r - What are the caveats of using source versus parse & eval? -

short version

can replace

source(filename, local = true, encoding = 'utf-8') 

with

eval(parse(filename, encoding = 'utf-8')) 

without risk of breakage, make utf-8 source files work on windows?

long version

i loading specific source files via

source(filename, local = true, encoding = 'utf-8') 

however, known this not work on windows, full stop.

as workaround, joe cheng suggested using instead

eval(parse(filename, encoding = 'utf-8')) 

this seems work quite well¹ after consulting source code of source, don’t understand how differ in 1 crucial detail:

both source , sys.source not parse , eval file content. instead, parse file content , iterate manually on parsed expressions, , eval them 1 one. not understand why necessary in sys.source (source @ least uses show verbose diagnostics, if instructed; sys.source nothing of kind):

for (i in seq_along(exprs)) eval(exprs[i], envir) 

what purpose of evaling statements separately? , why iterating on indices instead directly on sub-expressions? other caveats there?

_{to clarify: not concerned additional parameters of source , parse, of may set via options.}

---

¹ reason source tripped encoding parse isn’t boils down fact source attempts convert input text. parse no such thing, reads file’s byte content as-is , marks encoding utf-8 in memory.

this not full answer addresses seq_along part of question, lengthy include comments.

one key difference between seq_along followed [ vs using for in x approach (which believe similar seq_along followed [[ instead of [) former preserves expression. here example illustrate difference:

> txt <- "x <- 1 + 1 + # abnormal expression +   2 * +     3 + " > x <- parse(text=txt, keep.source=true) >  > for(i in x) print(i) x <- 1 + 1 2 * 3 > for(i in seq_along(x)) print(x[i]) expression(x <- 1 + 1) expression(2 *     3) 

alternatively:

> attributes(x[[2]]) null > attributes(x[2]) $srcref $srcref[[1]] 2 *     3 

whether has practical impact when comparing eval(parse(..., keep.source=t)), can could, can't imagine situation does.

note subsetting expression separately leads srcref business getting subset, conceivably useful (...maybe?).