Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
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
63 changes: 60 additions & 3 deletions lib/fetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -2,24 +2,33 @@
// SPDX-License-Identifier: MIT

const axios = require('axios')
const { default: axiosRetry } = require('axios-retry')

const defaultHeaders = Object.freeze({ 'User-Agent': 'clearlydefined.io crawler (clearlydefined@outlook.com)' })
const defaultMaximumAttempts = 3

axios.defaults.headers.common['User-Agent'] = defaultHeaders['User-Agent']

// Create a shared axios instance with retry configured for the default case
const defaultRetryInstance = axios.create()
axiosRetry(defaultRetryInstance, {
retries: defaultMaximumAttempts - 1,
retryDelay: (retryCount, error) => {
return axiosRetry.exponentialDelay(retryCount, error, 100)
}
})

function buildRequestOptions(request) {
let responseType = 'text'
if (request.json) {
responseType = 'json'
} else if (request.encoding === null) {
responseType = 'stream'
}

const validateOptions = {}
if (request.simple === false) {
validateOptions.validateStatus = () => true
}

return {
method: request.method,
url: request.url || request.uri,
Expand Down Expand Up @@ -62,4 +71,52 @@ async function getStream(opt) {
return await callFetch(request)
}

module.exports = { callFetch, withDefaults, defaultHeaders, getStream }
async function callFetchWithRetry(url, options = {}, retryOptions = {}) {
const { maxAttempts = defaultMaximumAttempts, ...otherRetryOpts } = retryOptions

// Use shared instance if using default settings, otherwise create a new one
let axiosInstance = defaultRetryInstance
const hasCustomRetryOptions = maxAttempts !== defaultMaximumAttempts || Object.keys(otherRetryOpts).length > 0

if (hasCustomRetryOptions) {
axiosInstance = axios.create()
axiosRetry(axiosInstance, {
retries: maxAttempts - 1,
retryDelay: (retryCount, error) => {
return axiosRetry.exponentialDelay(retryCount, error, 100)
},
...otherRetryOpts
})
}

const resolveWithFullResponse = options.resolveWithFullResponse ?? true
const request = { url, ...options, resolveWithFullResponse }
try {
const response = await callFetch(request, axiosInstance)
if (!resolveWithFullResponse) return response
return {
statusCode: response.status,
headers: response.headers,
body: response.data,
request: response.request
}
} catch (err) {
if (err.response && resolveWithFullResponse) {
return {
statusCode: err.response.status,
headers: err.response.headers,
body: err.response.data,
request: err.response.request
}
}
throw err
}
}

module.exports = {
callFetch,
callFetchWithRetry,
withDefaults,
defaultHeaders,
getStream
}
4 changes: 2 additions & 2 deletions lib/sourceDiscovery.js
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,6 @@

const { get, uniq, uniqWith } = require('lodash')
const ghrequestor = require('ghrequestor')
const request = require('requestretry')
const geit = require('geit')
// TODO why not parse-github-repo-url (10x more downloads)
const parseGitHubUrl = require('parse-github-url')
Expand Down Expand Up @@ -75,7 +74,8 @@ async function discoverFromGitHubRefs(version, candidate, options) {
headers,
maxAttempts: 3,
retryDelay: 250,
retryStrategy: request.RetryStrategies.HTTPOrNetworkError,
// TODO This is the default strategy in axios retry. We ghrequestor currently depends on requestretry, but we will replace it with axios retry soon.
// retryStrategy: request.RetryStrategies.HTTPOrNetworkError,
tokenLowerBound: 10,
json: true
})
Expand Down
23 changes: 7 additions & 16 deletions package-lock.json

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

1 change: 0 additions & 1 deletion package.json
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,6 @@
"parse-github-url": "^1.0.3",
"patch-package": "^8.0.1",
"qlimit": "^0.1.1",
"requestretry": "^8.0.0",
"semver": "^7.7.4",
"sha.js": "^2.4.12",
"spdx-correct": "^3.2.0",
Expand Down
19 changes: 9 additions & 10 deletions providers/fetch/nugetFetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@ const fs = require('fs')
const { mkdirp } = require('mkdirp')
const path = require('path')
const { promisify } = require('util')
const requestRetry = require('requestretry').defaults({ maxAttempts: 3, fullResponse: true })
const { callFetchWithRetry: requestRetry } = require('../../lib/fetch')
const FetchResult = require('../../lib/fetchResult')

