1. Verify01.exw Al
Hello Al,
Your code in verify01 compares a path and file with a drive letter.
I am attempting to use your code in my windows format Compare.exw.
How can I replaced the drive letter with a path and file? I tried and I get:
Comparing root paths:
F:\euphoria\clean\clean.exw
F:\euphoria\clean\clean.exw
problem file: F:\euphoria\clean\clean.exwclean.exw
Just like that. No typo here.
I should get no errors as they are the same file.
Don Cole
2. Re: Verify01.exw Al
don cole wrote:
>
>
> Hello Al,
>
> Your code in verify01 compares a path and file with a drive letter.
> I am attempting to use your code in my windows format Compare.exw.
>
> How can I replaced the drive letter with a path and file? I tried and I get:
>
> Comparing root paths:
> F:\euphoria\clean\clean.exw
> F:\euphoria\clean\clean.exw
>
> problem file: F:\euphoria\clean\clean.exwclean.exw
>
> Just like that. No typo here.
>
> I should get no errors as they are the same file.
>
>
> Don Cole
Hi Don,
Here is a new Verify01.exw file. I'm going to call it Verify02.exw now.
You already have all the required include files.
I'll try posting it here but you'll have to watch out for line wraps
so you might have to fix a few lines. If this doesnt work i'll either
send you the file or upload to the archive.
Take care,
Al
E boa sorte com sua programacao Euphoria!
My bumper sticker: "I brake for LED's"
From "Black Knight":
"I can live with losing the good fight,
but i can not live without fighting it".
"Well on second thought, maybe not."
[eucode]
--Written by Al Getz 04/2007, updated 07/2007.
--Use this when the two root directory names are different, such
--as:
--"C:\\download\\" and
--"C:\\Euphoria\\"
--The second path above can have a different drive letter too.
--This compares directories and file bytes of two directories
--on the same drive or different drives.
--Note this reports missing files and directories in the 'B' dir
--but will not report files that appear in the B dir that do not
--appear in the A dir. This is mainly to compare hard drive
--directories with CD drive directories after a CD write of
--those directories as in a backup, or to compare two directories
--on the hard drive when the second directory is a backup.
--This algorithm compares bytes counts before comparing each byte
--to save time when lots of files are different. It will not go
--through an entire 100MB file to find out the other file has 101MB,
--but will immediately report that the files are different.
--To use, enter (below) RootDirNameA and the CD drive letter in DriveLetterB
--and make RootPathB="", or simply enter the new entire path in RootPathB.
-- Note paths on CD drive must have the same names as those
-- on the hard drive, except for the drive letter, if the drive letter
-- method is used. If RootPathB is used, the root directories can
-- have any names.
--If you are using a program launcher of any kind be sure to specify the
-- working directory as well as the program path. The working directory
-- will be the same directory as this exw file is located in.
--Possible Report.txt error messages:
-- (note the A dir is the main directory, the B dir is the backup or copy, and
-- the A file resides under the A dir while the B file resides under the B
dir)
-- 1. "Could not find file" (the file in the A dir could not be found in the B
dir)
-- 2. "Bytes different in file" (A and B files have same number of bytes, but
at least
-- one byte is different)
-- 3. "Bytes count different in file" (one file has more or less bytes than
the other)
-- 4. "Could not find dir" (one complete directory could not be found in the B
dir)
-- The following error messages can only occur if one or more files have been
-- modified *AFTER* this program has been started. Of course this practice is
-- not recommended.
-- 5. "Could not open" (could not find the file in the directory)
-- 6. "More bytes in file" (more bytes in file B than in file A)
-- 7. "Not enough bytes to read in" (less bytes in file B than in file A)
-- Note that a totally blank Report.txt file (after program finishes) means
-- no changed files or directories have been found.
--Basic program operation:
--Make a list of all the directories in both root paths, then go through the A
dir
--directories (and files) one by one looking for directories in the B dir. Once
each
--dir is found, go through each file one by one comparing bytes count and if
necessary
--comparing each and every byte for equality. Files in the root directories are
--compared also.
