Resolves #34

Changes: No short option for --version (default in swift-argument-parser, can only be added back as a separate option), only outputs the number, without "Version: " (This can be easily changed back to the original behaviour) Added -qq for quieter as mentioned in #37 ~~--is-ci gets the default value from the CI environment variable used by most CI solutions.~~ (This would mean that specifying the --is-ci option sets it to false) XcbeautifyLib doesn't depend on the argument parser, only xcbeautify depends on it

Todo: Update other package managers/build tools, I only updated Package.swift

output of xcbeautify -h:

OVERVIEW: Format the output of `xcodebuild` and `swift`

EXAMPLE: xcodebuild test -project MyApp.xcodeproj -scheme MyApp -destination
'platform=iOS Simulator,OS=12.0,name=iPhone X' | xcbeautify

USAGE: xcbeautify [--quiet] [-qq] [--is-ci]

OPTIONS:
  -q, --quiet             Only print tasks that have warnings or errors. 
  -qq, --quieter          Only print tasks that have errors. 
  --is-ci                 Print test result too under quiet/quieter flag. 
  --version               Show the version.
  -h, --help              Show help information.

Update:

I have noticed xcbueatify cannot properly handle output from Xcode 11.2.1. So it may be a good idea to support different format of Xcode.

#################################

Could xcbeautify support different versions of OS and Xcode? If not yet, could these be implemented as new command line options?

Thanks!

The current implementation of xcbeautify performs regex checking for each line of the xcodebuild output. Due to the big amount of possible regexes, each line on average is checked for about N/2 times. which is a lot

Suggested implementation holds an array of 'inner parser' and try to reuse 'line parser/matchers' using frequency based array (basically, last matcher regex will be checked first). Due to the non-random structure of xcodebuil output, this approach allows to decrease xcbeautify running time from 160 s to 9 seconds when used on the xcodebuild output of size 240Mb

| Implementation | time | Log Size | |-|-|-| | xcbeautify-main | 209s | 256M | | xcbeautify-speedup | 8s | 256M | | xcpretty | 23s | 256M |

Testing approach: time cat LOGFILE | xcbeauitify

Possible 'bugs'

If there are multiple regexes which can match same line, this approach could return different results, depending on the order of the commands, appeared in log

Xcbeautify Speed

It seems that xcbeautify is waaaay slower than xcpretty. This PR will actually make xcbeautify faster

This was a weekend hack a few years ago to work around the parallel testing feature in Xcode 9 not being supported by xcpretty. Even though I don't use this anymore, it seems like there are still people that depends on it. I don't have time in maintaining this project for much longer, but also don't want to break anyone's builds. If you're interested in taking it over, please contact me.

Resolves https://github.com/thii/xcbeautify/issues/17.

• adds "CI" workflow
  • build release version
  • run tests and pipe output into release version just built
• adds some status badges on Readme 🙂

Hi, thanks for the great project.

One thing I'm missing from it is the warnings suppression - Travis cancel jobs after 4MB of logs and building React Native projects for IOS often generates many warnings for the third party libraries.

Using some flag like --errors-only would be very helpful in that case

• Change JunitComponent to capture the Regex used to create the component
• Relies on capture groups being consistent between parallel and non-parallel output
• Add unit test

Motivation

Fixes https://github.com/fastlane/fastlane/pull/19629#issuecomment-1013511227

The test summary printed after a run was showing wrong results if:

• Multiple test suites were being run
• Any tests were skipped

Expected: Tests Passed: 0 failed, 955 total (12.187 seconds) Actual: Tests Passed: 0 failed, 56 total (0.324 seconds)

The issue was because the test summary line parser assumed that there was only one line summary being printed. The line looks like 👇

