Cypress terminal report

This documentation is for cypress >= 12.0.0. For older versions see 4.x.x or 3.x.x branch.
Limitations • Install • Options • Integrations • After/before all • Logging to files • Development • Release Notes
Plugin for cypress that adds better terminal output for easier debugging. Prints cy commands, browser console logs, cy.request and cy.intercept data. Great for your pipelines.
- looks pretty in console
- logs all commands, requests and browser console logs
- supports logging to files
- option between logging only on failure (default) or always
- options for trimming and compacting logs
- support for multiple and nested mocha contexts
- log commands from before all and after all hooks (with a catch*)
Try it out by cloning cypress-terminal-report-demo.
Limitations and notes
- By default logs are not printed for successful tests. Please see option to change this.
console.log
usage was never meant to be used in the cypress test code. Using it will not log anything with this plugin. Using it also goes against the queue nature of cypress. Usecy.log
instead. See here for more details.
Install
Requirements
>=4.0.0
requires cypress>=10.0.0
and node>=12
>=3.0.0
requires cypress>=4.10.0
<3.0.0
requires cypress>=3.8.0
Install npm package.
npm i --save-dev cypress-terminal-report
- If using typescript and es6 imports ensure
esModuleInterop
is enabled. - Register the output plugin in
cypress.config.{js|ts}
module.exports = defineConfig({ e2e: { setupNodeEvents(on, config) { require('cypress-terminal-report/src/installLogsPrinter')(on) } } })
- Register the log collector support in
cypress/support/e2e.{js|ts}
require('cypress-terminal-report/src/installLogsCollector')()
Options
Options for the plugin install
require('cypress-terminal-report/src/installLogsPrinter')(on, options)
options.defaultTrimLength
integer; default: 800; Max length of cy.log and console.warn/console.error.
options.commandTrimLength
integer; default: 800; Max length of cy commands.
options.routeTrimLength
integer; default: 5000; Max length cy.request and XHR data.
options.compactLogs
integer?; default: null; If it is set to a number greater or equal to 0, this amount of logs
will be printed only around failing commands. Use this to have shorter output especially
for when there are a lot of commands in tests. When used with options.printLogsToConsole=always
for tests that don't have any severity=error
logs nothing will be printed.
options.outputCompactLogs
integer? | false; default: null; Overrides options.compactLogs
for the file log output specifically,
when options.outputTarget
is specified. Allows compacting of the terminal and the file output logs to different levels.
If options.outputCompactLogs
is unspecified, file output will use options.compactLogs
.
If set to false
, output file logs will not compact even if options.compactLogs
is set.
options.outputRoot
string; default: null; Required if options.outputTarget
provided. More details.
options.specRoot
string; default: null; Cypress specs root relative to package json. More details.
options.outputTarget
object; default: null; Output logs to files. More details.
options.printLogsToConsole
string; Default: 'onFail'. When to print logs to console, possible values: 'always', 'onFail', 'never' - When set to always logs will be printed to console for successful tests as well as failing ones.
options.printLogsToFile
string; Default: 'onFail'. When to print logs to file(s), possible values: 'always', 'onFail', 'never' - When set to always logs will be printed to file(s) for successful tests as well as failing ones.
options.includeSuccessfulHookLogs
boolean; Default: false. Commands from before all and after all hooks by default get logged only if
a command from them failed. This default is in accordance with the defaults on options.printLogsTo*
to
avoid printing too many, possibly irrelevant, information. However you can set this to true
if you
need more extensive logging, but be aware that commands will be logged to terminal from these hooks
regardless whether there were failing tests in the suite. This is because we can't know for sure in
advanced if a test fails or not.
options.outputVerbose
boolean; default: true; Toggles verbose output. Currently just writes out additional file write information, if any.
options.debug
boolean; default: false; Toggles debug output. Useful in cases of difficult to reproduce issues with the plugin.
options.collectTestLogs
*1
([spec, test, state], {type, message, severity}[]) => void; default: undefined;
Callback to collect each test case's logs after its run.
The first argument contains information about the test: the spec
(test file), test
(test title) and state
(test state) fields.