--If any inconsistancies are found they are printed to the file "Report.txt"
--that resides in the same directory as this exw file. The types of problems
--that can come up are listed above (1 through 7).
include misc.e
include sort.e
include GetDirsB.e
object AllDirNamesA,AllDirNamesB
integer ndirs
sequence RootDirNameA,RootDirNameB,RootPathB
integer DriveLetterB
atom fnReport
constant ReportFileName="Report.txt"
--program makes a report about missing or changed files
constant BackupFileListName="BackupList.txt"
--program makes a backup file path list
procedure printf_report(sequence formatstring, sequence info)
--prints the report file and the backup list file
atom fn
integer n
sequence name1,name2,dirnameA
n=info[1]
if n=1 then
name1=info[2]
name2=info[3]
else
name1=info[3]
name2=info[2]
end if
printf(1,"\nproblem file: %s\n",{name1})
fn=open(ReportFileName,"a")
printf(fn,formatstring,{name1})
close(fn)
fn=open(BackupFileListName,"a")
name1=info[2]
name2=info[3]
if length(name2)=0 then
dirnameA=name1[length(RootDirNameA)+1..length(name1)]
name2=RootPathB&dirnameA
end if
printf(fn,"%s\n",{name1})
printf(fn,"%s\n\n",{name2})
close(fn)
return
end procedure
--////////////////////////////////////////////////////////////////////
-- USER ENTERS DATA HERE
--enter root dir on drive C, and drive letter for the CD drive,
--or use alternate method by entering both root dir names...
--the root directory typically on the hard drive:
RootDirNameA="C:\\Euphoria\\Projects\\VerifyCD\\CD1\\"
--the root directory drive letter to compare (typically the CD drive letter):
-- (note dirs must have same names, different drive letter when using this
method)
DriveLetterB='i'
--make this next sequence null "" to use drive letter above alone,
--or else specify the entire root path and ignore drive letter above.
--This path is typically on the CD or in the backup directory...
RootPathB="C:\\Euphoria\\Projects\\VerifyCD\\CD2\\"
--The above example would compare RootDirNameA to RootPathB, going through all
the
--sub directories as well as the root directory.
-- END OF USER ENTERED DATA SECTION
--////////////////////////////////////////////////////////////////////
--makes sure both root paths end with a backslash:
if RootDirNameA[length(RootDirNameA)]!='\\' then
RootDirNameA=RootDirNameA&'\\'
end if
if RootPathB[length(RootPathB)]!='\\' then
RootPathB=RootPathB&'\\'
end if
--compute the CD drive root dir: (only used with drive letter method)
RootDirNameB=RootDirNameA
RootDirNameB[1]=DriveLetterB
--use the second path if it is not null:
if length(RootPathB)>0 then
RootDirNameB=RootPathB
end if
printf(1,"RootDirNameA: %s\n",{RootDirNameA})
printf(1,"RootDirNameB: %s\n",{RootDirNameB})
atom uDriveLetterB
uDriveLetterB=upper(RootDirNameB[1])
printf(1,"Starting... %s\n",{""})
printf(1,"Comparing root paths: %s\n",{""})
printf(1," %s\n",{RootDirNameA})
printf(1," %s\n\n",{RootDirNameB})
AllDirNamesA=GetDirs(RootDirNameA)
AllDirNamesB=GetDirs(RootDirNameB)
ndirs=length(AllDirNamesA)
printf(1,"Total number of Dirs to verify: %d\n",{ndirs})
printf(1,"See 'Report.txt' for any problem files. %s\n\n",{""})
sleep(1)
fnReport=open(ReportFileName,"w")
close(fnReport)
fnReport=open(BackupFileListName,"w")
printf(fnReport,"%s\n\n",{" Backup File List (no files here means no files"&
" have changed,\n while a terminating backslash means the entire directory"&
"has to be backed up):"})
close(fnReport)
--look through all the dirs to see if any files are different
--in bytes...
sequence dirnameA,dirnameB,udirnameA,filenameA,ufilenameA,filenameB
sequence FileListA,FileListB,attribs
sequence uRootPathA,uRootPathB,udirnameB
atom fnA,fnB,binA,binB, BytesA,BytesB
integer founddir,problems,foundfile,nFiles
atom t,fps,minfps
uRootPathA=upper(RootDirNameA)
uRootPathB=upper(RootDirNameB)
problems=0
for k=1 to ndirs do
printf(1,"Doing Dir Number: %d of %d\n",{k,ndirs})
founddir=0
dirnameA=AllDirNamesA[k]
printf(1,"dirnameA: %s\n",{dirnameA})
udirnameA=upper(dirnameA)
udirnameA=udirnameA[length(uRootPathA)+1..length(udirnameA)]
for j=1 to length(AllDirNamesB) do
dirnameB=AllDirNamesB[j]
udirnameB=upper(dirnameB)
udirnameB=udirnameB[length(uRootPathB)+1..length(udirnameB)]
if equal(udirnameA,udirnameB) then
--found dirname, now do files
founddir=1
FileListA=dir(dirnameA)
FileListB=dir(dirnameB)
FileListA=sort(FileListA)
FileListB=sort(FileListB)
nFiles=0
t=time()
minfps=1e10
for h=1 to length(FileListA) do
if h/5=floor(h/5) then
t=time()-t
if t=0 then
fps=1000
else
fps=5/t
end if
if fps<minfps then
minfps=fps
end if
--print: #of files done, files per second, and min files per second
printf(1,"File: %d files/second: %3.2f minfps=%3.2f
",{h,fps,minfps})
puts(1,13)
t=time()
if h=5 then
minfps=1e10
end if
end if
attribs=FileListA[h][2]
if find('d',attribs) then
--skip dir names
else
nFiles+=1
filenameA=FileListA[h][1]
ufilenameA=upper(filenameA)
foundfile=0
for i=1 to length(FileListB) do
attribs=FileListB[i][2]
if find('d',attribs) then
--skip dir names
else
filenameB=FileListB[i][1]
if equal(ufilenameA,upper(filenameB)) then
foundfile=1
--compare bytes counts before checking all individual bytes:
BytesA=FileListA[h][D_SIZE]
BytesB=FileListB[i][D_SIZE]
if BytesA!=BytesB then
--the files are different simply because they have a different
number of bytes
fnA=0
fnB=0
printf_report("Bytes count different in:
%s\n",{2,dirnameA&filenameA,dirnameB&filenameB})
problems+=1
exit
end if
--if byte count is the same then compare all byte values...
fnA=open(dirnameA&filenameA,"rb")
fnB=open(dirnameB&filenameB,"rb")
if fnA<1 then
printf_report("Could not open:
%s\n",{1,dirnameA&filenameA,""})
problems+=1
exit
end if
if fnB<1 then
printf_report("Could not open:
%s\n",{2,"",dirnameB&filenameB})
problems+=1
exit
end if
if fnA>1 and fnB>1 then
--compare file bytes
while 1 do
binA=getc(fnA)
if binA<0 then
--done comparing this file with the backup
binB=getc(fnB)
if binB>0 then
printf_report("More bytes in file:
%s\n",{2,dirnameA&filenameA,dirnameB&filenameB})
problems+=1
end if
exit
else
binB=getc(fnB)
if binB<0 then
printf_report("Not enough bytes to read in:
%s\n",{2,dirnameA&filenameA,dirnameB&filenameB})
problems+=1
exit
else
if binA=binB then
--both bytes the same so continue
else
printf_report("Bytes different in file:
%s\n",{2,dirnameA&filenameA,dirnameB&filenameB})
problems+=1
exit
end if
end if
end if
end while --1
end if --fnA and fnB
if fnA>0 then
close(fnA)
end if
if fnB>0 then
close(fnB)
end if
end if
end if
end for
if not foundfile then
printf_report("Could not find file: %s\n",{1,dirnameA&filenameA,""})
problems+=1
end if
end if
end for --h
printf(1,"\nFiles checked: %d\n",nFiles)
exit
end if
end for
if not founddir then
printf_report("Could not find dir: %s\n",{1,dirnameA,""})
problems+=1
end if
end for
printf(1,"\nProblem files or directories found: %d\n",{problems})
-- print time in minutes:
printf(1,"Time to complete: %8.1f minutes.\n", time()/60)
printf(1,"\n%s\n",{"Close console to exit..."})
sleep(500000) --user will close out console to exit program
[/eucode]
3. Re: Verify01.exw Al
- Posted by Al Getz <Xaxo at aol?c?m>
Jul 19, 2007
-
Last edited Jul 20, 2007
Don,
Sorry i didnt understand your question quite right as i read it
too fast. I thought you were asking about how to enter two paths
that were different.
