commit.committer.date;
} catch (error) {
throw new Error(
"Couldn't get project list of commits please make sure that you are pointing to a valid repo"
);
console.log(error);
return null;
}
}
export async function getProjectMetadata(
@@ -73,9 +72,8 @@ export async function getProjectMetadata(
});
return response.data;
} catch (error) {
throw new Error(
"Couldn't get project metadata please make sure that you are pointing to a valid repo"
);
console.log(error);
return null;
}
}
@@ -96,32 +94,13 @@ export async function getRepoContents(
ref: branch,
path: path,
});
const data = response.data as {
download_url?: string;
name: string;
size: number;
};
contents.push({
download_url: data.download_url,
name: data.name,
size: data.size,
});
const data = response.data as { download_url?: string, name: string, size: number };
contents.push({ download_url: data.download_url, name: data.name, size: data.size});
}
return contents;
} catch (error) {
if (
error.message ===
'This endpoint can only return blobs smaller than 100 MB in size. The requested blob is too large to fetch via the API, but you can always clone the repository via Git to obtain it.: {"resource":"Blob","field":"data","code":"too_large"}'
) {
throw new Error(
`The requested files ${files.join(
", "
)} are too big making it impossible to fetch via Github API`
);
}
throw new Error(
"Couldn't get project contents please make sure that you are pointing to a valid repo"
);
console.log(error);
return null;
}
}
@@ -141,20 +120,22 @@ export async function getProject(project: GithubProject, github_pat?: string) {
project.readme,
github_pat
);
let projectData = [];
if (project.files) {
projectData = await getRepoContents(
project.owner,
project.repo,
project.branch,
project.files,
github_pat
);
if (!projectReadme) {
return null;
}
const projectBase =
project.readme && project.readme.split("/").length > 1
? project.readme.split("/").slice(0, -1).join("/")
: "/";
const projectData = await getRepoContents(
project.owner,
project.repo,
project.branch,
project.files,
github_pat
);
if (!projectData) {
return null;
}
const projectBase = project.readme.split('/').length > 1
? project.readme.split('/').slice(0, -1).join('/')
: '/'
const last_updated = await getLastUpdated(
project.owner,
project.repo,
@@ -162,11 +143,5 @@ export async function getProject(project: GithubProject, github_pat?: string) {
projectBase,
github_pat
);
return {
...projectMetadata,
files: projectData,
readmeContent: projectReadme,
last_updated,
base_path: projectBase,
};
return { ...projectMetadata, files: projectData, readmeContent: projectReadme, last_updated, base_path: projectBase };
}

View File

@@ -3,8 +3,8 @@ const nextConfig = {
return {
beforeFiles: [
{
source: "/@:org/:project*",
destination: "/@org/:org/:project*",
source: '/@:org/:project*',
destination: '/@org/:org/:project*',
},
],
};

View File

@@ -27,7 +27,6 @@
"@tailwindcss/typography": "^0.5.9",
"autoprefixer": "^10.4.14",
"postcss": "^8.4.23",
"prettier": "2.8.8",
"tailwindcss": "^3.3.1"
}
},
@@ -4678,21 +4677,6 @@
"node": ">= 0.8.0"
}
},
"node_modules/prettier": {
"version": "2.8.8",
"resolved": "https://registry.npmjs.org/prettier/-/prettier-2.8.8.tgz",
"integrity": "sha512-tdN8qQGvNjw4CHbY+XXk0JgCXn9QiF21a55rBe5LJAU+kDyC4WQn4+awm2Xfk2lQMk5fKup9XgzTZtGkjBdP9Q==",
"dev": true,
"bin": {
"prettier": "bin-prettier.js"
},
"engines": {
"node": ">=10.13.0"
},
"funding": {
"url": "https://github.com/prettier/prettier?sponsor=1"
}
},
"node_modules/prop-types": {
"version": "15.8.1",
"resolved": "https://registry.npmjs.org/prop-types/-/prop-types-15.8.1.tgz",

View File

@@ -6,8 +6,7 @@
"dev": "next dev",
"build": "next build",
"start": "next start",
"lint": "next lint",
"prettier": "prettier --write ."
"lint": "next lint"
},
"dependencies": {
"@types/node": "18.16.0",
@@ -29,7 +28,6 @@
"@tailwindcss/typography": "^0.5.9",
"autoprefixer": "^10.4.14",
"postcss": "^8.4.23",
"prettier": "2.8.8",
"tailwindcss": "^3.3.1"
}
}

View File

@@ -1,31 +1,22 @@
import { NextSeo } from "next-seo";
import { promises as fs } from "fs";
import path from "path";
import getConfig from "next/config";
import { getProject, GithubProject } from "../../../lib/octokit";
import ReactMarkdown from "react-markdown";
import remarkGfm from "remark-gfm";
import Breadcrumbs from "../../../components/_shared/Breadcrumbs";
import { NextSeo } from 'next-seo';
import { promises as fs } from 'fs';
import path from 'path';
import getConfig from 'next/config';
import { getProject, GithubProject } from '../../../lib/octokit';
import ReactMarkdown from 'react-markdown';
import remarkGfm from 'remark-gfm';
import Breadcrumbs from '../../../components/_shared/Breadcrumbs';
export default function ProjectPage({ project }) {
const repoId = `@${project.repo_config.owner}/${project.repo_config.repo}`;
const repoId = `@${project.repo_config.owner}/${project.repo_config.repo}`
return (
<>
<NextSeo
title={`${repoId}${
project.base_path !== "/" ? "/" + project.base_path : ""
} - GitHub Datasets`}
/>
<NextSeo title={`${repoId}${project.base_path !== '/' ? '/' + project.base_path : ''} - GitHub Datasets`} />
<main className="prose mx-auto my-8">
<Breadcrumbs links={[{ title: repoId, href: "" }]} />
<h1 className="mb-0 mt-16">{project.repo_config.name || repoId}</h1>
<p className="mb-8">
<span className="font-semibold">Repository:</span>{" "}
<a target="_blank" href={project.html_url}>
{project.html_url}
</a>
</p>
<p className='mb-8'><span className='font-semibold'>Repository:</span> <a target="_blank" href={project.html_url}>{project.html_url}</a></p>
<h2 className="mb-0 mt-10">Files</h2>
<div className="inline-block min-w-full py-2 align-middle">
@@ -63,7 +54,7 @@ export default function ProjectPage({ project }) {
<hr />
<h2 className="uppercase font-black">Readme</h2>
<h2 className='uppercase font-black'>Readme</h2>
<ReactMarkdown remarkPlugins={[remarkGfm]}>
{project.readmeContent}
</ReactMarkdown>
@@ -74,14 +65,17 @@ export default function ProjectPage({ project }) {
// Generates `/posts/1` and `/posts/2`
export async function getStaticPaths() {
const jsonDirectory = path.join(process.cwd(), "datasets.json");
const repos = await fs.readFile(jsonDirectory, "utf8");
const jsonDirectory = path.join(
process.cwd(),
'datasets.json'
);
const repos = await fs.readFile(jsonDirectory, 'utf8');
return {
paths: JSON.parse(repos).map((repo) => {
const projectPath =
repo.readme && repo.readme.split("/").length > 1
? repo.readme.split("/").slice(0, -1)
repo.readme.split('/').length > 1
? repo.readme.split('/').slice(0, -1)
: null;
let path = [repo.repo];
if (projectPath) {
@@ -98,13 +92,16 @@ export async function getStaticPaths() {
}
export async function getStaticProps({ params }) {
const jsonDirectory = path.join(process.cwd(), "datasets.json");
const reposFile = await fs.readFile(jsonDirectory, "utf8");
const jsonDirectory = path.join(
process.cwd(),
'datasets.json'
);
const reposFile = await fs.readFile(jsonDirectory, 'utf8');
const repos: GithubProject[] = JSON.parse(reposFile);
const repo = repos.find((_repo) => {
const projectPath =
_repo.readme && _repo.readme.split("/").length > 1
? _repo.readme.split("/").slice(0, -1)
_repo.readme.split('/').length > 1
? _repo.readme.split('/').slice(0, -1)
: null;
let path = [_repo.repo];
if (projectPath) {

