Fix string format

[wch]

With R-devel, as of r85619 (r-devel/r-svn@f374bbf0), on my build machines I started getting failures when building rpart, because that commit added some string safety checks.

rpart.c: In function 'rpart':
rpart.c:203:9: error: format not a string literal and no format
arguments [-Werror=format-security]
  203 |         error(errmsg);
      |         ^~~~~

My understanding is that on most machines, this particular line of code raises a warning, but on mine it is an error. I'm not sure exactly why this is, but it seems like a good idea to address this anyway -- it is similar to a fix we recently had to do for a number of packages that CRAN sent emails about. For example:

• httpuv:
  • issue: warning: format string is not a string literal  rstudio/httpuv#388
  • fix: https://github.com/rstudio/httpuv/pull/389/files
• later:
  • issue: warning: format string is not a string literal r-lib/later#181
  • fix https://github.com/r-lib/later/pull/182/files

The change in this pull request does the same thing as those other pull requests.

[bethatkinson]

Kurt Hornik sent a note today suggesting changing error(errmsg); to error("%s", errmsg); Hopefully that fixes the issue. I'll try to get this submitted tomorrow. From: Winston Chang ***@***.***> Sent: Thursday, November 30, 2023 4:31 PM To: bethatkinson/rpart ***@***.***> Cc: Subscribed ***@***.***> Subject: [EXTERNAL] [bethatkinson/rpart] Fix string format (PR #52) With R-devel, as of r85619 ***@***.***<r-devel/r-svn@f374bbf0>), on my build machines I started getting failures when building rpart, because that commit added some string safety checks. rpart.c: In function 'rpart': rpart.c:203:9: error: format not a string literal and no format arguments [-Werror=format-security] 203 | error(errmsg); | ^~~~~ My understanding is that on most machines, this particular line of code raises a warning, but on mine it is an error. I'm not sure exactly why this is, but it seems like a good idea to address this anyway -- it is similar to a fix we recently had to do for a number of packages that CRAN sent emails about. For example: * httpuv: * issue: rstudio/httpuv#388<rstudio/httpuv#388> * fix: https://github.com/rstudio/httpuv/pull/389/files * later: * issue: r-lib/later#181<r-lib/later#181> * fix https://github.com/r-lib/later/pull/182/files The change in this pull request does the same thing as those other pull requests.

…

________________________________ You can view, comment on, or merge this pull request online at: #52 Commit Summary * 997e38e<997e38e> Fix string format File Changes (1 file<https://github.com/bethatkinson/rpart/pull/52/files>) * M src/rpart.c<https://github.com/bethatkinson/rpart/pull/52/files#diff-094c155360459643fa6e9239d9fcbe86f0604ca28ce2612afa267c8bf060f8f8> (2) Patch Links: * https://github.com/bethatkinson/rpart/pull/52.patch * https://github.com/bethatkinson/rpart/pull/52.diff - Reply to this email directly, view it on GitHub<#52>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ACWQG57XBSZQHP4SF2EGAXLYHECJVAVCNFSM6AAAAABAB5QZJOVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAYTSNRRGQ4DKMY>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.******@***.***>>

[wch]

Great, I'm glad to hear that Kurt also suggested the same fix.

[bethatkinson]

Fix should show up on cran shortly From: Winston Chang ***@***.***> Sent: Thursday, November 30, 2023 4:31 PM To: bethatkinson/rpart ***@***.***> Cc: Subscribed ***@***.***> Subject: [EXTERNAL] [bethatkinson/rpart] Fix string format (PR #52) With R-devel, as of r85619 ***@***.***<r-devel/r-svn@f374bbf0>), on my build machines I started getting failures when building rpart, because that commit added some string safety checks. rpart.c: In function 'rpart': rpart.c:203:9: error: format not a string literal and no format arguments [-Werror=format-security] 203 | error(errmsg); | ^~~~~ My understanding is that on most machines, this particular line of code raises a warning, but on mine it is an error. I'm not sure exactly why this is, but it seems like a good idea to address this anyway -- it is similar to a fix we recently had to do for a number of packages that CRAN sent emails about. For example: * httpuv: * issue: rstudio/httpuv#388<rstudio/httpuv#388> * fix: https://github.com/rstudio/httpuv/pull/389/files * later: * issue: r-lib/later#181<r-lib/later#181> * fix https://github.com/r-lib/later/pull/182/files The change in this pull request does the same thing as those other pull requests.

…

________________________________ You can view, comment on, or merge this pull request online at: #52 Commit Summary * 997e38e<997e38e> Fix string format File Changes (1 file<https://github.com/bethatkinson/rpart/pull/52/files>) * M src/rpart.c<https://github.com/bethatkinson/rpart/pull/52/files#diff-094c155360459643fa6e9239d9fcbe86f0604ca28ce2612afa267c8bf060f8f8> (2) Patch Links: * https://github.com/bethatkinson/rpart/pull/52.patch * https://github.com/bethatkinson/rpart/pull/52.diff - Reply to this email directly, view it on GitHub<#52>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ACWQG57XBSZQHP4SF2EGAXLYHECJVAVCNFSM6AAAAABAB5QZJOVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAYTSNRRGQ4DKMY>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.******@***.***>>