i am probably wrong, cuz it has been a long time since i wrote metatrader code that reads and writes files, but i think that you should be checking if you are at the end of the file prior to reading the file?
shouldnt it be like this?
if (FileIsEnding(h)) break;
string tmp = FileReadString(h);
also, when i am writing text files that i intend to later parse, i might insert lines that clue me in as to where i am currently at in the file, like this:
[RECORD_BEGIN]
first line of my file/record goes here....
second line of my file/record goes here....
[RECORD_BEGIN]
third line of my file/record goes here....
fourth line of my file/record goes here....
and then erroneous blank lines are easy to ignore/filter out.....
Just a guess, as I can't test this (you haven't supplied all the code), but try adding the line of code shown below:
Inserted Code
int h = FileOpen(FileName, FILE_CSV|FILE_READ,'~');
if (h==0) Comment("File "+FileName+" not found.");
// First pass loads data from file into arrays dt, val (i.e. date/time, and value)
for (int c=0; !FileIsEnding(h) && c<9999; c++) {
string tmp = FileReadString(h);
if (FileIsEnding(h)) break;
if (StringLen(StringTrimRight(tmp)) < 1) continue; // <<<<==== ADD THIS LINE OF CODE
StrToStringArray(tmp,arr,",");
int tc = StrToInteger(StringSubstr(arr[0],0,10));
dt[c] = StrToTime(TimeToStr(tc,TIME_DATE|TI ME_MINUTES|TIME_SECONDS));
val[c] = StrToNumber(arr[1]);
it didn't quite work but it got me close enough to fiddle around until it did work. I ended up putting
string pagetrim = StringTrimRight(page);
into the write side of things, which turns out has a maximum string length that I was exceeding too. So once I reduced the string length slightly, the CSV files were then written pre-trimmed.
THANK YOU SO MUCH
I'm gradually becoming semi-competent as a coder almost entire due to your help.
