*
About:
m u l t i d i f f:
This pogramm with the name multidiff was developed for
analysing binary data files.
If you have to reengineer a unkown fileformat you have to look
for differences in the file data and then you can understand what
is happening there. Of course is it a difficult way to check the
differences. An other problem was that the normal diff programm
can not give you an output. It tells only that the binary data files
differ. There are ways to do that but you have to do this and that
before there is what you need.
And you often need a html output to share the information with other
developer.
q m u l t i d i f f:
Because there are so many options for multidiff I have written
a qt frontend - called qmultidiff. It is only used to select
the used options. You can save the options in a xml file.
qmultidiff is part of multidiff.
If you do not have qt on your machine qmultidiff will not be compiled but
multidiff will.
*
Features:
Output modes
- output can be a html or a text file.
- output can be done into a file or to standard output.
Visu row modes
- you can diff the files in different visualisations rows:
- ascii
- hexadecimal
- decimal
- octal
- bit (each bit is compared)
- you can see the position on a visualisations file-position-row.
- you can change the order of the different visualisitions rows.
Information and statistics
- you can add a ascii file with hexadecimal value and the explaniation
of that as strings. So you get the information visualisation. The lenght of the strings is automatic adapted to the length of the values of the
hexadecimal values.
row.
- you can get an statistics output table for the hex-string.
informations
Diff
- you can diff with more than 2 files.
- the different values between to files are colored.
- you can start a automatic diff and the line will be moved
in right place.
- you can give also manually diff positions and the values
will be moved to that place.
- but some problems are still there.
List of tested different diffing algorithms
1. Analyses of the diff work with 4 files
2. Analyses of the diff work with 4 files
3. Analyses of the diff work with 4 files with other algorithm
Else
- you can hide identical start or stop ranges of the files.
- you can give manualy the start und stop position to look on.
- you can tell how much spaces should be between 2 values.
- you can change the colors, the fonts, the look very much.
- you can add a .multidiff initialisation file in your home
directory where you can put all the options you are using by your own default. The ini file is using the same options strings like that you did use on the command line
*
Screenshots:
1. A simple example (2 files)
2. A simple block example (2 files)
3. Hide equal start values (2 files)
4. The Rows: ascii bit hexadecimal decimal (2 files)
5. The info mode (2 files)
6. The block mode (with and without linking) (2 files)
7. Different ascii and row posibilties (2 files)
8. The row ordering plus start and stop position (2 files)
9. The moving diff mode (2 files)
10. The moving diff mode (more then 2 files)
Is not really ready jet!!
11. The info statistik (simple and detailed)
Detailed is not ready jet!!
12. Filelist and block list with and without links
Image will comeing soon!
13. Vertical mode
Is not ready jet!!
14. Output into multiple html pages
Is not ready jet!!
*
Initialisation file example:
-wsasc p
-wsinf '#'
-col "FF0000 FF00FF 00FFFF FFFF00 00FF00 "
*
Infofile file example:
One line with the hexadecimal values. A -1 is used to indicate the
end of the hexadecimal values line.
You can also use a -9 instead of a hexadecimal value for one unknown
value.
01 ff 02 04 03 02 00 02 -1
Header_part_1
21 05 00 04 01 ff 03 ff 00 -1
Header_part_4a
10 00 00 02 19 -1
Header_part_6
-9 -9 01 18 -9 08 1a -9 -1
cm_or_zoll
c8 00 01 18 04 08 1A 01 -1
Header_part_6b_(cm)
*
Options:
usage: ./multidiff [Options] <file1> <file2> ..
Options:
--------
* = default
-help = Prints this full list with all options.
-v = Prints the actual version of multidiff and exit.
-c x = Number of data values in one line
(default: the whole file in one line)
-ufna = * Use file names for the diff table
-ufnu = Use file numbers for the diff table
The file numbers option is only useful together
with the file list option (-fl)
-o x = Give the output filename.
(default: for html-mode: <result.html>
for txt-mode : <result.txt> )
Modes:
- - - -
-vm = Vertical display mode (not useful yet!!)