View File

@@ -1,6 +1,6 @@
import { AppProps } from "next/app";
import Head from "next/head";
import "./styles.css";
import { AppProps } from 'next/app';
import Head from 'next/head';
import './styles.css';
function CustomApp({ Component, pageProps }: AppProps) {
return (

View File

@@ -1,19 +1,21 @@
import { promises as fs } from "fs";
import path from "path";
import { getProject } from "../lib/octokit";
import getConfig from "next/config";
import ExternalLinkIcon from "../components/icons/ExternalLinkIcon";
import TimeAgo from "react-timeago";
import Link from "next/link";
import { NextSeo } from "next-seo";
import { promises as fs } from 'fs';
import path from 'path';
import { getProject } from '../lib/octokit';
import getConfig from 'next/config';
import ExternalLinkIcon from '../components/icons/ExternalLinkIcon';
import TimeAgo from 'react-timeago';
import Link from 'next/link';
export async function getStaticProps() {
const jsonDirectory = path.join(process.cwd(), "/datasets.json");
const repos = await fs.readFile(jsonDirectory, "utf8");
const jsonDirectory = path.join(
process.cwd(),
'/datasets.json'
);
const repos = await fs.readFile(jsonDirectory, 'utf8');
const github_pat = getConfig().serverRuntimeConfig.github_pat;
const projects = await Promise.all(
JSON.parse(repos).map(async (repo) => {
(JSON.parse(repos)).map(async (repo) => {
const project = await getProject(repo, github_pat);
return { ...project, repo_config: repo };
})
@@ -27,112 +29,88 @@ export async function getStaticProps() {
export function Datasets({ projects }) {
return (
<>
<NextSeo title="GitHub Datasets" />
<div className="bg-white min-h-screen">
<div className="mx-auto max-w-7xl px-6 py-16 sm:py-24 lg:px-8">
<div className="text-center">
<h2 className="text-3xl font-bold leading-10 tracking-tight">
GitHub Datasets
</h2>
<p className="mt-3 mx-auto max-w-2xl text-base leading-7 text-gray-500">
Data catalog with datasets hosted on GitHub by{" "}
<Link
target="_blank"
className="underline"
href="https://portaljs.org/"
>
🌀 PortalJS
</Link>
</p>
</div>
<div className="mt-20">
<div className="-mx-4 -my-2 overflow-x-auto sm:-mx-6 lg:-mx-8">
<div className="inline-block min-w-full py-2 align-middle sm:px-6 lg:px-8">
<table className="min-w-full divide-y divide-gray-300">
<thead>
<tr>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Name
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Repository
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Description
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Last updated
</th>
<th
scope="col"
className="relative py-3.5 pl-3 pr-4 sm:pr-0"
></th>
<div className="bg-white min-h-screen">
<div className="mx-auto max-w-7xl px-6 py-16 sm:py-24 lg:px-8">
<div className='text-center'>
<h2 className="text-3xl font-bold leading-10 tracking-tight">
GitHub Datasets
</h2>
<p className="mt-3 mx-auto max-w-2xl text-base leading-7 text-gray-500">
Data catalog with datasets hosted on GitHub by <Link target="_blank" className='underline' href="https://portaljs.org/">🌀 PortalJS</Link>
</p>
</div>
<div className="mt-20">
<div className="-mx-4 -my-2 overflow-x-auto sm:-mx-6 lg:-mx-8">
<div className="inline-block min-w-full py-2 align-middle sm:px-6 lg:px-8">
<table className="min-w-full divide-y divide-gray-300">
<thead>
<tr>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Name
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Repository
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Description
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Last updated
</th>
<th
scope="col"
className="relative py-3.5 pl-3 pr-4 sm:pr-0"
></th>
</tr>
</thead>
<tbody className="divide-y divide-gray-200">
{projects.map((project) => (
<tr key={project.id}>
<td className="whitespace-nowrap px-3 py-6 text-sm text-gray-500">
{project.repo_config.name
? project.repo_config.name
: project.full_name + (project.base_path === '/' ? '' : '/' + project.base_path)}
</td>
<td className="whitespace-nowrap px-3 py-6 text-sm group text-gray-500 hover:text-gray-900 transition-all duration-250">
<a href={project.html_url} target="_blank" className='flex items-center'>@{project.full_name} <ExternalLinkIcon className='ml-1' /></a>
</td>
<td className="px-3 py-4 text-sm text-gray-500">
{project.repo_config.description
? project.repo_config.description
: project.description}
</td>
<td className="whitespace-nowrap px-3 py-6 text-sm text-gray-500">
<TimeAgo date={new Date(project.last_updated)} />
</td>
<td className="relative whitespace-nowrap py-6 pl-3 pr-4 text-right text-sm font-medium sm:pr-0">
<a
href={`/@${project.repo_config.owner}/${project.repo_config.repo}/${project.base_path === '/' ? '' : project.base_path}`}
className='border border-gray-900 text-gray-900 px-4 py-2 transition-all hover:bg-gray-900 hover:text-white'
>
info
</a>
</td>
</tr>
</thead>
<tbody className="divide-y divide-gray-200">
{projects.map((project) => (
<tr key={project.id}>
<td className="whitespace-nowrap px-3 py-6 text-sm text-gray-500">
{project.repo_config.name
? project.repo_config.name
: project.full_name +
(project.base_path === "/"
? ""
: "/" + project.base_path)}
</td>
<td className="whitespace-nowrap px-3 py-6 text-sm group text-gray-500 hover:text-gray-900 transition-all duration-250">
<a
href={project.html_url}
target="_blank"
className="flex items-center"
>
@{project.full_name}{" "}
<ExternalLinkIcon className="ml-1" />
</a>
</td>
<td className="px-3 py-4 text-sm text-gray-500">
{project.repo_config.description
? project.repo_config.description
: project.description}
</td>
<td className="whitespace-nowrap px-3 py-6 text-sm text-gray-500">
<TimeAgo date={new Date(project.last_updated)} />
</td>
<td className="relative whitespace-nowrap py-6 pl-3 pr-4 text-right text-sm font-medium sm:pr-0">
<a
href={`/@${project.repo_config.owner}/${
project.repo_config.repo
}/${
project.base_path === "/" ? "" : project.base_path
}`}
className="border border-gray-900 text-gray-900 px-4 py-2 transition-all hover:bg-gray-900 hover:text-white"
>
info
</a>
</td>
</tr>
))}
</tbody>
</table>
</div>
))}
</tbody>
</table>
</div>
</div>
</div>
</div>
</>
</div>
);
}

