Validator error with config file, fixed with ignoreSubjectConsistency

[ins0mniac2]

Using 1.6.2 with no command line options, when using a config file, we get the following error:

Unhandled rejection (
reason: TypeError: Cannot read property 'split' of undefined
at Minimatch.match (/usr/lib/node_modules/bids-validator/node_modules/minimatch/minimatch.js:717:9)
at minimatch (/usr/lib/node_modules/bids-validator/node_modules/minimatch/minimatch.js:107:42)
at Object.ignoredFile (/usr/lib/node_modules/bids-validator/utils/config.js:11:13)
at Function.format (/usr/lib/node_modules/bids-validator/utils/issues/index.js:65:16)
at Function.exceptionHandler (/usr/lib/node_modules/bids-validator/utils/issues/index.js:191:25)
at validateMisc.then.then.then.then.then.then.then.then.catch.err (/usr/lib/node_modules/bids-validator/validators/bids/fullTest.js:217:35)
at processTicksAndRejections (internal/process/task_queues.js:86:5)
).

The error doesn't occur when we add --ignoreSubjectConsistency option to the command line, which we do to avoid the error described here: https://neurostars.org/t/validator-internal-error-due-to-dataset-size/17829

[sappelhoff]

Hi @ins0mniac2, could you please post the contents of your config file here?

Does the validator run normally if you do not pass a config file?

Did you try running the validator on a different dataset using the config file of the other dataset?

[ins0mniac2]

@sappelhoff , thanks for responding. The validator throws the error whether we explicitly use the "-c" argument to specify the config file or don't use any arguments. We haven't tried it with a different dataset yet. I attached the config file as a zip file -- it wasn't formatted correctly when I pasted the content.
config.zip

[sappelhoff]

for convenience, here it is pasted again (after downloading and unzipping your file)

{
    "ignore": [],
    "warn": [],
    "error": [],
    "ignoredFiles": [
          "/docs/**",
          "/code/**",
          "/ignoredir/**",
          "**/*BAD*.nii.gz",
          "**/*BAD*.json"
    ]

}

If the error also occurs without the config file, it's likely that the config file is not a problem 🤔

Finding out if the validator throws this issue on other datasets would be the next step. you can try some on https://github.com/bids-standard/bids-examples

[ins0mniac2]

Just to clarify, the error does NOT happen when the config file is not there at all, but does happen with it being there, regardless of whether it's specified on the command line with "-c" or not. I assume even without "-c", it is picking it up as the file is saved under the default name it is looking for ( .bids-validator-config.json ) ?

Thanks for posting the content. I need to find the "verbatim" tag in here!

[sappelhoff]

Just to clarify, the error does NOT happen when the config file is not there at all,

ahh okay

I assume even without "-c", it is picking it up as the file is saved under the default name it is looking for ( .bids-validator-config.json ) ?

yes 👍

I need to find the "verbatim" tag in here!

you can use markdown code blocks:

```put the language here
anything written in between backticks is code
```

for example

```python
def fun(a):
    return a**2
```

renders:

def fun(a):
    return a**2

and

```json
{a: true}
```

renders:

{a: true}

even marking a syntax error, because I should have written "a" and not just a.

see: https://guides.github.com/features/mastering-markdown/

[ins0mniac2]

ah, markdown. thanks!

based on the above clarification, do you have any further thoughts or trying on another dataset is what you would still suggest next ?

[sappelhoff]

yes, trying on another dataset would be imporant now. I don't have more ideas unfortunately

[sappelhoff]

Hey @ins0mniac2 I hope you were able to resolve your issue in the meantime. I am closing this now because we'd need more information to clarify whether this is a bug to be fixed, or something else. Happy to re-open if this is still a problem and if you can provide further info.