The second argument contains the test logs. 'type' is from the same list as for the collectTypes
support install option (see below). Severity can be of ['', 'error', 'warning'].
options.logToFilesOnAfterRun
boolean; default: false;
When set to true it enables additional log write pass to files using the cypress after:run
plugin
hook. This option can only be used with cypress 6.2.0 onwards, and with the additional
experimentalRunEvents
configuration on versions smaller than 6.7.0.
Options for the support install
require('cypress-terminal-report/src/installLogsCollector')(options);
options.collectTypes
array; default: ['cons:log','cons:info', 'cons:warn', 'cons:error', 'cy:log', 'cy:xhr', 'cy:request', 'cy:intercept', 'cy:command'] What types of logs to collect and print. By default all types are enabled. The 'cy:command' is the general type that contain all types of commands that are not specially treated.
options.commandTimings
null | 'timestamp' | 'seconds'; default: null; Record and log the time when the logs were generated. 'timestamp' is the epoch timestamp in milliseconds. 'seconds' is the number of seconds since the test start, displaying the milliseconds as decimals.
options.filterLog
null | ({type, message, severity}) => boolean; default: undefined;
Callback to filter logs manually.
The type is from the same list as for the collectTypes
option. Severity can be of ['', 'error', 'warning'].
options.processLog
null | ({type, message, severity}) => {type, message, severity}; default: undefined;
Callback to process logs manually.
The type is from the same list as for the collectTypes
option. Severity can be of ['', 'error', 'warning'].
options.collectTestLogs
*2
(mochaRunnable, {type, message, severity}[]) => void; default: undefined;
Callback to collect each test case's logs after its run.
The mochaRunnable
is of type Test | Hook
from the mocha library.
The type is from the same list as for the collectTypes
option. Severity can be of ['', 'error', 'warning'].
options.xhr.printBody
boolean; default true; Whether to print response data. Controls request body as well when printRequestData
is
enabled. Note that currently response body is logged only on failing requests and when this config is enabled.
options.xhr.printHeaderData
boolean; default false; Whether to print header data for XHR requests.
options.xhr.printRequestData
boolean; default false; Whether to print request data for XHR requests besides response data.
options.enableExtendedCollector
boolean; default false; Enables an extended collector which will also collect command logs from before all and after all hooks.
options.enableContinuousLogging
boolean; default false; Enables logging logs to terminal continuously / immediately as they are registered. This feature is unstable and has an impact on pipeline performance. This option has no effect for extended collector, only works for the simple collector. Use only for debugging purposes in case the pipelines / tests are timing out.
NOTE: In case of this option enabled, logs will come before the actual title of the test. Also the
printLogsToConsole
option will be ignored. Logging to files might also get impacted.
Example for options for the support install
// ...
// Options for log collector
const options = {
// Log console output only
collectTypes: ['cons:log', 'cons:info', 'cons:warn', 'cons:error']
}
// Register the log collector
require('cypress-terminal-report/src/installLogsCollector')(options)
// ...
Integrations
cypress-fail-fast
Logging to files does not work out of the box. To enable support use the
logToFilesOnAfterRun
option.
cypress-mochawesome-reporter
The following example demonstrates adding logs to context for all tests (snippet from e2e.js
):
import 'cypress-mochawesome-reporter/register'
afterEach(() => {
cy.wait(50, {log: false}).then(() => cy.addTestContext(Cypress.TerminalReport.getLogs('txt')))
})
// Ensure that after plugin installation is after the afterEach handling the integration.
require('cypress-terminal-report/src/installLogsCollector')()
For typescript support add to your tsconfig types cypress-terminal-report
.
Logging after all and before all hooks
Commands from before all and after all hooks are not logged by default. A new experimental feature introduces
support for logging commands from these hooks: enableExtendedCollector
.
This feature is by default disabled as it relies much more heavily on internals of cypress and
mocha, thus there is a higher chance of something breaking, especially with cypress upgrades.