View File

@@ -3,4 +3,4 @@ module.exports = {
tailwindcss: {},
autoprefixer: {},
},
};
}

View File

@@ -8,5 +8,8 @@ module.exports = {
theme: {
extend: {},
},
plugins: [require("@tailwindcss/typography")],
};
plugins: [
require('@tailwindcss/typography')
],
}

View File

@@ -1,138 +1,45 @@
import { expect, test } from 'vitest';
import { getAllProjectsFromOrg, getProjectDataPackage } from '../lib/project';
import { loadDataPackage } from '../lib/loader';
import { getProjectMetadata } from '../lib/project';
import { getCsv, parseCsv } from '../components/Table';
import { Octokit } from 'octokit';
import { assert, expect, test } from 'vitest'
import { getProjectDataPackage } from '../lib/octokit';
test(
'Test OS-Data',
async () => {
const repos = await getAllProjectsFromOrg(
'os-data',
'main',
process.env.VITE_GITHUB_PAT
);
if (repos.failed.length > 0) console.log(repos.failed);
expect(repos.failed.length).toBe(0);
},
{ timeout: 100000 }
);
export async function getAllDataPackagesFromOrg(
org: string,
branch?: string,
github_pat?: string
) {
const octokit = new Octokit({ auth: github_pat });
const repos = await octokit.rest.repos.listForOrg({ org, type: 'public', per_page: 100 });
let failedDataPackages = [];
const datapackages = await Promise.all(
repos.data.map(async (_repo) => {
const datapackage = await getProjectDataPackage(
org,
_repo.name,
branch ? branch : 'main',
github_pat
);
if (!datapackage) {
failedDataPackages.push(_repo.name)
return null
};
return {...datapackage, repo: _repo.name};
})
);
return {
datapackages: datapackages.filter((item) => item !== null),
failedDataPackages,
};
}
test(
'Test Gift-Data',
async () => {
const repos = await getAllProjectsFromOrg(
'gift-data',
'main',
process.env.VITE_GITHUB_PAT
);
if (repos.failed.length > 0) console.log(repos.failed);
expect(repos.failed.length).toBe(0);
},
{ timeout: 100000 }
);
test('Test OS-Data', async () => {
const repos = await getAllDataPackagesFromOrg('os-data', 'main', process.env.VITE_GITHUB_PAT)
if (repos.failedDataPackages.length > 0) console.log(repos.failedDataPackages)
expect(repos.failedDataPackages.length).toBe(0)
}, {timeout: 100000})
test(
'Test getting one dataset from github',
async () => {
const datapackage = await getProjectDataPackage(
'os-data',
'berlin-berlin',
'main',
process.env.VITE_GITHUB_PAT
);
const repo = await getProjectMetadata(
'os-data',
'berlin-berlin',
process.env.VITE_GITHUB_PAT
);
const project = loadDataPackage(datapackage, repo);
delete project['datapackage'];
delete project.files[0]['dialect'];
delete project.files[0]['schema'];
expect(project).toStrictEqual({
name: 'berlin-berlin',
title: 'Berlin-Berlin',
description: null,
owner: {
name: 'os-data',
logo: 'https://avatars.githubusercontent.com/u/13695166?v=4',
title: 'os-data',
},
repo: {
name: 'berlin-berlin',
full_name: 'os-data/berlin-berlin',
url: 'https://github.com/os-data/berlin-berlin',
},
files: [
{
name: 'berlin-gesamt',
format: 'csv',
path: 'https://storage.openspending.org/berlin-berlin/berlin-gesamt.csv',
mediatype: 'text/csv',
bytes: 81128743,
encoding: 'utf-8',
},
],
author: 'Michael Peters <michael.peters@okfn.de>',
cityCode: 'Berlin',
countryCode: 'DE',
fiscalPeriod: { start: '2014-01-01', end: '2019-12-31' },
readme: '',
});
},
{ timeout: 100000 }
);
test('Test Gift-Data', async () => {
const repos = await getAllDataPackagesFromOrg('gift-data', 'main', process.env.VITE_GITHUB_PAT)
if (repos.failedDataPackages.length > 0) console.log(repos.failedDataPackages)
expect(repos.failedDataPackages.length).toBe(0)
}, {timeout: 100000})
test(
'Test getting one section of csv from R2',
async () => {
const rawCsv = await getCsv(
'https://storage.openspending.org/state-of-minas-gerais-brazil-planned-budget/__os_imported__br-mg-ppagloc.csv'
);
const parsedCsv = await parseCsv(rawCsv);
expect(parsedCsv.errors.length).toBe(1);
expect(parsedCsv.data.length).toBe(10165);
expect(parsedCsv.meta.fields).toStrictEqual([
'function_name',
'function_label',
'product_name',
'product_label',
'area_name',
'area_label',
'subaction_name',
'subaction_label',
'region_label_map',
'region_reg_map',
'region_name',
'region_label',
'municipality_map_id',
'municipality_name',
'municipality_map_code',
'municipality_label',
'municipality_map_name_simple',
'municipality_map_name',
'cofog1_label_en',
'cofog1_name',
'cofog1_label',
'amount',
'subprogramme_name',
'subprogramme_label',
'time_name',
'time_year',
'time_month',
'time_day',
'time_week',
'time_yearmonth',
'time_quarter',
'time',
'action_name',
'action_label',
'subfunction_name',
'subfunction_label',
'programme_name',
'programme_label',
]);
},
{ timeout: 100000 }
);