const providerMap = {
Expand Down Expand Up @@ -69,9 +69,8 @@ class NuGetFetch extends AbstractFetch {
// https://docs.microsoft.com/en-us/nuget/api/registration-base-url-resource
// Example: https://api.nuget.org/v3/registration5/moq/4.8.2.json and follow catalogEntry
// https://api.nuget.org/v3/registration5-gz-semver2/microsoft.powershell.native/7.0.0-preview.1.json
const { body, statusCode } = await requestRetry.get(
`${baseUrl}/v3/registration5-gz-semver2/${spec.name.toLowerCase()}/${spec.revision}.json`,
{ gzip: true }
const { body, statusCode } = await requestRetry(
`${baseUrl}/v3/registration5-gz-semver2/${spec.name.toLowerCase()}/${spec.revision}.json`
)
return statusCode !== 200 || !body ? null : JSON.parse(body)
}
Expand All @@ -87,7 +86,7 @@ class NuGetFetch extends AbstractFetch {
// https://docs.microsoft.com/en-us/nuget/api/package-base-address-resource
// Example: https://api.nuget.org/v3-flatcontainer/moq/index.json
const baseUrl = providerMap.nuget
const { body, statusCode } = await requestRetry.get(`${baseUrl}/v3-flatcontainer/${name}/index.json`, {
const { body, statusCode } = await requestRetry(`${baseUrl}/v3-flatcontainer/${name}/index.json`, {
json: true
})
// If statusCode is not 200, XML may be returned
Expand All @@ -99,9 +98,9 @@ class NuGetFetch extends AbstractFetch {
}

async _getPackage(zip, packageContentUrl) {
const response = await requestRetry(packageContentUrl, { json: false, encoding: null })
return new Promise((resolve, reject) => {
requestRetry
.get(packageContentUrl, { json: false, encoding: null })
response.body
.pipe(fs.createWriteStream(zip))
.on('finish', () => resolve(null))
.on('error', reject)
Expand All @@ -110,7 +109,7 @@ class NuGetFetch extends AbstractFetch {

async _getManifest(catalogEntryUrl) {
// Example: https://api.nuget.org/v3/catalog0/data/2018.10.29.04.23.22/xunit.core.2.4.1.json
const { body, statusCode } = await requestRetry.get(catalogEntryUrl)
const { body, statusCode } = await requestRetry(catalogEntryUrl)
if (statusCode !== 200) return null
return JSON.parse(body)
}
Expand All @@ -119,7 +118,7 @@ class NuGetFetch extends AbstractFetch {
async _getNuspec(spec) {
// https://docs.microsoft.com/en-us/nuget/api/package-base-address-resource#download-package-manifest-nuspec
// Example: https://api.nuget.org/v3-flatcontainer/newtonsoft.json/11.0.1/newtonsoft.json.nuspec
const { body, statusCode } = await requestRetry.get(
const { body, statusCode } = await requestRetry(
`https://api.nuget.org/v3-flatcontainer/${spec.name.toLowerCase()}/${spec.revision}/${spec.name.toLowerCase()}.nuspec`
)
if (statusCode !== 200) return null
Expand All @@ -146,7 +145,7 @@ class NuGetFetch extends AbstractFetch {
if (licenseUrl.toLowerCase().includes('license_url_here_or_delete_this_line')) return
const downloadedLicenseDirName = path.join(dirName, 'clearlydefined', 'downloaded')
await mkdirp(downloadedLicenseDirName)
const { body, statusCode } = await requestRetry.get(licenseUrl)
const { body, statusCode } = await requestRetry(licenseUrl)
if (statusCode !== 200) return
await promisify(fs.writeFile)(path.join(downloadedLicenseDirName, 'LICENSE'), body)
}
Expand Down
5 changes: 2 additions & 3 deletions providers/fetch/packagistFetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -2,10 +2,9 @@
// SPDX-License-Identifier: MIT

const AbstractFetch = require('./abstractFetch')
const requestRetry = require('requestretry').defaults({ maxAttempts: 3, fullResponse: true })
const fs = require('fs')
const { get } = require('lodash')
const { getStream } = require('../../lib/fetch')
const { getStream, callFetchWithRetry: requestRetry } = require('../../lib/fetch')
const { promisify } = require('util')
const readdir = promisify(fs.readdir)
const FetchResult = require('../../lib/fetchResult')
Expand Down Expand Up @@ -40,7 +39,7 @@ class PackagistFetch extends AbstractFetch {

async _getRegistryData(spec) {
const baseUrl = providerMap.packagist
const { body, statusCode } = await requestRetry.get(`${baseUrl}p2/${spec.namespace}/${spec.name}.json`, {
const { body, statusCode } = await requestRetry(`${baseUrl}p2/${spec.namespace}/${spec.name}.json`, {
json: true
})
if (statusCode !== 200 || !body) return null
Expand Down
5 changes: 2 additions & 3 deletions providers/fetch/podFetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -3,12 +3,11 @@

const { clone, get } = require('lodash')
const AbstractFetch = require('./abstractFetch')
const { callFetch: request } = require('../../lib/fetch')
const { callFetch: request, callFetchWithRetry: requestRetry } = require('../../lib/fetch')
const fs = require('fs')
const path = require('path')
const crypto = require('crypto')
const { exec } = require('child_process')
const requestRetry = require('requestretry').defaults({ maxAttempts: 3, fullResponse: true })
const FetchResult = require('../../lib/fetchResult')

const services = {
Expand Down Expand Up @@ -131,7 +130,7 @@ class PodFetch extends AbstractFetch {

async _getVersion(spec) {
// Example: https://trunk.cocoapods.org/api/v1/pods/SwiftLCS
const { body, statusCode } = await requestRetry.get(`${services.trunk}/pods/${spec.name}`, {
const { body, statusCode } = await requestRetry(`${services.trunk}/pods/${spec.name}`, {
json: true
})

