Auto-normalize versions in input flags and from parsing of indexes.

[cmaglie]

Please check if the PR fulfills these requirements

See how to contribute

• The PR has no duplicates (please search among the Pull Requests
  before creating one)
• The PR follows
  our contributing guidelines
• Tests for the changes have been added (for bug fixes / features)
• Docs have been added / updated (for bug fixes / features)
• UPGRADING.md has been updated with a migration guide (for breaking changes)
• configuration.schema.json updated if new parameters are added.

What kind of change does this PR introduce?

The versions parsed from command input flags and JSON indexes are now automatically "normalized", i.e. 2.0 is automatically converted to 2.0.0.

What is the current behavior?

Sometimes there are discrepancies from the command output and the accepted command input, as described in detail in #1727.

$ ./arduino-cli lib install [email protected]
Downloading [email protected]...
Installing [email protected]...
Installed [email protected]
$ ./arduino-cli lib list --format json                      
[
  {
    "library": {
      "name": "ILI9341_t3",
      "author": "Limor Fried, Adafruit, Paul Stoffregen",
      "maintainer": "Paul Stoffregen",
      "sentence": "Optimized ILI9341 (320x240 Color TFT) Display",
      "website": "https://github.com/PaulStoffregen/ILI9341_t3",
      "category": "Display",
      "architectures": [
        "*"
      ],
      "install_dir": "/Users/a.kitta/Documents/Arduino/libraries/ILI9341_t3",
      "source_dir": "/Users/a.kitta/Documents/Arduino/libraries/ILI9341_t3",
      "real_name": "ILI9341_t3",
      "version": "1.0",              <----- listed as "1.0"
      "license": "Unspecified",
      "location": 1,
      "provides_includes": [
        "ILI9341_t3.h"
      ]
    }
  }
]
$ ./arduino-cli lib install "ILI9341_t3"@1.0               <--- did not accept "1.0"
Error installing ILI9341_t3: Library '[email protected]' not found

What is the new behavior?

$ ./arduino-cli lib install [email protected]
Downloading [email protected]...
Installing [email protected]...
Installed [email protected]
$ ./arduino-cli lib list --format json                      
[
  {
    "library": {
      "name": "ILI9341_t3",
      "author": "Limor Fried, Adafruit, Paul Stoffregen",
      "maintainer": "Paul Stoffregen",
      "sentence": "Optimized ILI9341 (320x240 Color TFT) Display",
      "website": "https://github.com/PaulStoffregen/ILI9341_t3",
      "category": "Display",
      "architectures": [
        "*"
      ],
      "install_dir": "/Users/a.kitta/Documents/Arduino/libraries/ILI9341_t3",
      "source_dir": "/Users/a.kitta/Documents/Arduino/libraries/ILI9341_t3",
      "real_name": "ILI9341_t3",
      "version": "1.0.0",
      "license": "Unspecified",
      "location": 1,
      "provides_includes": [
        "ILI9341_t3.h"
      ]
    }
  }
]
$ ./arduino-cli lib install [email protected]
Already installed [email protected]

Does this PR introduce a breaking change, and is titled accordingly?

The change should be backward compatible.

Other information

Fix #1727

It is based on bugst/relaxed-semver#11, it should be merged before this one and the go.mod file updated accordingly.

[cmaglie]

Ok, looks like it's not so simple:

    debug_test.go:70: 
        	Error Trace:	/home/runner/work/arduino-cli/arduino-cli/commands/debug/debug_test.go:70
        	Error:      	Not equal: 
        	            	expected: "/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/data_dir/packages/arduino-test/tools/arm-none-eabi-gcc/7-2017q4/bin/arm-none-eabi-gdb --interpreter=console -ex set remotetimeout 5 -ex target extended-remote | \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/data_dir/packages/arduino-test/tools/openocd/0.10.0-arduino7/bin/openocd\" -s \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/data_dir/packages/arduino-test/tools/openocd/0.10.0-arduino7/share/openocd/scripts/\" --file \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/custom_hardware/arduino-test/samd/variants/arduino_zero/openocd_scripts/arduino_zero.cfg\" -c \"gdb_port pipe\" -c \"telnet_port 0\" /home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/hello/build/arduino-test.samd.arduino_zero_edbg/hello.ino.elf"
        	            	actual  : "{runtime.tools.arm-none-eabi-gcc-7-2017q4.path}/bin/arm-none-eabi-gdb --interpreter=console -ex set remotetimeout 5 -ex target extended-remote | \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/data_dir/packages/arduino-test/tools/openocd/0.10.0-arduino7/bin/openocd\" -s \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/data_dir/packages/arduino-test/tools/openocd/0.10.0-arduino7/share/openocd/scripts/\" --file \"/home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/custom_hardware/arduino-test/samd/variants/arduino_zero/openocd_scripts/arduino_zero.cfg\" -c \"gdb_port pipe\" -c \"telnet_port 0\" /home/runner/work/arduino-cli/arduino-cli/commands/debug/testdata/hello/build/arduino-test.samd.arduino_zero_edbg/hello.ino.elf"

The effect of "normalize-everything" is more disruptive than expected, it seems we have to keep the relaxed version otherwise the properties {runtime.tools.arm-none-eabi-gcc-7-2017q4.path} will fail to resolve.

I'll look for another solution...