The icu ibm library currently used in the project does not comply with the uts35/ldml standard completely. For example a number pattern #0.0#,# is valid as per uts35 but the library raises an exception for this.
There are test cases by W3C containing such patterns. eg: Validation test case - 282, 285
In ICU IBM library, they have used 2 grouping separators - primary grouping separator and secondary grouping separator. The primary grouping size used for the least significant integer group, and the secondary grouping size used for more significant groups.
We have decided to leave this issue as of now. Reasons for this are,
Lack of clarity on what approach to take with the grouping separators in the fractional part.
We do not deal with the numbers with grouping separators in fractional part.
Effort to get this functionality to the ICU library seems quite large (approx 2 - 3 weeks)
ICU might ask us to implement the same feature for its C++ implementation as well (Contributing to ICU library says, unless its a very simple fix/feature, the developers who open a PR for the Java part should implement it in C++ version as well)
Maybe once everything else is done with this project (CSV-Validation), we can have a go at this issue.
The IDE configuration to use when working on the ICU.IBM project (java part)
The icu ibm library currently used in the project does not comply with the uts35/ldml standard completely. For example a number pattern #0.0#,# is valid as per uts35 but the library raises an exception for this.
There are test cases by W3C containing such patterns. eg: Validation test case - 282, 285
Test cases: 155, 158 are also ignored as of now.
Update:
Issue/bug raised by us https://unicode-org.atlassian.net/browse/ICU-21689 The issue raised by us is already known to them (ICU IBM) and the related issue are https://unicode-org.atlassian.net/browse/CLDR-7166 https://unicode-org.atlassian.net/browse/ICU-10794
In ICU IBM library, they have used 2 grouping separators - primary grouping separator and secondary grouping separator. The primary grouping size used for the least significant integer group, and the secondary grouping size used for more significant groups.
When it comes to the fractional part grouping, it is ambiguous whether there should be 2 grouping separators (primary and secondary). This is the part in uts-35 /ldml which talks about the grouping separators http://www.unicode.org/reports/tr35/tr35-31/tr35-numbers.html#Number_Symbols
We have decided to leave this issue as of now. Reasons for this are,
Lack of clarity on what approach to take with the grouping separators in the fractional part.
We do not deal with the numbers with grouping separators in fractional part.
Effort to get this functionality to the ICU library seems quite large (approx 2 - 3 weeks)
ICU might ask us to implement the same feature for its C++ implementation as well (Contributing to ICU library says, unless its a very simple fix/feature, the developers who open a PR for the Java part should implement it in C++ version as well)
Maybe once everything else is done with this project (CSV-Validation), we can have a go at this issue.
The IDE configuration to use when working on the ICU.IBM project (java part)
https://app.zenhub.com/files/304276433/5e35035f-440d-400b-9a0e-a0f57993f2cc/download