Break issue - ComponentRange error

[neemanjabu]

Reproduction steps:

1. run solver as vrp-cli solve pragmatic problem.json (two problem.json attached, both cause the same error)
2. error occurs with the following message:

thread 'main' panicked at vrp-pragmatic/src/lib.rs:51:97:
called `Result::unwrap()` on an `Err` value: ComponentRange { name: "timestamp", minimum: -377705116800, maximum: 253402300799, value: 9223372036854775807, conditional_range: false }
stack backtrace:
   0:     0x713215e158b2 - <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt::h56c396e6ec1a6d25
   1:     0x713215dc2a9b - core::fmt::write::h8d9f62e4196a6251
   2:     0x713215dedeb2 - std::io::Write::write_fmt::hfeda4b672809964d
   3:     0x713215e1ab89 - std::sys_common::backtrace::print::h92792842c9200fcf
   4:     0x713215e1a3ae - std::panicking::default_hook::{{closure}}::h4fa40162feafd3ea
   5:     0x713215e1b5d4 - std::panicking::rust_panic_with_hook::h33e85abbf3f92159
   6:     0x713215e1aef2 - std::panicking::begin_panic_handler::{{closure}}::h6e89e0968a11dcc7
   7:     0x713215e1ae49 - std::sys_common::backtrace::__rust_end_short_backtrace::h6f80b72bde67d6fa
   8:     0x713215e1ae36 - rust_begin_unwind
   9:     0x713215c24ec2 - core::panicking::panic_fmt::hb37c2646452904e3
  10:     0x713215c252d5 - core::result::unwrap_failed::h95d9de04097dd2c4
  11:     0x713215f47a4d - vrp_pragmatic::format_time::hfb0f8b0c521a6f24
  12:     0x713215fc6310 - vrp_pragmatic::format::solution::solution_writer::create_tour::h951bc14512f91ca7
  13:     0x713215fc37c7 - vrp_pragmatic::format::solution::solution_writer::create_solution::hbe1d921567365674
  14:     0x713215c84d19 - vrp_pragmatic::format::solution::write_pragmatic::h0eafee619269b34d
  15:     0x713215d638e3 - vrp_cli::commands::solve::add_pragmatic::{{closure}}::h199290cbbed79869
  16:     0x713215d69af4 - vrp_cli::commands::solve::run_solve::h8817575db3bbc162
  17:     0x713215d7e678 - vrp_cli::cli::run_subcommand::h564d9de34c729393
  18:     0x713215d6fb66 - vrp_cli::main::hc0f39bb4a070fa2e
  19:     0x713215c91cb3 - std::sys_common::backtrace::__rust_begin_short_backtrace::h1cf440c697908a17
  20:     0x713215d81716 - main

It seems that the issue lies somewhere in having a break with a short duration and the break time window at the same time as the job's time window end.

problem.json
problem2.json

[reinterpretcat]

Thanks for reporting! I can confirm the issue: there is an overflow happens in activity place schedule times. I would need to find time to dig into it to understand what went wrong..

[reinterpretcat]

Basically, the reason is that you're using required break feature which is kind of experimental:

	"breaks": [
		{
			"time": {
				"earliest": "2025-02-10T12:00:00.000Z",
				"latest": "2025-02-10T12:00:00.000Z"
			},
			"duration": 300
		}
	]

One possible solution is to switch to optional break feature which is considered as a stable:

	"breaks": [
		{
			"time": ["2025-02-10T12:00:00.000Z", "2025-02-10T12:00:00.000Z"],
			"places": [{ "duration": 300 }]
		}
	]

The crash still needs to be fixed, it is good that you provided minimalistic reproducible example, this will help a lot when I find time for it.