My program was made to compare whole *directories*, not individual
files. You enter two paths but the paths are directory names
such as:
c:\download\files\
At present, you cant point to an individual file unless you use a trick.
If you dont feel like updating it to do that, if
you really want to compare two files, then one thing
you can do is to copy the two files into two temp directories.
You can then point to those directories.
See, when i made this program i wanted it to be able to dig down
into sub directories to compare all the files, given only the
root directory names.
Take care,
Al
E boa sorte com sua programacao Euphoria!
My bumper sticker: "I brake for LED's"
From "Black Knight":
"I can live with losing the good fight,
but i can not live without fighting it".
"Well on second thought, maybe not."
4. Re: Verify01.exw Al
Al Getz wrote:
>
> Don,
>
> Sorry i didnt understand your question quite right as i read it
> too fast. I thought you were asking about how to enter two paths
> that were different.
>
> My program was made to compare whole *directories*, not individual
> files. You enter two paths but the paths are directory names
> such as:
> c:\download\files\
>
> At present, you cant point to an individual file unless you use a trick.
>
> If you dont feel like updating it to do that, if
> you really want to compare two files, then one thing
> you can do is to copy the two files into two temp directories.
> You can then point to those directories.
>
> See, when i made this program i wanted it to be able to dig down
> into sub directories to compare all the files, given only the
> root directory names.
>
>
> Al
>
> E boa sorte com sua programacao Euphoria!
>
>
> My bumper sticker: "I brake for LED's"
>
Hello Al,
I got that verify02 working with my program. It compares 2 directories and
reports the files that are different. It does not however tell you which line
numbers in the files are different. I think that's what you just said. I read
too fast sometimes also. I could send you what I have but I should email it. It
has several includes.
Don Cole
5. Re: Verify01.exw Al
don cole wrote:
>
> Al Getz wrote:
> >
> > Don,
> >
> > Sorry i didnt understand your question quite right as i read it
> > too fast. I thought you were asking about how to enter two paths
> > that were different.
> >
> > My program was made to compare whole *directories*, not individual
> > files. You enter two paths but the paths are directory names
> > such as:
> > c:\download\files\
> >
> > At present, you cant point to an individual file unless you use a trick.
> >
> > If you dont feel like updating it to do that, if
> > you really want to compare two files, then one thing
> > you can do is to copy the two files into two temp directories.
> > You can then point to those directories.
> >
> > See, when i made this program i wanted it to be able to dig down
> > into sub directories to compare all the files, given only the
> > root directory names.
> >
> >
> > Al
> >
> > E boa sorte com sua programacao Euphoria!
> >
> >
> > My bumper sticker: "I brake for LED's"
> >
>
> Hello Al,
>
> I got that verify02 working with my program. It compares 2 directories and
> reports the files that are different. It does not however tell you which line
> numbers in the files are different. I think that's what you just said. I read
> too fast sometimes also. I could send you what I have but I should email it.
> It has several includes.
>
>
> Don Cole
Hi Don,
Im glad you had some luck with it finally.
The report does not include any files line numbers because it does
a binary compare, not a 'text' or 'line by line' compare. I had
it do a binary compare because many of the files are not text files
(or euphoria source files) but are instead .exe files, .dll files,
etc., with some text files too.
I guess what you would want to see is the compare action taken
commensurate with the type of file? I guess that's not a bad idea
really. We'd have to also supply a list of file extensions and
how those files should be handled, but i guess that's not too much
work to do. On the other hand, the program would then also have to
work with different formats, like .doc and .rtf and be able to
handle unicode. Im not sure if i want to get into that myself
even though it could be done.
Maybe you want to be able to compare Euphoria source files too?
For now though it looks like it will compare in binary mode only,
however it would not take much to get it to report exactly which
byte was different, and then you could go back and open the files
and provide a dump of part of the two files. Even so, only the
first difference found would be reported as (at least currently)
the algo exits at the first difference found just after writing
to the report file.
Another thing i should mention (and this is written in the file
itself in a comment section) is that for two directories, call
them A and B (and the variable names reflect this convention with
their last letter being A and B respectively), the file names
(and directory names) not found in B are reported (B is the backup)
but the file names and dir names in B (if there are any) that are
not in A are *not* reported. This is because if for some reason
you feel like including an extra file in the backup for some reason
and you dont really need it on the hard drive, you dont have to
skip reading it in the report file. If this is a problem however,
it doesnt take much to have the 'missing' A files (and dir names)
reported in the report too.