View File

@@ -9,7 +9,7 @@ export default function DatasetCard({ dataset }: { dataset: Project }) {
className="overflow-hidden rounded-xl border border-gray-200"
>
<Link
href={`/@${dataset.owner.name}/${dataset.repo.name}`}
href=""
className="flex items-center gap-x-4 border-b border-gray-900/5 bg-gray-50 p-6"
>
<img
@@ -60,8 +60,8 @@ export default function DatasetCard({ dataset }: { dataset: Project }) {
<dd className="flex items-start gap-x-2">
<div className="font-medium text-gray-900">
<Link
// TODO: this link may be incorrect for some datasets
href={`https://github.com/${dataset.owner.name}/${dataset.repo.name}/blob/main/datapackage.json`}
// TODO: where do we get the info needed for this link?
href=""
target="_blank"
className="flex items-center hover:text-gray-700"
>

View File

@@ -2,26 +2,9 @@ import { useForm } from 'react-hook-form';
import DatasetsGrid from './DatasetsGrid';
import { Project } from '../lib/project.interface';
import { Index } from 'flexsearch';
import {
ChevronDoubleLeftIcon,
ChevronDoubleRightIcon,
ChevronLeftIcon,
ChevronRightIcon,
} from '@heroicons/react/24/solid';
import { useState } from 'react';
export default function DatasetsSearch({
datasets,
availableCountries,
}: {
datasets: Project[];
availableCountries;
}) {
const itemsPerPage = 6;
const [page, setPage] = useState(1);
export default function DatasetsSearch({ datasets }: { datasets: Project[] }) {
const index = new Index({ tokenize: 'full' });
datasets.forEach((dataset: Project) =>
index.add(
dataset.name,
@@ -38,38 +21,12 @@ export default function DatasetsSearch({
},
});
const filteredDatasets = datasets
.filter((dataset: Project) =>
watch().searchTerm && watch().searchTerm !== ''
? index.search(watch().searchTerm).includes(dataset.name)
: true
)
.filter((dataset) =>
watch().country && watch().country !== ''
? dataset.countryCode === watch().country
: true
)
// TODO: Does that really makes sense?
// What if the fiscalPeriod is 2015-2017 and inputs are
// set to 2015-2016. It's going to be filtered out but
// it shouldn't.
.filter((dataset) =>
watch().minDate && watch().minDate !== ''
? dataset.fiscalPeriod?.start >= watch().minDate
: true
)
.filter((dataset) =>
watch().maxDate && watch().maxDate !== ''
? dataset.fiscalPeriod?.end <= watch().maxDate
: true
);
const paginatedDatasets = filteredDatasets.slice(
(page - 1) * itemsPerPage,
(page - 1) * itemsPerPage + itemsPerPage
);
const pageCount = Math.ceil(filteredDatasets.length / itemsPerPage) || 1;
const allCountries = datasets
.map((item) => item.countryCode)
.filter((v) => v) // Filters false values
.filter((v, i, a) => a.indexOf(v) === i) // Remove duplicates
// TODO: title should be the full name
.map((code) => ({ code, title: code }));
return (
<>
@@ -80,7 +37,7 @@ export default function DatasetsSearch({
<input
placeholder="Search datasets"
aria-label="Search datasets"
{...register('searchTerm', { onChange: () => setPage(1) })}
{...register('searchTerm')}
className="h-[3em] relative w-full rounded-lg bg-white py-2 pl-3 pr-10 text-left shadow-md focus:outline-none focus-visible:border-emerald-500 focus-visible:ring-2 focus-visible:ring-white focus-visible:ring-opacity-75 focus-visible:ring-offset-2 focus-visible:ring-offset-emerald-400 sm:text-sm"
/>
{watch().searchTerm !== '' && (
@@ -98,10 +55,10 @@ export default function DatasetsSearch({
<label className="text-sm text-gray-600 font-medium">Country</label>
<select
className="h-[3em] w-full rounded-lg bg-white py-2 pl-3 pr-10 text-left shadow-md focus:outline-none focus-visible:border-emerald-500 focus-visible:ring-2 focus-visible:ring-white focus-visible:ring-opacity-75 focus-visible:ring-offset-2 focus-visible:ring-offset-emerald-400 sm:text-sm"
{...register('country', { onChange: () => setPage(1) })}
{...register('country')}
>
<option value="">All</option>
{availableCountries.map((country) => {
{allCountries.map((country) => {
return (
<option key={country.code} value={country.code}>
{country.title}
@@ -116,9 +73,17 @@ export default function DatasetsSearch({
<input
aria-label="Min. date"
type="date"
{...register('minDate', { onChange: () => setPage(1) })}
{...register('minDate')}
className="h-[3em] w-full rounded-lg bg-white py-2 pl-3 pr-10 text-left shadow-md focus:outline-none focus-visible:border-emerald-500 focus-visible:ring-2 focus-visible:ring-white focus-visible:ring-opacity-75 focus-visible:ring-offset-2 focus-visible:ring-offset-emerald-400 sm:text-sm"
/>
{watch().minDate !== '' && (
<button
onClick={() => resetField('minDate')}
className="absolute right-3 top-1/2 -translate-y-1/2 text-gray-500"
>
<CloseIcon />
</button>
)}
</div>
</div>
<div className="sm:basis-1/6">
@@ -127,56 +92,48 @@ export default function DatasetsSearch({
<input
aria-label="Max. date"
type="date"
{...register('maxDate', { onChange: () => setPage(1) })}
{...register('maxDate')}
className="h-[3em] w-full rounded-lg bg-white py-2 pl-3 pr-10 text-left shadow-md focus:outline-none focus-visible:border-emerald-500 focus-visible:ring-2 focus-visible:ring-white focus-visible:ring-opacity-75 focus-visible:ring-offset-2 focus-visible:ring-offset-emerald-400 sm:text-sm"
/>
{watch().maxDate !== '' && (
<button
onClick={() => resetField('maxDate')}
className="absolute right-3 top-1/2 -translate-y-1/2 text-gray-500"
>
<CloseIcon />
</button>
)}
</div>
</div>
</div>
<div className="mt-5 mb-5">
<span className="text-lg font-medium">
{filteredDatasets.length} datasets found
</span>
</div>
<div className="min-w-full align-middle">
<DatasetsGrid datasets={paginatedDatasets} />
<div className="w-full flex justify-center mt-10">
<button
onClick={() => setPage(1)}
disabled={page <= 1}
className="disabled:text-gray-400"
>
<ChevronDoubleLeftIcon className="w-6 h-6" />
</button>
<button
onClick={() => {
if (page > 1) setPage((prev) => --prev);
}}
disabled={page <= 1}
className="disabled:text-gray-400"
>
<ChevronLeftIcon className="w-6 h-6" />
</button>
<span className="mx-5">
Page {page} of {pageCount}
</span>
<button
onClick={() => {
if (page < pageCount) setPage((prev) => ++prev);
}}
disabled={page >= pageCount}
className="disabled:text-gray-400"
>
<ChevronRightIcon className="w-6 h-6" />
</button>
<button
onClick={() => setPage(pageCount)}
disabled={page >= pageCount}
className="disabled:text-gray-400"
>
<ChevronDoubleRightIcon className="w-6 h-6" />
</button>
</div>
<div className="min-w-full mt-10 align-middle">
<DatasetsGrid
datasets={datasets
.filter((dataset: Project) =>
watch().searchTerm && watch().searchTerm !== ''
? index.search(watch().searchTerm).includes(dataset.name)
: true
)
.filter((dataset) =>
watch().country && watch().country !== ''
? dataset.countryCode === watch().country
: true
)
// TODO: Does that really makes sense?
// What if the fiscalPeriod is 2015-2017 and inputs are
// set to 2015-2016. It's going to be filtered out but
// it shouldn't.
.filter((dataset) =>
watch().minDate && watch().minDate !== ''
? dataset.fiscalPeriod?.start >= watch().minDate
: true
)
.filter((dataset) =>
watch().maxDate && watch().maxDate !== ''
? dataset.fiscalPeriod?.end <= watch().maxDate
: true
)}
/>
</div>
</>
);

