R-dev-day issue 86 changes

Author: eddelbuettel

src/library/base/R/dcf.R

@@ -75,6 +75,10 @@ function(file, fields = NULL, all = FALSE, keep.white = NULL)
 
     lines <- readLines(file, skipNul = TRUE, encoding = "bytes")
 
+    ## Remove comment lines (if any)
+    comments <- grepl("^#", lines)
+    if (any(comments)) lines <- lines[!comments]
+
     ## Try to find out about invalid things: mostly, lines which do not
     ## start with blanks but have no ':' ...
     ind <- grep(paste0("^[^", ascii_blank, "][^:]*$"), lines)

src/library/base/man/dcf.Rd

@@ -106,8 +106,7 @@ write.dcf(x, file = "", append = FALSE, useBytes = FALSE,
   \url{https://www.debian.org/doc/debian-policy/ch-controlfields.html}.
 
   Note that \R does not require encoding in UTF-8, which is a recent
-  Debian requirement.  Nor does it use the Debian-specific sub-format
-  which allows comment lines starting with \samp{#}.
+  Debian requirement.
 }
 
 \note{

src/main/dcf.c

@@ -72,7 +72,7 @@ attribute_hidden SEXP do_readDCF(SEXP call, SEXP op, SEXP args, SEXP env)
     bool blank_skip, field_skip = false;
     int whatlen, dynwhat, buflen = 8096; // was 100, but that re-alloced often
     char *line, *buf;
-    regex_t blankline, contline, trailblank, regline, eblankline;
+    regex_t blankline, contline, trailblank, regline, eblankline, commentline;
     regmatch_t regmatch[1];
     SEXP file, what, what2, retval, retval2, dims, dimnames;
     Rconnection con = NULL;
@@ -122,6 +122,7 @@ attribute_hidden SEXP do_readDCF(SEXP call, SEXP op, SEXP args, SEXP env)
     tre_regcompb(&contline, "^[ \t]+", REG_EXTENDED);
     tre_regcompb(&regline, "^[^:]+:[ \t]*", REG_EXTENDED);
     tre_regcompb(&eblankline, "^[ \f\n\r\t\v]+\\.[ \f\n\r\t\v]*$", REG_EXTENDED);
+    tre_regcompb(&commentline, "^#", REG_EXTENDED);
 
     k = 0;
     lastm = -1; /* index of the field currently being recorded */
@@ -149,6 +150,8 @@ attribute_hidden SEXP do_readDCF(SEXP call, SEXP op, SEXP args, SEXP env)
 		field_fold = true;
 		n_eblanklines = 0;
 	    }
+        } else if(tre_regexecb(&commentline, line, 0, 0, 0) == 0) {
+            /* comment-line detected. skipping it per DCF specification */
 	} else {
 	    blank_skip = false;
 	    if(tre_regexecb(&contline, line, 1, regmatch, 0) == 0) {
@@ -308,6 +311,7 @@ attribute_hidden SEXP do_readDCF(SEXP call, SEXP op, SEXP args, SEXP env)
     tre_regfree(&trailblank);
     tre_regfree(&regline);
     tre_regfree(&eblankline);
+    tre_regfree(&commentline);
 
     if(!blank_skip) k++;