Expand Down
3 changes: 1 addition & 2 deletions providers/fetch/pypiFetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,6 @@
// SPDX-License-Identifier: MIT

const AbstractFetch = require('./abstractFetch')
const requestRetry = require('./requestRetryWithDefaults')
const nodeRequest = require('../../lib/fetch')
const fs = require('fs')
const spdxCorrect = require('spdx-correct')
Expand Down Expand Up @@ -46,7 +45,7 @@ class PyPiFetch extends AbstractFetch {

async _getRegistryData(spec) {
const baseUrl = providerMap.pypi
const { body, statusCode } = await requestRetry.get(`${baseUrl}/pypi/${spec.name}/json`, {
const { body, statusCode } = await nodeRequest.callFetchWithRetry(`${baseUrl}/pypi/${spec.name}/json`, {
json: true
})
if (statusCode !== 200 || !body) return null
Expand Down
3 changes: 0 additions & 3 deletions providers/fetch/requestRetryWithDefaults.js

This file was deleted.

4 changes: 2 additions & 2 deletions providers/fetch/rubyGemsFetch.js
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@

const AbstractFetch = require('./abstractFetch')
const { getStream } = require('../../lib/fetch')
const requestRetry = require('requestretry').defaults({ maxAttempts: 3, fullResponse: true })
const requestRetry = require('../../lib/fetch').callFetchWithRetry
const fs = require('fs')
const zlib = require('zlib')
const path = require('path')
Expand Down Expand Up @@ -48,7 +48,7 @@ class RubyGemsFetch extends AbstractFetch {

async _getRegistryData(spec) {
const baseUrl = providerMap.rubyGems
const { body, statusCode } = await requestRetry.get(`${baseUrl}/api/v1/gems/${spec.name}.json`, {
const { body, statusCode } = await requestRetry(`${baseUrl}/api/v1/gems/${spec.name}.json`, {
json: true
})
return statusCode === 200 && body ? body : null
Expand Down
19 changes: 12 additions & 7 deletions providers/process/top.js
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,8 @@ const ghrequestor = require('ghrequestor')
const linebyline = require('linebyline')
const path = require('path')
const Request = require('../../ghcrawler').request
const requestRetry = require('requestretry').defaults({ json: true, maxAttempts: 3, fullResponse: false })
const requestRetry = require('../../lib/fetch').callFetchWithRetry
const defaultOptions = { json: true, resolveWithFullResponse: false }

class TopProcessor extends AbstractProcessor {
canHandle(request) {
Expand Down Expand Up @@ -89,7 +90,8 @@ class TopProcessor extends AbstractProcessor {
if (!end || end - start <= 0) end = start + 1000
const initialOffset = Math.floor(start / 36) * 36
for (let offset = initialOffset; offset < end; offset += 36) {
const response = await requestRetry.get(`https://www.npmjs.com/browse/depended?offset=${offset}`, {
const response = await requestRetry(`https://www.npmjs.com/browse/depended?offset=${offset}`, {
...defaultOptions,
headers: { 'x-spiferack': 1 }
})
const packages = response.packages || []
Expand Down Expand Up @@ -154,8 +156,9 @@ class TopProcessor extends AbstractProcessor {
if (!end || end - start <= 0) end = start + 1000
for (let offset = start; offset < end; offset += 100) {
const page = offset / 100 + 1
const response = await requestRetry.get(
`https://crates.io/api/v1/crates?page=${page}&per_page=100&sort=downloads`
const response = await requestRetry(
`https://crates.io/api/v1/crates?page=${page}&per_page=100&sort=downloads`,
defaultOptions
)
const requestsPage = response.crates.map(
x => new Request('package', `cd:/crate/cratesio/-/${x.name}/${x.max_version}`)
Expand Down Expand Up @@ -316,8 +319,9 @@ class TopProcessor extends AbstractProcessor {
if (!start || start < 0) start = 0
if (!end || end - start <= 0) end = start + 1000
for (let offset = start; offset < end; offset += pageSize) {
const topComponents = await requestRetry.get(
`https://api-v2v3search-0.nuget.org/query?prerelease=false&skip=${offset}&take=${pageSize}`
const topComponents = await requestRetry(
`https://api-v2v3search-0.nuget.org/query?prerelease=false&skip=${offset}&take=${pageSize}`,
defaultOptions
)
const requests = topComponents.data.map(component => {
return new Request('package', `cd:/nuget/nuget/-/${component.id}`)
Expand Down Expand Up @@ -349,7 +353,8 @@ class TopProcessor extends AbstractProcessor {
})
const requests = []
for (let i = 0; i < repos.length; i++) {
const commits = await requestRetry.get(`https://api.github.com/repos/${namespace}/${repos[i].name}/commits`, {
const commits = await requestRetry(`https://api.github.com/repos/${namespace}/${repos[i].name}/commits`, {
...defaultOptions,
headers
})
if (commits.length > 0) {
Expand Down
Loading