-hm = * Horizontal display mode
-ono = Write no output
-ofil = * Write (html/txt) output info file
-ostd = Write (html/txt) output info stdout
-ohtm = * Write as html output
-otxt = Write as text output without frames
-otxt2 = Write as text output with frames
Rows:
- - -
-ahl = * ASCII html output for the first file
-nahl = No ASCII html output for the first file
-abl = ASCII html output at the other files for diffing
-nabl = * No ASC html output at the other files for diffing
-d = Decimal html output
-nd = * No decimal html output
-7 = Octal html output
-n7 = * No octal html output
-h = * Hexadecimal html output
-nh = No hexadecimal html output
-b = Bits html output
-nb = * No bits html output
-p = * File position html output
-np = No file position html output
-fprd = * File position row as decimal values
-fprh = File position row as hexadecimal values
-m = Memory position html row.
-nm = * No memory position html row.
The option (-m) is very useful in the diff mode
(-mh and -md)
-inf x = Set a file as input for regex values
This is used for the info rows in the first file.
So you can get an explanation what the hex-code
means.
x = Name of the file with regex infos
-ninf = * Do not set a file as input for regex values
-infa x = Set a file as input for regex values
This is used for the info rows at the all files
(first file and the other files).
So you can get an explanition what the hex-code
means.
x = Name of the file with regex infos
(default: Not activated)
-mprd = * Memory position row as decimal values
-mprh = Memory position row as hexadecimal values
-prv = Vertical pos row mode: each char of
a pos value will be displayed one
below the other.
(default: Not activated)
-cr = * Cut row html output
-ncr = No Cut row html output
-pref = Position row in each file
(default: Not activated)
-ordh x = Order of the html rows (only the header)
(default: -ordh piadohb)
-ordb x = Order of the html rows (only the body)
(default: -ordb piadohb)
-ooci = Length of the information text (row/column) is
only 1 character. (hex) -> (h)
(default: all characters)
Order Information:
- p for pos (this works only with the md-mode)
- i for info
- a for asc
- d for dec
- h for hex
- b for bit
- o for oct
Example : -ord adhibp
Colors:
- - - -
-crcol x = Background color of the cut row
(default: <000000> )
-mmcol x = Background color of the global memory row
(default: -mmcol DDDDDD )
-fpcol x = Background color of the global file position row
(default: -fpcol DDDDDD )
-dfcol x = Color of the different expression
Only if the first value is existing
(default: -dfcol FF0000 )
-dfcol2 x= Color of the different expression
Only if the first value is not existing
Only in -md mode
(default: -dfcol2 FF8800 )
-bccol x = Background color of the block counter column/row
(default: -bccol DDDDDD )
-bgcol x = Background color of the webpage background
(default: -bgcol FFFFFF )
-brcol x = Background color of the reg text elements
(default: -brcol F0F020 )
-col "x" = Background colors of the files (1 or more)
example: -col "FFFFFF DDDDDD EEEEEE 543452"
(default: -col "FFDDDD DDFFDD DDDDFF" )
Start - stop:
- - - - - - -
-nequs = Hide an identical start range
-equs = * Do not hide an identical start range
The option -nequs is not allowed with -s
-neque = Hide an identical end range
-eque = * Do not hide an identical end range
The option -neque is not allowed with -e
-s xx = Set the start byte number to display
This option is not allowed with -nequs !!!
(default: 0)
-e xx = Set the end byte number to display
This option is not allowed with -neque !!!
(default: last byte of the biggest file)
Text output:
- - - - - - -
-txtwc = Use colored diff text in text mode
-ntxtwc = * Do not use colored diff text in text mode
-hf = Generate from each input file one output file
with the binary input data as ascii hex data
and with only one value per line
(default: Not activated)
Statistics table:
- - - - - - - - -
-st = Show statistics on html page
(default: Not activated)
-stpt = * Statistics position on top of html page
-stpb = Statistics position on bottom of html page
-stwl = Show statistics with links
(default: Not activated)
-stfna = * Statistics: Use file names
-stfnu = Statistics: Use file number
Use file number is only useful together with the
file list option -fl
File table:
- - - - - -
-fl = Print a list of all files first
-nfl = * Do not print a list of all files first
Block counting:
- - - - - - - -
-bc = Block counting
-nbc = * No block counting
-bcl = Block counting linking info on page
-nbcl = * No block counting linking info on page
Diff:
- - -
-md = Move diff visualize option
-pref is automatically set
-nmd = * No move diff visualize option
(default: -nmd)
-mh x y z= Move diff visualize option by hand
x=filenumber y=filepostion z=movewidth
This option can be added multiple times
(default: Not activated)
-pmlmh = Print the move elements like move hand elements
(default: Not activated)
-pmeof x = Print the move elements in the file x
(default: Not activated)
Width:
- - - -
-trsm = Try smallest value width (0)
If a value is wider it will always use the
widest width!
