Michael comments

Results 20 comments of


                                            Michael

Enhance validator output to make filtering by error type more straightforward

This is definitely needed to use at scale and / or programmatically.

@quoted seems to have no effect on validation result

> As noted in the documentation, in practice the CSV libraries tend to handle quoting (ie strip it) before the tests are applied, so this is not implemented in the...

@quoted seems to have no effect on validation result

@DavidUnderdown @adamretter I was only looking for insight - ironically this might break my schema, since the csv files I'm working with only have quotes "where needed", e.g. when a...

@quoted seems to have no effect on validation result

@DavidUnderdown @adamretter it would be great if `quoted` support could be optional in a future release 😉

Provide a list of commonly used schemas

What's the expected schema format? Does csvlint accept CSV Schema Language V1.1? Reference: https://github.com/digital-preservation/csv-schema

Update link to JSON Table Schema specification and explain variations

Or this language: https://github.com/digital-preservation/csv-schema ?

Update link to JSON Table Schema specification and explain variations

@Stephen-Gates @quadrophobiac seems like the variations are substantial. I tried to use a schema with minimum lengths of 1 for all columns, with any empty values i.e. "" indicated as...

Allow additional columns

From a purist point of view, isn't the option to allow extensions part of the schema definition?

Clarification: @optional means full, empty, or partially full?

Thanks for clarifying. So "column" is shorthand for "the cell in this column in this row". As I surmised.

Infinite number of (unnamed) columns

Forgive my naiveté @adbre but couldn't the CSV files be pre-proccessed to pad them with max - n blank columns before validation, where max = most columns seen (this time)...