Executed 8 tests, with 4 failures (0 unexpected) in 0.154 (0.155) seconds

However... this line gets printed multiple times after each suite gets run 👇

Test Case '-[Tests_iOS.Tests_iOS testExample5ThatSometimesFails]' failed (0.019 seconds).                                                              
Test Suite 'Tests_iOS' failed at 2022-01-15 21:31:49.073.                                                                                              
         Executed 8 tests, with 4 failures (0 unexpected) in 0.154 (0.155) seconds                                                                     
Test Suite 'Tests iOS.xctest' failed at 2022-01-15 21:31:49.073.                                                                                       
         Executed 8 tests, with 4 failures (0 unexpected) in 0.154 (0.155) seconds                                                                     
Test Suite 'All tests' failed at 2022-01-15 21:31:49.073.                                                                                              
         Executed 8 tests, with 4 failures (0 unexpected) in 0.154 (0.156) seconds

There is also a case that was unaccounted when tests were skipped 👇

Executed 56 tests, with 3 tests skipped and 0 failures (0 unexpected) in 1.019 (1.029) seconds 

Description

• Introduced 2 new line matchers for when a suite is finished for "All tests"
• Introduced 1 new line matcher for "Executed with skipped tests"

Previously, the testSummary in Parser would get override by the latest summary which was the root cause of this issue.

Now, a testSummary will only get parsed when it follows either Test Suite 'All tests' passed at or Test Suite 'All tests' failed at. There can also be multiple summaries so this will also add all the summaries together for:

• test count
• skipped count
• failure count
• unexpected count
• time

It also adds logic now for parsing the two different variation of the "executed" lines:

• Executed 8 tests, with 4 failures (0 unexpected) in 0.154 (0.155) seconds
• Executed 56 tests, with 3 tests skipped and 0 failures (0 unexpected) in 1.019 (1.029) seconds

Fastlane ships with xcpretty but when running tests in parallel, it doesnot create correct output. Hence we want to use xcbeautify with fastlane. Is there any way we can use that with action run_tests? https://docs.fastlane.tools/actions/run_tests/

Is it possible to add a command line options to display warnings or errors only? Because for a huge project, most of output "[XXX] Compiling YYY.cpp" are of less importance and they would hide real issues.

    NSArray *lines  = [log componentsSeparatedByString:@"\n"];
    Parser *parser = [[Parser alloc] init];
    for (NSString *line in lines) {
        NSString *beautyLog = [parser parseWithLine:line colored:NO];
        if (beautyLog.length) {
            DDLogInfo(@"%@", beautyLog);
        }
    }