View File

@@ -1,82 +1,53 @@
import Image from 'next/image';
import { Container } from './Container';
import logo from '../public/logo.svg';
import Link from 'next/link';
import { useRouter } from 'next/router';
import { Bars3Icon } from '@heroicons/react/24/outline';
import { useState } from 'react';
import Image from 'next/image'
import { Button } from './Button'
import { Container } from './Container'
import logo from "../public/logo.svg"
import Link from 'next/link'
import { useRouter } from 'next/router'
export function Header() {
const [menuOpen, setMenuOpen] = useState<boolean>(false);
const router = useRouter();
const isActive = (navLink) => {
return router.asPath.split('?')[0] == navLink.href;
};
return router.asPath.split("?")[0] == navLink.href;
}
const navLinks = [
{
title: 'Home',
href: '/',
title: "Home",
href: "/#header"
},
{
title: 'Datasets',
href: '/#datasets',
title: "Datasets",
href: "/#datasets"
},
// {
// title: "Community",
// href: "https://community.openspending.org/"
// }
];
{
title: "Community",
href: "https://community.openspending.org/"
}
]
return (
<header className="relative z-50 pb-11 lg:pt-11">
<Container className="flex flex-wrap items-center justify-between lg:flex-nowrap mt-10 lg:mt-0">
<Link href="/" className="lg:mt-0 lg:grow lg:basis-0 flex items-center">
<header className="z-50 pb-5 lg:pt-11 sticky top-0 backdrop-blur" id="header">
<Container className="flex flex-wrap items-center justify-center sm:justify-between lg:flex-nowrap">
<div className="mt-10 lg:mt-0 lg:grow lg:basis-0 flex items-center">
<Image src={logo} alt="OpenSpending" className="h-12 w-auto" />
</Link>
<ul className="hidden list-none sm:flex gap-x-5 text-base font-medium">
</div>
<ul className='list-none flex gap-x-5 text-base font-medium'>
{navLinks.map((link, i) => (
<li key={`nav-link-${i}`}>
<Link
className={`text-emerald-900 hover:text-emerald-600 ${
isActive(link) ? 'text-emerald-600' : ''
}`}
className={`text-emerald-900 hover:text-emerald-600 ${isActive(link) ? "text-emerald-600" : ""}`}
href={link.href}
scroll={false}
>
{link.title}
</Link>
</li>
))}
</li>))}
</ul>
<div className="hidden xl:block xl:grow"></div>
<div className="sm:hidden sm:mt-10 lg:mt-0 lg:grow lg:basis-0 lg:justify-end">
<button onClick={() => setMenuOpen(!menuOpen)}>
<Bars3Icon className="w-8 h-8" />
</button>
<div className="hidden sm:mt-10 sm:flex lg:mt-0 lg:grow lg:basis-0 lg:justify-end">
</div>
{menuOpen && (
<div className={`sm:hidden basis-full mt-5 text-center`}>
<ul className="gap-x-5 text-base font-medium">
{navLinks.map((link, i) => (
<li key={`nav-link-${i}`}>
<Link
className={`text-emerald-900 hover:text-emerald-600 ${
isActive(link) ? 'text-emerald-600' : ''
}`}
href={link.href}
scroll={false}
>
{link.title}
</Link>
</li>
))}
</ul>
</div>
)}
</Container>
</header>
);
</header >
)
}

View File

@@ -1,9 +1,9 @@
import { Button } from './Button';
import { Container } from './Container';
import { Button } from './Button'
import { Container } from './Container'
export function Hero({ countriesCount, datasetsCount, filesCount }) {
export function Hero() {
return (
<div className="relative pb-20 pt-10 sm:py-40" id="hero">
<div className="relative pb-20 pt-10 sm:py-40">
<div className="absolute inset-x-0 -bottom-14 -top-48 overflow-hidden bg-green-50 bg-opacity-50">
<div className="absolute inset-x-0 top-0 h-40 bg-gradient-to-b from-white" />
<div className="absolute inset-x-0 bottom-0 h-40 bg-gradient-to-t from-white" />
@@ -15,13 +15,12 @@ export function Hero({ countriesCount, datasetsCount, filesCount }) {
</h1>
<div className="mt-6 space-y-6 font-display text-2xl tracking-tight text-emerald-900">
<p>
By understanding how governments spend money in our name can we
have a say in how that money will affect our own lives. The
journey starts here.
By understanding how governments spend money in our name can we have a say
in how that money will affect our own lives. The journey starts here.
</p>
<p>
OpenSpending is a free, open and global platform to search,
visualise and analyse fiscal data in the public sphere.
OpenSpending is a free, open and global platform to search, visualise and analyse
fiscal data in the public sphere.
</p>
</div>
<Button href="#datasets" className="mt-10">
@@ -29,11 +28,9 @@ export function Hero({ countriesCount, datasetsCount, filesCount }) {
</Button>
<dl className="mt-10 grid grid-cols-2 gap-x-10 gap-y-6 sm:mt-16 sm:gap-x-16 sm:gap-y-10 sm:text-center lg:auto-cols-auto lg:grid-flow-col lg:grid-cols-none lg:justify-start lg:text-left">
{[
// Added the plus sign because some datasets do not
// contain defined countries
['Countries', '+' + countriesCount],
['Datasets', datasetsCount],
['Files', filesCount],
['Countries', '75'],
['Datasets', '2091'],
['Files', '9230'],
].map(([name, value]) => (
<div key={name}>
<dt className="font-mono text-sm text-emerald-600">{name}</dt>
@@ -46,5 +43,5 @@ export function Hero({ countriesCount, datasetsCount, filesCount }) {
</div>
</Container>
</div>
);
)
}

