From 30d4fdbc3a7e609f8b7f49b0c910f1a29be0ce0d Mon Sep 17 00:00:00 2001 From: Larry Golding Date: Sun, 18 Nov 2018 10:41:18 -0800 Subject: [PATCH] #283: Note closure in Editor's Report. --- EditorsReports/Editor's report 2018-11-28.md | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/EditorsReports/Editor's report 2018-11-28.md b/EditorsReports/Editor's report 2018-11-28.md index b2b19e14..1c7fab53 100644 --- a/EditorsReports/Editor's report 2018-11-28.md +++ b/EditorsReports/Editor's report 2018-11-28.md @@ -29,7 +29,13 @@ Presented at TC Meeting #28, November 28th, 2018 1. [Issue #278](https://github.com/oasis-tcs/sarif-spec/issues/278): "Should the sections be reordered?" - This is an editorial issue raised by Jim. Michael and I discussed and we feel the current ordering approach reasonable as a whole (although we might decide to make small changes consistent with the current approach). + This is an editorial issue raised by Jim. Michael and I discussed and we feel the current ordering approach reasonable as a whole (although we might decide to make small changes consistent with the current approach). + + 1. [Issue #283](https://github.com/oasis-tcs/sarif-spec/issues/283): "result.message SHALL be present constraint dropped from schema" + + Both the spec and the schema are correct. The spec says that _either_ `message` must be present (if the log specifies its messages "inline"), _or_ `ruleMessageId` must be present (if the log just provides a resource identifier together that locates the message string), or both. + + Now even if the log file specifies a message indirectly _via_ `ruleMessageId`, you still need `message` if the resource string has any replacement sequences `{n}` that need to be filled from `message.arguments`. But if the resource string has no replacement sequences, you don't need `message` at all. 1. I made the following changes at editorial discretion: