//
// Split_Lines: C
//
// Given a string series, split lines on CR-LF. Give back array of strings.
//
// Note: The definition of "line" in POSIX is a sequence of characters that
// end with a newline. Hence, the last line of a file should have a newline
// marker, or it's not a "line")
//
// https://stackoverflow.com/a/729795
//
// This routine does not require it.
//
// !!! CR support is likely to be removed...and CR will be handled as a normal
// character, with special code needed to process it.
//
REBARR *Split_Lines(const REBVAL *str)
{
REBDSP dsp_orig = DSP;
REBCNT len = VAL_LEN_AT(str);
REBCNT i = VAL_INDEX(str);
if (i == len)
return Make_Array(0);
DECLARE_MOLD (mo);
Push_Mold(mo);
REBCHR(const*) cp = VAL_STRING_AT(str);
REBUNI c;
cp = NEXT_CHR(&c, cp);
for (; i < len; ++i, cp = NEXT_CHR(&c, cp)) {
if (c != LF && c != CR) {
Append_Codepoint(mo->series, c);
continue;
}
Init_Text(DS_PUSH(), Pop_Molded_String(mo));
SET_CELL_FLAG(DS_TOP, NEWLINE_BEFORE);
Push_Mold(mo);
if (c == CR) {
REBCHR(const*) tp = NEXT_CHR(&c, cp);
if (c == LF) {
++i;
cp = tp; // treat CR LF as LF, lone CR as LF
}
}
}
// If there's any remainder we pushed in the buffer, consider the end of
// string to be an implicit line-break
if (STR_SIZE(mo->series) == mo->offset)
Drop_Mold(mo);
else {
Init_Text(DS_PUSH(), Pop_Molded_String(mo));
SET_CELL_FLAG(DS_TOP, NEWLINE_BEFORE);
}
return Pop_Stack_Values_Core(dsp_orig, ARRAY_FLAG_NEWLINE_AT_TAIL);
}
void
Read_Input_Files(int argc, const char *argv[], int round) {
int n;
Init_Text(argc);
Init_Token_Array();
/* Assume all texts to be new */
Number_Of_New_Texts = Number_Of_Texts;
/* Read the files */
for (n = 0; n < Number_Of_Texts; n++) {
const char *fname = argv[n];
struct text *txt = &Text[n];
if (round == 1 && !is_set_option('T')) {
fprintf(Output_File, "File %s: ", fname);
}
txt->tx_fname = fname;
txt->tx_pos = 0;
txt->tx_start =
txt->tx_limit = Text_Length();
if (is_new_old_separator(fname)) {
if (round == 1 && !is_set_option('T')) {
fprintf(Output_File, "separator\n");
}
Number_Of_New_Texts = n;
}
else {
if (!Open_Text(First, txt)) {
if (round == 1 && !is_set_option('T')) {
fprintf(Output_File,
">>>> cannot open <<<< ");
}
/* the file has still been opened
with a null file for uniformity
*/
}
while (Next_Text_Token_Obtained(First)) {
if (!Token_EQ(lex_token, End_Of_Line)) {
Store_Token(lex_token);
}
}
Close_Text(First, txt);
txt->tx_limit = Text_Length();
/* report */
if (round == 1 && !is_set_option('T')) {
fprint_count(Output_File,
txt->tx_limit - txt->tx_start,
token_name
);
fprintf(Output_File, ", ");
fprint_count(Output_File, lex_nl_cnt-1, "line");
if (lex_non_ascii_cnt) {
fprintf(Output_File, ", ");
fprint_count(Output_File,
lex_non_ascii_cnt,
"non-ASCII character"
);
}
fprintf(Output_File, "\n");
}
#ifdef DB_TEXT
db_print_text(txt);
#endif /* DB_TEXT */
}
fflush(Output_File);
}
/* report total */
if (round == 1 && !is_set_option('T')) {
fprintf(Output_File, "Total: ");
fprint_count(Output_File, Text_Length() - 1, token_name);
fprintf(Output_File, "\n\n");
fflush(Output_File);
}
}
void
Read_Input_Files(int argc, const char *argv[]) {
int n;
Init_Text(argc);
Init_Token_Array();
/* Initially assume all texts to be new */
Number_of_New_Texts = Number_of_Texts;
/* Read the files */
for (n = 0; n < Number_of_Texts; n++) {
const char *fname = argv[n];
struct text *txt = &Text[n];
if (!is_set_option('T')) {
fprintf(Output_File, "File %s: ", fname);
}
txt->tx_fname = fname;
txt->tx_pos = 0;
txt->tx_start = Token_Array_Length();
txt->tx_limit = Token_Array_Length();
if (is_new_old_separator(fname)) {
if (!is_set_option('T')) {
fprintf(Output_File, "new/old separator\n");
}
if (Number_of_New_Texts == Number_of_Texts) {
Number_of_New_Texts = n;
} else fatal("more than one new/old separator");
}
else {
int file_opened = 0;
if (Open_Text(First_Pass, txt)) {
file_opened = 1;
} else {
/* print a warning */
if (is_set_option('T')) {
/* the file name has not yet been
printed; print it now
*/
fprintf(Output_File, "File %s: ",
fname);
}
fprintf(Output_File,
">>>> cannot open <<<<\n");
/* the file has still been opened
with a null file for uniformity
*/
}
while (Next_Text_Token_Obtained()) {
if (!Token_EQ(lex_token, End_Of_Line)) {
Store_Token(lex_token);
}
}
Close_Text(First_Pass, txt);
txt->tx_limit = Token_Array_Length();
txt->tx_EOL_terminated =
Token_EQ(lex_token, End_Of_Line);
/* report */
if (file_opened && !is_set_option('T')) {
fprint_count(Output_File,
txt->tx_limit - txt->tx_start,
Token_Name
);
fprintf(Output_File, ", ");
fprint_count(Output_File,
lex_nl_cnt - 1 +
(!txt->tx_EOL_terminated ? 1 : 0),
"line"
);
if (!txt->tx_EOL_terminated) {
fprintf(Output_File,
" (not NL-terminated)");
}
if (lex_non_ascii_cnt) {
fprintf(Output_File, ", ");
fprint_count(Output_File,
lex_non_ascii_cnt,
"non-ASCII character"
);
}
fprintf(Output_File, "\n");
}
#ifdef DB_TEXT
db_print_text(txt);
#endif /* DB_TEXT */
}
fflush(Output_File);
}
/* report total */
int sep_present = (Number_of_Texts != Number_of_New_Texts);
fprintf(Output_File, "Total input: ");
fprint_count(Output_File,
(!sep_present ? Number_of_Texts : Number_of_Texts - 1),
"file"
);
fprintf(Output_File, " (%d new, %d old), ",
Number_of_New_Texts,
(!sep_present ? 0 : Number_of_Texts - Number_of_New_Texts - 1)
);
fprint_count(Output_File, Token_Array_Length() - 1, Token_Name);
fprintf(Output_File, "\n\n");
fflush(Output_File);
}
