inline expressions do not permit errors any more; closes #651

NEWS.md

@@ -54,6 +54,8 @@
 
 - the default value for the chunk option `tidy` is `FALSE` now, which means the R source code in chunks will no longer be reformatted by `formatR::tidy.source()` by default; this feature must be explicitly turned on by `tidy=TRUE`, and it has brought a lot of confusion in the past, so it is perhaps a good idea not to reformat the source code by default
 
+- inline R expressions will no longer be evaluated in `try()`, which means errors in inline R code will be emitted immediately
+
 - the first argument of the `plot` hook is the filename of the plot now; in previous versions, it was a vector of length 2 (basename and file extension); see `?hook_plot`
 
 - in the previous version, we can set `options(knitr.foo = value)` so that **knitr** can adjust the package options `opts_knit$set(foo = value)` before knitting a document; now the prefix for package options has been changed to `knitr.package.`, i.e. we should set `options(knitr.package.foo)` to achieve `opts_knit$set(foo)`; besides, it is also possible to change the default chunk options using `options(knitr.chunk.foo)` now, but you are warned that this may bring reproducibility issues, so please use with care

R/block.R

@@ -295,7 +295,7 @@ call_inline = function(block) {
 
 inline_exec = function(
   block, eval = eval_lang(opts_chunk$get('eval')), envir = knit_global(),
-  error = eval_lang(opts_chunk$get('error')), hook = knit_hooks$get('inline')
+  hook = knit_hooks$get('inline')
 ) {
 
   # run inline code and substitute original texts
@@ -305,19 +305,13 @@ inline_exec = function(
   loc = block$location
   for (i in 1:n) {
     res = if (eval) {
-      (if (error) try else identity)(
-        {
-          v = withVisible(eval(parse_only(code[i]), envir = envir))
-          if (v$visible) v$value
-        }
-      )
+      v = withVisible(eval(parse_only(code[i]), envir = envir))
+      if (v$visible) v$value
     } else '??'
     d = nchar(input)
     # replace with evaluated results
     str_sub(input, loc[i, 1], loc[i, 2]) = if (length(res)) {
-      if (inherits(res, 'try-error')) {
-        knit_hooks$get('error')(str_c('\n', res, '\n'), opts_chunk$get())
-      } else hook(res)
+      paste(hook(res), collapse = '')
     } else ''
     if (i < n) loc[(i + 1):n, ] = loc[(i + 1):n, ] - (d - nchar(input))
     # may need to move back and forth because replacement may be longer or shorter

R/template.R

@@ -118,5 +118,5 @@ knit_expand = function(file, ..., text = readLines(file, warn = FALSE),
   env = list(...)
   env = if (length(env)) list2env(env, parent = parent.frame()) else parent.frame()
   inline_exec(list(code = mat, input = txt, location = loc),
-              eval = TRUE, envir = env, error = FALSE, hook = identity)
+              eval = TRUE, envir = env, hook = identity)
 }

inst/examples/knitr-spin.R

@@ -20,8 +20,7 @@ mean(x)
 #' is the mean of x plus 2.
 #' The code itself may contain braces, but these are not checked.  Thus,
 #' perfectly valid (though very strange) R code such as `{{2 + 3}} - {{4 - 5}}`
-#' can produce different results compared to simply running the script:
-{{2 + 3}} - {{4 - 5}}
+#' can lead to errors because `2 + 3}} - {{4 - 5` will be treated as inline code.
 #'
 #' Now we continue writing the report. We can draw plots as well.
 

inst/examples/knitr-spin.Rmd

@@ -22,8 +22,7 @@ You can use the special syntax {{code}} to embed inline expressions, e.g.
 is the mean of x plus 2.
 The code itself may contain braces, but these are not checked.  Thus,
 perfectly valid (though very strange) R code such as `{{2 + 3}} - {{4 - 5}}`
-can produce different results compared to simply running the script:
-`r 2 + 3}} - {{4 - 5`
+can lead to errors because `2 + 3}} - {{4 - 5` will be treated as inline code.
 
 Now we continue writing the report. We can draw plots as well.