ibsystem_kernel.dylib         0x7fff6c5b4afe read$NOCANCEL + 10
libsystem_c.dylib              0x7fff6c500b25 __srefill1 + 24
libsystem_c.dylib              0x7fff6c4fd017 getdelim + 251
libswiftCore.dylib             0x7fff6bc18903 swift_stdlib_readLine_stdin + 35
libswiftCore.dylib             0x7fff6b9a9465 $ss8readLine16strippingNewlineSSSgSb_tF + 37
OCObfuscator                   0x10004497a $sSS12OCObfuscatorE8beautify7pattern7coloredSSSgAA7PatternO_SbtF + 7242
OCObfuscator                   0x1000396d2 $s12OCObfuscator6ParserC5parse4line7coloredSSSgSS_SbtF + 1826
OCObfuscator                   0x10003ae9c $s12OCObfuscator6ParserC5parse4line7coloredSSSgSS_SbtFTo + 92
OCObfuscator                   0x10000b2cd -[ObfuscatorTool receivedData:] + 1101
CoreFoundation                 0x7fff34d9a35f __CFNOTIFICATIONCENTER_IS_CALLING_OUT_TO_AN_OBSERVER__ + 12
CoreFoundation                 0x7fff34d9a2f3 ___CFXRegistrationPost1_block_invoke + 63
CoreFoundation                 0x7fff34d9a268 _CFXRegistrationPost1 + 372
CoreFoundation                 0x7fff34d99ebe ___CFXNotificationPost_block_invoke + 97
CoreFoundation                 0x7fff34d697e2 -[_CFXNotificationRegistrar find:object:observer:enumerator:] + 1575
CoreFoundation                 0x7fff34d68c82 _CFXNotificationPost + 1351
Foundation                     0x7fff373eea22 -[NSNotificationCenter postNotificationName:object:userInfo:] + 59
Foundation                     0x7fff375271f3 _performFileHandleSource + 1106
CoreFoundation                 0x7fff34da4b21 __CFRUNLOOP_IS_CALLING_OUT_TO_A_SOURCE0_PERFORM_FUNCTION__ + 17
CoreFoundation                 0x7fff34da4ac0 __CFRunLoopDoSource0 + 103
CoreFoundation                 0x7fff34da48d4 __CFRunLoopDoSources0 + 209
CoreFoundation                 0x7fff34da3740 __CFRunLoopRun + 1272
CoreFoundation                 0x7fff34da2bd3 CFRunLoopRunSpecific + 499
HIToolbox                      0x7fff338f865d RunCurrentEventLoopInMode + 292
HIToolbox                      0x7fff338f82a9 ReceiveNextEventCommon + 356
HIToolbox                      0x7fff338f8127 _BlockUntilNextEventMatchingListInModeWithFilter + 64
AppKit                         0x7fff31f68ba4 _DPSNextEvent + 990
AppKit                         0x7fff31f67380 -[NSApplication(NSEvent) _nextEventMatchingEventMask:untilDate:inMode:dequeue:] + 1352
AppKit                         0x7fff31f5909e -[NSApplication run] + 658
AppKit                         0x7fff31f2b465 NSApplicationMain + 777
OCObfuscator                   0x100017e1f main + 47
libdyld.dylib                  0x7fff6c4727fd start + 1

then i get the dead lock code

    private func formatCompileWarning(pattern: Pattern) -> String? {
        let groups = capturedGroups(with: pattern)
        let filePath = groups[0]
        let reason = groups[2]

        // Read 2 additional lines to get the warning line and cursor position
        let line: String = readLine() ?? "" // ===>this is the dead lock !!!!!!!!!!!!!
        let cursor: String = readLine() ?? ""
        ....
       .....
}

Where can i find more information on the usage of xcbeautify?

For e.g. xcodebuild [flags] | xcbeautify

• what are the different flags that can be used here?

How to generate a JUnit report from xcbeautify?

 xcodebuild test -project "MyApp.xcodeproj" -scheme "MyApp" -sdk iphonesimulator -destination "OS=16.0,name=iPhone 13 mini,platform=iOS Simulator" -derivedDataPath "~/temp2" -destination-timeout 300 -parallel-testing-worker-count 4 -parallel-testing-enabled YES |  ~/.mint/bin/xcbeautify --is-ci --quiet --report junit --junit-report-filename junitresult.xml --report-path . 

The resulting file junitresult.xml shows 0 tests run :(

My project in xcode14 would build with some error like this:

❌  error: Dependency for P1:target-MyApp-fce55c0e14fedcc57d73ec039f604d040bdc1e3e5eb2ad770e53e15b1d5e91df-:Ad-Hoc-Mirror:SwiftDriver Compilation Requirements xiaoheihe normal arm64 com.apple.xcode.tools.swift.compiler is not absolute (Pods/Headers/Public/ActionSheetPicker_3_0/ActionSheetPicker-3.0.modulemap). (in target 'MyApp' from project 'MyApp')

And xcbeautify will stock after encounter such error.

This PR mainly refactors how Regex implementation is used. We now have a Matching protocol, ideally, in the near future, we could experiment with different implementations dynamically.

I am writing a package for compiling and running apps using Emacs and I pipe the output from xcodebuild to xcbeautify but Emacs wont recognize the errors when this tool adds warning/info and error emojis. Is it possible to disable them somehow for just when an error occurs?