Once the feature enabled, logs from these hooks will only appear in console if:
- a command from the hook fails
- hook passes and
printLogsToConsole
==always
- hook passes and
printLogsToConsole
==onFail
andincludeSuccessfulHookLogs
==true
Global after all
hooks need to be registered before the registration of the support install
, otherwise
they will not be added to file outputs, if such is configured. Example e2e.js
:
after(() => cy.log('this log will appear in the output files'))
require('cypress-terminal-report/src/installLogsCollector')(config)
after(() => cy.log('this log will NOT appear in the files'))
Logging to files
To enable logging to file you must add the following configuration options to the plugin install.
setupNodeEvents(on, config) {
// ...
const options = {
outputRoot: config.projectRoot + '/logs/',
outputTarget: {
'out.txt': 'txt',
'out.json': 'json',
'out.html': 'html',
}
};
require('cypress-terminal-report/src/installLogsPrinter')(on, options);
// ...
}
The outputTarget
needs to be an object where the key is the relative path of the
file from outputRoot
and the value is the type of format to output.
Supported types: txt
, json
, html
.
Log specs in separate files
To create log output files per spec file instead of one single file change the
key in the outputTarget
to the format {directory}|{extension}
, where
{directory}
the root directory where to generate the files and {extension}
is the file extension for the log files. The generated output will have the
same structure as in the cypress specs root directory.
setupNodeEvents(on, config) {
const options = {
outputRoot: config.projectRoot + '/logs/',
// Used to trim the base path of specs and reduce nesting in the generated output directory.
specRoot: 'cypress/e2e',
outputTarget: {
'cypress-logs|json': 'json',
}
};
}
Custom output log processor
If you need to output in a custom format you can pass a function instead of a string
to the outputTarget
value. This function will be called with the list of messages
per spec per test. It is called right after one spec finishes, which means on each
iteration it will receive for one spec the messages. See for example below.
NOTE: The chunks have to be written in a way that after every write the file is in a valid format. This has to be like this since we cannot detect when cypress runs the last test. This way we also make the process faster because otherwise the more tests would execute the more RAM and processor time it would take to rewrite all the logs to the file.
Inside the function you will have access to the following API:
this.size
- Current char size of the output file.this.atChunk
- The count of the chunk to be written.this.initialContent
- The initial content of the file. Defaults to''
. Set this before the first chunk write in order for it to work.this.chunkSeparator
- Chunk separator string. Defaults to''
. This string will be written between each chunk. If you need a special separator between chunks use this as it is internally handled to properly write and replace the chunks.this.writeSpecChunk(specPath, dataString, positionInFile?)
- Writes a chunk of data in the output file.
// ...
const options = {
outputTarget: {
'custom.output': function (allMessages, options) {
// allMessages= {[specPath: string]: {[testTitle: string]: [type: string, message: string, severity: string][]}}
Object.entries(allMessages).forEach(([spec, tests]) => {
let text = `${spec}:\n`
Object.entries(tests).forEach(([test, messages]) => {
text += ` ${test}\n`
messages.forEach(({type, message, severity}) => {
text += ` ${type} (${severity}): ${message}\n`
})
})
// .. Process the tests object into desired format ..
// Insert chunk into file, by default at the end.
this.writeSpecChunk(spec, text)
// Or before the last two characters.
this.writeSpecChunk(spec, text, -2)
})
}
}
}
// ...
See JsonOutputProcessor implementation as a good example demonstrating both conversion of data into string and chunk write position alternation.
HTML output log processor
The HTML output processor has default styles:
body {
font-family: monospace;
}
p {
margin: 0;
padding: 0;
}
pre {
display: inline;
margin: 0;
}
h2 {
margin: 0;
font-size: 1.2em;
}
To pass custom CSS styles:
const HtmlOutputProcessor = require('cypress-terminal-report/src/outputProcessor/HtmlOutputProcessor')
// ...
setupNodeEvents(on, config) {
const options = {
// ...
outputTarget: {
'html': function (this) {
return new HtmlOutputProcessor(
this.file,
this.options,
`
/* Custom CSS */
`
)
},
};
}
Development
Testing
Tests can be found under /test
. The primary expectations are run with mocha and these tests in fact
start cypress run instances and assert on their output. So that means there is a cypress suite that
is used to emulate the usage of the plugin, and a mocha suite to assert on those emulations.
To add tests you need to first add a case to existing cypress spec or create a new one and then
add the case as well in the /test/test.js
. To run the tests you can use npm test
in the test \
directory. You should add it.only
to the test case you are working on to speed up development.