View File

@@ -1,94 +0,0 @@
import {
QueryClient,
QueryClientProvider,
useQuery,
useQueryClient,
} from 'react-query';
import Papa from 'papaparse';
import { Grid } from '@githubocto/flat-ui';
const queryClient = new QueryClient();
export async function getCsv(url: string) {
const response = await fetch(url, {
headers: {
Range: 'bytes=0-5132288',
},
});
const data = await response.text();
return data;
}
export async function parseCsv(file: string): Promise<any> {
return new Promise((resolve, reject) => {
Papa.parse(file, {
header: true,
dynamicTyping: true,
skipEmptyLines: true,
transform: (value: string): string => {
return value.trim();
},
complete: (results: any) => {
return resolve(results);
},
error: (error: any) => {
return reject(error);
},
});
});
}
const Spinning = () => {
return (
<div role="status w-fit mx-auto">
<svg
aria-hidden="true"
className="w-8 h-8 mr-2 text-gray-200 animate-spin dark:text-gray-600 fill-emerald-600"
viewBox="0 0 100 101"
fill="none"
xmlns="http://www.w3.org/2000/svg"
>
<path
d="M100 50.5908C100 78.2051 77.6142 100.591 50 100.591C22.3858 100.591 0 78.2051 0 50.5908C0 22.9766 22.3858 0.59082 50 0.59082C77.6142 0.59082 100 22.9766 100 50.5908ZM9.08144 50.5908C9.08144 73.1895 27.4013 91.5094 50 91.5094C72.5987 91.5094 90.9186 73.1895 90.9186 50.5908C90.9186 27.9921 72.5987 9.67226 50 9.67226C27.4013 9.67226 9.08144 27.9921 9.08144 50.5908Z"
fill="currentColor"
/>
<path
d="M93.9676 39.0409C96.393 38.4038 97.8624 35.9116 97.0079 33.5539C95.2932 28.8227 92.871 24.3692 89.8167 20.348C85.8452 15.1192 80.8826 10.7238 75.2124 7.41289C69.5422 4.10194 63.2754 1.94025 56.7698 1.05124C51.7666 0.367541 46.6976 0.446843 41.7345 1.27873C39.2613 1.69328 37.813 4.19778 38.4501 6.62326C39.0873 9.04874 41.5694 10.4717 44.0505 10.1071C47.8511 9.54855 51.7191 9.52689 55.5402 10.0491C60.8642 10.7766 65.9928 12.5457 70.6331 15.2552C75.2735 17.9648 79.3347 21.5619 82.5849 25.841C84.9175 28.9121 86.7997 32.2913 88.1811 35.8758C89.083 38.2158 91.5421 39.6781 93.9676 39.0409Z"
fill="currentFill"
/>
</svg>
<span className="sr-only">Loading...</span>
</div>
);
};
export const Table: React.FC<{ url: string }> = ({ url }) => {
return (
// Provide the client to your App
<QueryClientProvider client={queryClient}>
<TableInner url={url} />
</QueryClientProvider>
);
};
const TableInner: React.FC<{ url: string }> = ({ url }) => {
const { data: csvString, isLoading: isDownloadingCSV } = useQuery(
['dataCsv', url],
() => getCsv(url)
);
const { data: parsedData, isLoading: isParsing } = useQuery(
['dataPreview', csvString],
() => parseCsv(csvString),
{ enabled: !!csvString }
);
if (isParsing || isDownloadingCSV)
<div className="w-full">
<Spinning />
</div>;
if (parsedData)
return (
<div className="h-[500px] overflow-scroll">
<Grid data={parsedData.data} />
</div>
);
};

View File

@@ -1,10 +0,0 @@
import { Header } from '../Header';
export default function Layout({ children }) {
return (
<div className="bg-white min-h-screen">
<Header />
{children}
</div>
);
}

View File

@@ -0,0 +1,27 @@
[
{
"owner": "os-data",
"branch": "main",
"name": "mongolia-budget-2016-2017"
},
{
"owner": "os-data",
"branch": "main",
"name": "gb-country-regional-analysis"
},
{
"owner": "os-data",
"branch": "main",
"name": "berlin-berlin"
},
{
"owner": "os-data",
"branch": "main",
"name": "state-of-minas-gerais-brazil-planned-budget"
},
{
"owner": "os-data",
"branch": "main",
"name": "wesel"
}
]

View File

@@ -98,7 +98,6 @@ export interface TabularDataResource {
key?: string;
path?: string;
size?: number;
bytes?: number;
}
export interface Field {

View File

@@ -5,14 +5,13 @@ export function loadDataPackage(datapackage: FiscalDataPackage, repo): Project {
return {
name: datapackage.name,
title: datapackage.title,
description: datapackage.description || null,
owner: {
name: repo.owner.login,
logo: repo.owner.avatar_url,
// TODO: make this title work
title: repo.owner.login,
},
repo: { name: repo.name, full_name: repo.full_name, url: repo.html_url },
repo: { name: repo, full_name: repo.full_name },
files: datapackage.resources,
author: datapackage.author ? datapackage.author : null,
cityCode: datapackage.cityCode ? datapackage.cityCode : null,

View File

@@ -13,7 +13,8 @@ export interface GithubProject {
export async function getProjectReadme(
owner: string,
repo: string,
branch: string = 'main',
branch: string,
readme: string,
github_pat?: string
) {
const octokit = new Octokit({ auth: github_pat });
@@ -21,7 +22,7 @@ export async function getProjectReadme(
const response = await octokit.rest.repos.getContent({
owner,
repo,
path: 'README.md',
path: readme,
ref: branch,
});
const data = response.data as { content?: string };
@@ -124,6 +125,7 @@ export async function getProject(project: GithubProject, github_pat?: string) {
project.owner,
project.repo,
project.branch,
project.readme,
github_pat
);
@@ -183,43 +185,8 @@ export async function getProjectDataPackage(
}
const decodedContent = Buffer.from(fileContent, 'base64').toString();
const datapackage = JSON.parse(decodedContent);
return { ...datapackage, repo };
return {...datapackage, repo };
} catch (error) {
return null;
}
}
export async function getAllProjectsFromOrg(
org: string,
branch?: string,
github_pat?: string
) {
const octokit = new Octokit({ auth: github_pat });
const repos = await octokit.rest.repos.listForOrg({
org,
type: 'public',
per_page: 100,
});
let failedProjects = [];
const projects = await Promise.all(
repos.data.map(async (_repo) => {
const project = await getProjectDataPackage(
org,
_repo.name,
branch ? branch : 'main',
github_pat
);
if (!project) {
failedProjects.push(_repo.name);
return null;
}
return { datapackage: project, repo: _repo };
})
);
return {
results: projects.filter((item) => item !== null),
failed: failedProjects,
};
}

