static void
pass2_txt(struct text *txt) {
struct position *pos;
size_t old_nl_cnt;
if (!txt->tx_pos) /* no need to scan the file */
return;
/* Open_Text() initializes lex_nl_cnt and lex_tk_cnt */
if (!Open_Text(Second_Pass, txt)) {
fprintf(stderr, ">>>> File %s disappeared <<<<\n",
txt->tx_fname
);
return;
}
/* Sort the positions so they can be matched to the file; the linked
list of struct positions snakes through the struct positions in the
struct chunks in the struct runs.
*/
#ifdef DB_POS
db_print_pos_list("before sorting", txt);
#endif /* DB_POS */
sort_pos_list(&txt->tx_pos);
#ifdef DB_POS
db_print_pos_list("after sorting", txt);
#endif /* DB_POS */
#ifdef DB_NL_BUFF
db_print_nl_buff(txt->tx_nl_start, txt->tx_nl_limit);
#endif /* DB_NL_BUFF */
#ifdef DB_POS
fprintf(Debug_File, "\n**** DB_PRINT_SCAN of %s ****\n", txt->tx_fname);
#endif /* DB_POS */
old_nl_cnt = 1;
pos = txt->tx_pos;
while (pos) {
/* we scan the pos list and the file in parallel */
/* find the corresponding line */
while (pos->ps_tk_cnt >= lex_tk_cnt) {
/* pos does not refer to this line, try the next */
/* shift the administration */
old_nl_cnt = lex_nl_cnt;
/* and get the next eol position */
if (!Next_Text_EOL_Obtained()) {
/* reached end of file without obtaining EOL */
if (!txt->tx_EOL_terminated) {
/* that's OK then */
} else {
fprintf(stderr,
">>>> File %s modified <<<<\n",
txt->tx_fname
);
}
break;
}
#ifdef DB_POS
db_print_lex(txt->tx_fname);
#endif /* DB_POS */
}
/* fill in the pos */
switch (pos->ps_type) {
case 0: /* first token of run */
pos->ps_nl_cnt = old_nl_cnt;
break;
case 1: /* last token of run */
pos->ps_nl_cnt = lex_nl_cnt;
break;
}
/* and get the next pos */
pos = pos->ps_next;
}
#ifdef DB_POS
db_print_pos_list("after scanning", txt);
#endif /* DB_POS */
/* Flush the flex buffers; it's easier than using YY_BUFFER_STATE. */
while (Next_Text_EOL_Obtained());
Close_Text(Second_Pass, txt);
}
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);
}