[mgxd]

We've also hit this in nipreps/fmriprep#2409 after bumping the bids-validator to the 1.7.x series. However, it is a bit strange because:

• It is only happening with one dataset (ds005)
• I haven't been able to replicate it locally

Command and config

bids-validator /tmp/data/ds005 -c /tmp/tmphmn8b3fi

{
  "ignore": [
    "EVENTS_COLUMN_ONSET",
    "EVENTS_COLUMN_DURATION",
    "TSV_EQUAL_ROWS",
    "TSV_EMPTY_CELL",
    "TSV_IMPROPER_NA",
    "VOLUME_COUNT_MISMATCH",
    "BVAL_MULTIPLE_ROWS",
    "BVEC_NUMBER_ROWS",
    "DWI_MISSING_BVAL",
    "INCONSISTENT_SUBJECTS",
    "INCONSISTENT_PARAMETERS",
    "BVEC_ROW_LENGTH",
    "B_FILE",
    "PARTICIPANT_ID_COLUMN",
    "PARTICIPANT_ID_MISMATCH",
    "TASK_NAME_MUST_DEFINE",
    "PHENOTYPE_SUBJECTS_MISSING",
    "STIMULUS_FILE_MISSING",
    "DWI_MISSING_BVEC",
    "EVENTS_TSV_MISSING",
    "TSV_IMPROPER_NA",
    "ACQTIME_FMT",
    "Participants age 89 or higher",
    "DATASET_DESCRIPTION_JSON_MISSING",
    "FILENAME_COLUMN",
    "WRONG_NEW_LINE",
    "MISSING_TSV_COLUMN_CHANNELS",
    "MISSING_TSV_COLUMN_IEEG_CHANNELS",
    "MISSING_TSV_COLUMN_IEEG_ELECTRODES",
    "UNUSED_STIMULUS",
    "CHANNELS_COLUMN_SFREQ",
    "CHANNELS_COLUMN_LOWCUT",
    "CHANNELS_COLUMN_HIGHCUT",
    "CHANNELS_COLUMN_NOTCH",
    "CUSTOM_COLUMN_WITHOUT_DESCRIPTION",
    "ACQTIME_FMT",
    "SUSPICIOUSLY_LONG_EVENT_DESIGN",
    "SUSPICIOUSLY_SHORT_EVENT_DESIGN",
    "MALFORMED_BVEC",
    "MALFORMED_BVAL",
    "MISSING_TSV_COLUMN_EEG_ELECTRODES",
    "MISSING_SESSION"
  ],
  "error": [
    "NO_T1W"
  ],
  "ignoredFiles": [
    "/dataset_description.json",
    "/participants.tsv"
  ]
}

[effigies]

@mgxd I don't think this is the same issue. Or if so, it's likely to get buried because it's closed. Do you want to open a new one?

[yarikoptic]

not sure if it is the same issue, but also not sure if the original issue filed here was resolved (no PR backref), so here is what I get with some elderly validator (well -- all is in container, used for other studies and actually for this one before with less subjects etc -- didn't investigate yet when started to fail)

bids-validator@1.5.8

Unhandled rejection (
  reason: TypeError: Cannot read property 'split' of undefined
    at Minimatch.match (/usr/local/lib/node_modules/bids-validator/node_modules/minimatch/minimatch.js:717:9)
    at minimatch (/usr/local/lib/node_modules/bids-validator/node_modules/minimatch/minimatch.js:107:42)
    at Object.ignoredFile (/usr/local/lib/node_modules/bids-validator/utils/config.js:11:13)
    at Function.format (/usr/local/lib/node_modules/bids-validator/utils/issues/index.js:65:16)
    at Function.exceptionHandler (/usr/local/lib/node_modules/bids-validator/utils/issues/index.js:191:25)
    at validateMisc.then.then.then.then.then.then.then.then.catch.err (/usr/local/lib/node_modules/bids-validator/validators/bids/fullTest.js:217:35)
    at process._tickCallback (internal/process/next_tick.js:68:7)
).

WARNING: validator exited with exit code 3

and the config file used is

{
    "ignore": [
       "TOTAL_READOUT_TIME_NOT_DEFINED"
    ],
    "warn": [],
    "error": [],
    "ignoredFiles": [
          "/.heudiconv/*", "/.heudiconv/*/*", "/.heudiconv/*/*/*", "/.heudiconv/*/*/*/*",  
          "/.heudiconv/.git*", 
          "/.heudiconv/.git/*",
          "/.heudiconv/.git/*/*",
          "/.heudiconv/.git/*/*/*",
          "/.heudiconv/.git/*/*/*/*",
          "/.heudiconv/.git/*/*/*/*/*",
          "/.heudiconv/.git/*/*/*/*/*/*",
          "/.git*", 
          "/.datalad/*", "/.datalad/.*", 
          "/.*/.datalad/*", "/.*/.datalad/.*", 
          "/sub*/ses*/*/*__dup*", "/sub*/*/*__dup*"
    ]
}