View File

@@ -5,11 +5,10 @@ import {
export interface Project {
owner: { name: string; logo?: string; title?: string }; // Info about the owner of the data repo
repo: { name: string; full_name: string; url: string }; // Info about the the data repo
repo: { name: string; full_name: string }; // Info about the the data repo
files: TabularDataResource[];
name: string;
title?: string;
description?: string;
author?: string;
cityCode?: string;
countryCode?: string;

View File

@@ -1,3 +0,0 @@
[
"os-data"
]

File diff suppressed because it is too large Load Diff

View File

@@ -10,8 +10,6 @@
"test": "vitest"
},
"dependencies": {
"@githubocto/flat-ui": "^0.14.1",
"@heroicons/react": "^2.0.18",
"@octokit/plugin-throttling": "^5.2.2",
"@types/flexsearch": "^0.7.3",
"@types/node": "18.16.0",
@@ -25,20 +23,17 @@
"next": "13.3.1",
"next-seo": "^6.0.0",
"octokit": "^2.0.14",
"papaparse": "^5.4.1",
"prettier": "^2.8.8",
"react": "18.2.0",
"react-dom": "18.2.0",
"react-hook-form": "^7.43.9",
"react-markdown": "^8.0.7",
"react-query": "^3.39.3",
"react-timeago": "^7.1.0",
"remark-gfm": "^3.0.1",
"typescript": "5.0.4"
},
"devDependencies": {
"@tailwindcss/typography": "^0.5.9",
"@types/papaparse": "^5.3.7",
"autoprefixer": "^10.4.14",
"postcss": "^8.4.23",
"tailwindcss": "^3.3.1",

View File

@@ -0,0 +1,126 @@
import { NextSeo } from 'next-seo';
import { promises as fs } from 'fs';
import path from 'path';
import getConfig from 'next/config';
import { getProject, GithubProject } from '../../../lib/octokit';
import ReactMarkdown from 'react-markdown';
import remarkGfm from 'remark-gfm';
import Breadcrumbs from '../../../components/_shared/Breadcrumbs';
export default function ProjectPage({ project }) {
const repoId = `@${project.repo_config.owner}/${project.repo_config.repo}`
return (
<>
<NextSeo title={`${repoId}${project.base_path !== '/' ? '/' + project.base_path : ''} - GitHub Datasets`} />
<main className="prose mx-auto my-8">
<Breadcrumbs links={[{ title: repoId, href: "" }]} />
<h1 className="mb-0 mt-16">{project.repo_config.name || repoId}</h1>
<p className='mb-8'><span className='font-semibold'>Repository:</span> <a target="_blank" href={project.html_url}>{project.html_url}</a></p>
<h2 className="mb-0 mt-10">Files</h2>
<div className="inline-block min-w-full py-2 align-middle">
<table className="min-w-full divide-y divide-gray-300">
<thead>
<tr>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Name
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Size
</th>
</tr>
</thead>
<tbody className="divide-y divide-gray-200">
{project.files?.map((file) => (
<tr key={file.download_url}>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
<a href={file.download_url}>{file.name}</a>
</td>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{file.size} Bytes
</td>
</tr>
))}
</tbody>
</table>
</div>
{project.readmeContent && <>
<hr />
<h2 className='uppercase font-black'>Readme</h2>
<ReactMarkdown remarkPlugins={[remarkGfm]}>
{project.readmeContent}
</ReactMarkdown>
</>}
</main>
</>
);
}
// Generates `/posts/1` and `/posts/2`
export async function getStaticPaths() {
const jsonDirectory = path.join(
process.cwd(),
'datasets.json'
);
const repos = await fs.readFile(jsonDirectory, 'utf8');
return {
paths: JSON.parse(repos).map((repo) => {
const projectPath =
repo.readme && repo.readme.split('/').length > 1
? repo.readme.split('/').slice(0, -1)
: null;
let path = [repo.name];
if (projectPath) {
projectPath.forEach((element) => {
path.push(element);
});
}
return {
params: { org: repo.owner, path },
};
}),
fallback: false, // can also be true or 'blocking'
};
}
export async function getStaticProps({ params }) {
const jsonDirectory = path.join(
process.cwd(),
'datasets.json'
);
const reposFile = await fs.readFile(jsonDirectory, 'utf8');
const repos: GithubProject[] = JSON.parse(reposFile);
const repo = repos.find((_repo) => {
const projectPath =
_repo.readme && _repo.readme.split('/').length > 1
? _repo.readme.split('/').slice(0, -1)
: null;
let path = [_repo.name];
if (projectPath) {
projectPath.forEach((element) => {
path.push(element);
});
}
return (
_repo.owner == params.org &&
JSON.stringify(path) === JSON.stringify(params.path)
);
});
const github_pat = getConfig().serverRuntimeConfig.github_pat;
const project = await getProject(repo, github_pat);
return {
props: {
project: { ...project, repo_config: repo },
},
};
}

