Skip to content
This repository has been archived by the owner on Nov 16, 2023. It is now read-only.

add an event retro fit command #5

Merged
merged 2 commits into from
Mar 6, 2017
Merged
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
add an event retro fit command
  • Loading branch information
jeffmcaffer committed Mar 6, 2017
commit 0107b413845c89f09a6cbc2303614971176d4ef2
71 changes: 66 additions & 5 deletions bin/cc
Original file line number Diff line number Diff line change
Expand Up @@ -2,12 +2,15 @@
// Copyright (c) Microsoft Corporation. All rights reserved.
// Licensed under the MIT License.

const argSplit = require('argv-split');
const commander = require('commander');
const CrawlerClient = require('../lib/crawlerClient');
const readline = require('readline');
const split = require('argv-split');
const Q = require('q');
const qlimit = require('qlimit');
const request = require('request');
const split = require('split');
const throttle = require('throttle');

const crawlerClient = new CrawlerClient();
let promise = Q();
Expand Down Expand Up @@ -39,7 +42,7 @@ function getCommands() {
.description('Stop all processing in the crawler')
.action(() => configureCount(0));
commands
.command('queue <requests>')
.command('queue <requests...>')
.description('Queue the given list of orgs and/or repos to be processed.')
.action(requests => queueRequests(requests));
commands
Expand All @@ -66,6 +69,11 @@ function getCommands() {
.option('-d, --delete <urn>', 'Delete the deadletter with the given urn')
.description('Manage deadletters')
.action(options => deadletters(options));
commands
.command('events <url>')
.option('-z, --test', 'Test mode. Process but do not actually queue the events.')
.description('Backfill the events in the resource at the URL. The resource is assumed to be one JSON object per line')
.action((url, options) => queueEvents(url, options));
commands
.command('exit')
.description('Exit this tool')
Expand All @@ -83,7 +91,7 @@ function startReplLoop(commands) {
rl.prompt();

rl.on('line', (line) => {
const command = split(line);
const command = argSplit(line);
// pad up the command line to keep commander happy
command.unshift('node', 'cc');
commands.parse(command);
Expand Down Expand Up @@ -121,8 +129,8 @@ function setTokens(tokens) {
}

function queueRequests(specs) {
let requests = JSON.parse(specs);
requests = Array.isArray(requests) ? requests : [requests];
let requests = Array.isArray(specs) ? specs : [specs];
requests = requests.map(request => request.trim().startsWith('{') ? JSON.parse(request) : request);
promise = crawlerClient.queueRequests(requests).then(() => console.log(`Queued ${requests.length} requests`));
}

Expand Down Expand Up @@ -175,3 +183,56 @@ function requeueDeadletter(urn) {
function deleteDeadletter(urn) {
promise = crawlerClient.deleteDeadletter(urn).then(() => console.log(`Deleted ${urn}`));
}

function queueEvents(url, options) {
let requests = [];
let count = 0;
const response = request(url);
response
.pipe(split())
.on('data', (line) => {
const request = createEventRequest(line);
if (!request) {
return;
}
requests.push(request);
if (requests.length === 10 && !options.test) {
const toQueue = requests.slice();
requests = [];
response.pause();
count += 10;
crawlerClient.queueRequests(toQueue, 'later').then(
() => {
console.log(`Queued ${count} events so far...`);
response.resume();
},
error =>
console.log(error)
);
}
})
.on('error', error => {
console.log(error);
})
.on('end', () => {
console.log('ended');
});
}

function createEventRequest(line) {
try {
const event = JSON.parse(line);
delete event._id;
const eventUrlBase = event.repo ? event.repo.url : event.org.url;
const request = { type: event.type, url: `${eventUrlBase}/events/${event.id}` };
request.payload = { body: event, etag: 1 };
if (event.created_at) {
request.payload.fetchedAt = event.created_at;
}
request.policy = 'default:self';
return request;
} catch (error) {
console.log(error);
return null;
}
}
3 changes: 2 additions & 1 deletion package.json
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,8 @@
"q": "1.4.1",
"qlimit": "^0.1.1",
"readline": "^1.3.0",
"request": "^2.79.0"
"request": "^2.79.0",
"split": "^1.0.0"
},
"devDependencies": {}
}