BackupList.txt:
The other convention used is that file names end with no backslash,
while directory names *always* end with a backslash. This is so
another program can go through the list in the BackupList.txt file
and detect when it has to write an entire directory to disk rather
than a single file. Writing an entire directory means writing
the root files and every sub directory and all those files too,
whereas writing a single missing file is a simple 'copy file'.
Take care,
Al
E boa sorte com sua programacao Euphoria!
My bumper sticker: "I brake for LED's"
From "Black Knight":
"I can live with losing the good fight,
but i can not live without fighting it".
"Well on second thought, maybe not."
6. Re: Verify01.exw Al
Al Getz wrote:
>
> don cole wrote:
> >
> > Al Getz wrote:
> > >
> > > Don,
> > >
> > > Sorry i didnt understand your question quite right as i read it
> > > too fast. I thought you were asking about how to enter two paths
> > > that were different.
> > >
> > > My program was made to compare whole *directories*, not individual
> > > files. You enter two paths but the paths are directory names
> > > such as:
> > > c:\download\files\
> > >
> > > At present, you cant point to an individual file unless you use a trick.
> > >
> > > If you dont feel like updating it to do that, if
> > > you really want to compare two files, then one thing
> > > you can do is to copy the two files into two temp directories.
> > > You can then point to those directories.
> > >
> > > See, when i made this program i wanted it to be able to dig down
> > > into sub directories to compare all the files, given only the
> > > root directory names.
> > >
> > >
> > > Al
> > >
> > > E boa sorte com sua programacao Euphoria!
> > >
> > >
> > > My bumper sticker: "I brake for LED's"
> > >
> >
> > Hello Al,
> >
> > I got that verify02 working with my program. It compares 2 directories
> > and
> > reports the files that are different. It does not however tell you which
> > line
> > numbers in the files are different. I think that's what you just said. I
> > read
> > too fast sometimes also. I could send you what I have but I should email it.
> > It has several includes.
> >
> >
> > Don Cole
>
> Hi Don,
>
>
> Im glad you had some luck with it finally.
>
> The report does not include any files line numbers because it does
> a binary compare, not a 'text' or 'line by line' compare. I had
> it do a binary compare because many of the files are not text files
> (or euphoria source files) but are instead .exe files, .dll files,
> etc., with some text files too.
> I guess what you would want to see is the compare action taken
> commensurate with the type of file? I guess that's not a bad idea
> really. We'd have to also supply a list of file extensions and
> how those files should be handled, but i guess that's not too much
> work to do. On the other hand, the program would then also have to
> work with different formats, like .doc and .rtf and be able to
> handle unicode. Im not sure if i want to get into that myself
> even though it could be done.
> Maybe you want to be able to compare Euphoria source files too?
>
> For now though it looks like it will compare in binary mode only,
> however it would not take much to get it to report exactly which
> byte was different, and then you could go back and open the files
> and provide a dump of part of the two files. Even so, only the
> first difference found would be reported as (at least currently)
> the algo exits at the first difference found just after writing
> to the report file.
>
> Another thing i should mention (and this is written in the file
> itself in a comment section) is that for two directories, call
> them A and B (and the variable names reflect this convention with
> their last letter being A and B respectively), the file names
> (and directory names) not found in B are reported (B is the backup)
> but the file names and dir names in B (if there are any) that are
> not in A are *not* reported. This is because if for some reason
> you feel like including an extra file in the backup for some reason
> and you dont really need it on the hard drive, you dont have to
> skip reading it in the report file. If this is a problem however,
> it doesnt take much to have the 'missing' A files (and dir names)
> reported in the report too.
>
> BackupList.txt:
> The other convention used is that file names end with no backslash,
> while directory names *always* end with a backslash. This is so
> another program can go through the list in the BackupList.txt file
> and detect when it has to write an entire directory to disk rather
> than a single file. Writing an entire directory means writing
> the root files and every sub directory and all those files too,
> whereas writing a single missing file is a simple 'copy file'.
>
>
> Al
>
> E boa sorte com sua programacao Euphoria!
>
>
> My bumper sticker: "I brake for LED's"
>
Thanks Al,
Don Cole
