Xml resources have invalid formatter

[krwq]

i.e. https://github.com/dotnet/corefx/blob/master/src/System.Private.Xml/src/Resources/Strings.resx#L1722

{facets} instead of i.e. {0}. This is likely causing some random exceptions to some people

[am11]

Seems like there are 13 invalid cases:

Sch_InvalidAllMax: 'all' must have {max occurs}=1.
Sch_InvalidAllElementMax: The {max occurs} of all the particles in the {particles} of an all group must be 0 or 1.
Sch_InvalidExamplar: Cannot be nominated as the {substitution group affiliation} of any other declaration.
Sch_GroupBaseRestNotEmptiable: All particles in the {particles} of the base particle which are not mapped to by any particle in the {particles} of the derived particle should be emptiable - All:All,Sequence:Sequence -- Recurse Rule 2 / Choice:Choice -- RecurseLax.
Sch_AllRefMinMax: The group ref to 'all' must have {min occurs}= 0 or 1 and {max occurs}=1.
Sch_AttributeRestrictionInvalidFromWildcard: The {base type definition} must have an {attribute wildcard} and the {target namespace} of the R's {attribute declaration} must be valid with respect to that wildcard.
Sch_LengthGtBaseLength: It is an error if 'length' is among the members of {facets} of {base type definition} and {value} is greater than the {value} of the parent 'length'.
Sch_MinLengthGtBaseMinLength: It is an error if 'minLength' is among the members of {facets} of {base type definition} and {value} is less than the {value} of the parent 'minLength'.
Sch_MaxLengthGtBaseMaxLength: It is an error if 'maxLength' is among the members of {facets} of {base type definition} and {value} is greater than the {value} of the parent 'maxLength'.
Sch_MaxMinLengthBaseLength: It is an error for both 'length' and either 'minLength' or 'maxLength' to be members of {facets}, unless they are specified in different derivation steps. In which case the following must be true: the {value} of 'minLength' <= the {value} of 'length' <= the {value} of 'maxLength'.
Sch_FacetBaseFixed: Values that are declared as {fixed} in a base type can not be changed in a derived type.
Sch_WhiteSpaceRestriction1: It is an error if 'whiteSpace' is among the members of {facets} of {base type definition}, {value} is 'replace' or 'preserve', and the {value} of the parent 'whiteSpace' is 'collapse'.
Sch_WhiteSpaceRestriction2: It is an error if 'whiteSpace' is among the members of {facets} of {base type definition}, {value} is 'preserve', and the {value} of the parent 'whiteSpace' is 'replace'."

found by navigating to raw view https://raw.githubusercontent.com/dotnet/corefx/master/src/System.Private.Xml/src/Resources/Strings.resx and in the Firefox's console, running:

// edit: works in other new-gen browsers
[...  // spread operator to convert NodeList to Array object
  new DOMParser()
  .parseFromString(
    document
      .querySelector('body > pre').innerText, 'text/xml')
      .documentElement.childNodes
].filter(node => node.querySelector && node.querySelector('data > value') &&
    /\{(\d*[a-zA-Z\s]+\d*)+\}/.test(node.querySelector('data > value').textContent))
 .map(node =>
    `${node.getAttribute('name')}: ${node.querySelector('data > value').textContent}`)
 .join('\n')

[krwq]

@am11 I've done it with just regex: \{[^\d] but that works too

[am11]

It was just to capture mixed cases like {123abc}, {abc123}, {123 abc456 xyz} etc.

[danmoseley]

String.Format throws FormatException if there are more replacement markers than supplied replacements, but does not and has never thrown if there are fewer markers. (For some reason)

So if some {0} etc are missing, I would not expect String.Format to throw.

In other projects, I've sometimes had a Debug.Assert at the point we do the String.Format that asserts that there are not fewer replacement markers than replacements. Assuming we fix this and are otherwise consistent, that might be worth considering. Looks like in this case it's maybe the one in XmlException.CreateMessage().

[danmoseley]

BTW, I see these strings look the same in .NET Framework. I do not believe this meets the bar for 3.0 at this point (assuming it's even a bug and it isn't somehow meant to be this way)

[mrj001]

I'd like to tackle this as a first contribution. I was not able to find any way to claim this issue. I think I need to be added for this to be assigned to me.

I've created pull request #33890 for one of the 13 resources for feedback before tackling all of them.

[danmoseley]

Hi @mrj001 glad to have you on board. I assigned you this issue.