View File

@@ -1,234 +0,0 @@
import { NextSeo } from 'next-seo';
import getConfig from 'next/config';
import {
getAllProjectsFromOrg,
getProjectDataPackage,
getProjectMetadata,
getProjectReadme,
} from '../../../lib/project';
import ReactMarkdown from 'react-markdown';
import remarkGfm from 'remark-gfm';
import { loadDataPackage } from '../../../lib/loader';
import Layout from '../../../components/_shared/Layout';
import Link from 'next/link';
import { Project } from '../../../lib/project.interface';
import ExternalLinkIcon from '../../../components/icons/ExternalLinkIcon';
export default function ProjectPage({
project,
readme,
}: {
project: Project;
readme: string;
}) {
// Get description from datapackage or calculate
// excerpt from README by getting all the content
// up to the first dot.
const description =
project.description || (readme && readme.slice(0, readme.indexOf('.') + 1));
return (
<Layout>
<NextSeo title={`${project.title} - OpenSpending`} />
<main className="prose mx-auto my-8">
<h1 className="mb-1 mt-16">{project.title || project.name}</h1>
<Link target="_blank" href={project.repo.url}>
@{project.repo.full_name}
</Link>
{description && (
<div className="inline-block min-w-full py-2 align-middle mt-5">
{description}
</div>
)}
<div className="inline-block min-w-full py-2 align-middle">
<table className="min-w-full divide-y divide-gray-300">
<thead>
<tr>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Name
</th>
{project.datapackage.countryCode && (
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Country
</th>
)}
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Metadata
</th>
</tr>
</thead>
<tbody className="divide-y divide-gray-200">
<tr>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{project.name}
</td>
{project.datapackage.countryCode && (
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{project.datapackage.countryCode}
</td>
)}
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
<Link
// TODO: this link may be incorrect for some datasets
href={`https://github.com/${project.owner.name}/${project.repo.name}/blob/main/datapackage.json`}
target="_blank"
className="flex items-center hover:text-gray-700"
>
datapackage.json <ExternalLinkIcon className="ml-1" />
</Link>
</td>
</tr>
</tbody>
</table>
</div>
<h3 className="mb-1 mt-10">Data files</h3>
<p>
This dataset contains {project.files.length} file
{project.files.length != 1 ? '' : 's'}
</p>
<div className="inline-block min-w-full py-2 align-middle">
<table className="mt-0 min-w-full divide-y divide-gray-300">
<thead>
<tr>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Name
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Format
</th>
<th
scope="col"
className="px-3 py-3.5 text-left text-sm font-semibold text-gray-900"
>
Size
</th>
</tr>
</thead>
<tbody className="divide-y divide-gray-200">
{project.files?.map((file) => {
let size: number | string = file.size;
if (!size) {
if (file.bytes) {
if (file.bytes > 1000000) {
size = (file.bytes / 1000000).toFixed(2) + ' MB';
} else {
size = (file.bytes / 1000).toFixed(2) + ' kB';
}
}
}
return (
<tr key={file.name}>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{file.name}
</td>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{file.format}
</td>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
{size}
</td>
<td className="whitespace-nowrap px-3 py-4 text-sm text-gray-500">
<Link
target="_blank"
href={
file.path.startsWith('http')
? file.path
: `https://raw.githubusercontent.com/${project.owner.name}/${project.repo.name}/main/${file.path}`
}
>
Download
</Link>
</td>
</tr>
);
})}
</tbody>
</table>
</div>
{readme && (
<>
<hr />
<h2 className="uppercase font-black">Readme</h2>
<ReactMarkdown remarkPlugins={[remarkGfm]}>{readme}</ReactMarkdown>
</>
)}
</main>
</Layout>
);
}
// Generates `/posts/1` and `/posts/2`
export async function getStaticPaths() {
const github_pat = getConfig().serverRuntimeConfig.github_pat;
const allProjects = await getAllProjectsFromOrg(
'os-data',
'main',
github_pat
);
console.log(allProjects)
const paths = allProjects.results.map((project) => ({
params: {
// TODO: dynamize the org
org: 'os-data',
project: project.repo.name,
},
}));
return {
paths,
fallback: false, // can also be true or 'blocking'
};
}
export async function getStaticProps({ params }) {
const { org: orgName, project: projectName } = params;
const github_pat = getConfig().serverRuntimeConfig.github_pat;
const datapackage = await getProjectDataPackage(
orgName,
projectName,
'main',
github_pat
);
const repo = await getProjectMetadata(orgName, projectName, github_pat);
const project = loadDataPackage(datapackage, repo);
// TODO: should this be moved to the loader?
const readme = await getProjectReadme(orgName, projectName, 'main', github_pat);
return {
props: {
project,
readme,
},
};
}

View File

@@ -1,11 +1,13 @@
import { AppProps } from 'next/app';
import Head from 'next/head';
import './styles.css';
import { NextSeo } from 'next-seo';
function CustomApp({ Component, pageProps }: AppProps) {
return (
<>
<NextSeo title="OpenSpending" />
<Head>
<title>GitHub Datasets</title>
</Head>
<main className="app">
<Component {...pageProps} />
</main>

View File

@@ -1,58 +1,65 @@
import { getAllProjectsFromOrg } from '../lib/project';
import { promises as fs } from 'fs';
import path from 'path';
import {
GithubProject,
getProjectDataPackage,
getProjectMetadata,
} from '../lib/octokit';
import getConfig from 'next/config';
import ExternalLinkIcon from '../components/icons/ExternalLinkIcon';
import TimeAgo from 'react-timeago';
import Link from 'next/link';
import { Hero } from '../components/Hero';
import { Header } from '../components/Header';
import { Container } from '../components/Container';
import { FiscalDataPackage } from '../lib/datapackage.interface';
import { loadDataPackage } from '../lib/loader';
import DatasetsSearch from '../components/DatasetsSearch';
import Layout from '../components/_shared/Layout';
export async function getStaticProps() {
// TODO: support other orgs
// const orgsListPath = path.join(process.cwd(), '/orgs.json');
// const orgs = await fs.readFile(orgsListPath, 'utf8');
const jsonDirectory = path.join(process.cwd(), '/datasets.json');
const repos = await fs.readFile(jsonDirectory, 'utf8');
const github_pat = getConfig().serverRuntimeConfig.github_pat;
const datapackages = await Promise.all(
JSON.parse(repos).map(async (_repo: GithubProject) => {
const datapackage = await getProjectDataPackage(
_repo.owner,
_repo.name,
'main',
github_pat
);
const repo = await getProjectMetadata(
_repo.owner,
_repo.name,
github_pat
);
const allProjects = await getAllProjectsFromOrg(
'os-data',
'main',
github_pat
return {
datapackage,
repo,
};
})
);
const projects = allProjects.results.map(
const projects = datapackages.map(
(item: { datapackage: FiscalDataPackage & { repo: string }; repo: any }) =>
loadDataPackage(item.datapackage, item.repo)
);
const availableCountries = projects
.map((item) => item.countryCode)
.filter((v) => v) // Filters false values
.filter((v, i, a) => a.indexOf(v) === i) // Remove duplicates
// TODO: title should be the full name
.map((code) => ({ code, title: code }));
return {
props: {
projects: JSON.stringify(projects),
availableCountries,
},
};
}
export function Home({ projects, availableCountries }) {
export function Datasets({ projects }) {
projects = JSON.parse(projects);
return (
<Layout>
<Hero
countriesCount={availableCountries.length}
datasetsCount={projects.length}
filesCount={projects.reduce(
(partialSum, a) => partialSum + a.files.length,
0
)}
/>
<div className="bg-white min-h-screen">
<Header />
<Hero />
<section className="py-20 sm:py-32">
<Container>
<div className="mx-auto max-w-2xl lg:mx-0">
@@ -67,15 +74,12 @@ export function Home({ projects, availableCountries }) {
</p>
</div>
<div className="mt-10">
<DatasetsSearch
datasets={projects}
availableCountries={availableCountries}
/>
<DatasetsSearch datasets={projects} />
</div>
</Container>
</section>
</Layout>
</div>
);
}
export default Home;
export default Datasets;