(default: 3)
-ec x = Number of whitespaces between two values
(default: 1)
Look:
- - -
-wsasc x = Whitespace for a not showable character at the
ascii row.
(default: < >)
-wsinf x = Whitespace for a unknown value at the info row.
(default: < >)
-fsfn x = Font size of the filenames. (-5 ... +5)
(default: 0)
-fstd x = Font size of the type description. (-5 ... +5)
(default: 0)
-fsda x = Font size of the data range. (-5 ... +5)
(default: -1)
-tbcol x = Table border color
(default: DDDDDD)
-tb x = Table border width
(default: 2)
-tcs x = Table cell spacing
(default: 0)
-tp x = Table padding
(default: 5)
-pt x = Page title: Title of the window.
(default: empty)
-pa x = Page author .
(default: empty)
-pe x = Page author - email address.
(default: empty)
-ph x = Page title: Title of the page.
(default: empty)
Debug:
- - - -
-deb x = Print some debug information
It's a bit code:
* - 0 = don't print debug infos
- 1 = print only reg-tag debug infos
- 2 = print only detailed diff debug infos
- 4 = print only statistics debug infos
- 8 = print only general diff debug infos
- 16 = print only memory alloc debug infos
- 32 = print detailed memory alloc debug infos
example : -deb 3
-sizes = Print the sizes of all structures and exit.
*
Download:
22. aug 2002:
multidiff-0.1.0.tar.gz
17. feb 2002:
hex_diff-0.0.98.tar.gz
10. feb 2002:
hex_diff-0.0.97.tar.gz
22. jan 2002:
hex_diff-0.0.95.tar.gz
11. jan 2002:
hex_diff-0.0.9.tar.gz
configure
make
make install
*
Mail:
Feel free to send comments/suggestions/patches !
ennobartels@t-online.de
*
Todo:
new:
-----
- free at moveItems segfault ??
- change the moveItem structure to moveItem
- bug: -s-e option real block counter !!!
- text mode: color for diff: -tdcol 31 -tdcol2 33
- rebuild the move diff function. with diff3 or anything else ????
- bcl should be possible without bc and should be moved to -wl (with links)
this is for -fl and -bc if on is -wl then okay
- add -ms and -me for memory start and stop position
Output:
-------
- output into different files
- only possible in html mode
- -ohf0 output html file mode in one file (default)
-ohf1 output html file mode in 3 files: index, statistic, diff
-ohf2 output html file mode in x files: index, statistic, diff, each block
->fpout[] as array dito name of the output files names
- onb x number of blocks per html page
PROG_TEST listing
-----------------
- add check_moveHand, sort_moveHand, swap_moveHand, move_handProceed,
- add moveItem_auto
- add print_size
- add row_head, row_foot,
- add re_mem, ma_mem
- add get_diffItems
- add moveHand -> moveItems
Rows:
-----
- short int row:
- [n]rsi0 -> floating point row output (short int = 2Bytes) with offset 0
- [n]rsi1 -> floating point row output (short int = 2Bytes) with offset 1
- float row:
- [n]rf0 -> floating point row output (float = 4Bytes) with offset 0
- [n]rf1 -> floating point row output (float = 4Bytes) with offset 1
- [n]rf2 -> floating point row output (float = 4Bytes) with offset 2
- [n]rf3 -> floating point row output (float = 4Bytes) with offset 3
- double row:
- [n]rd0 -> floating point row output (float = 8Bytes) with offset 0
- [n]rd1 -> floating point row output (float = 8Bytes) with offset 1
- [n]rd2 -> floating point row output (float = 8Bytes) with offset 2
- [n]rd3 -> floating point row output (float = 8Bytes) with offset 3
- [n]rd4 -> floating point row output (float = 8Bytes) with offset 4
- [n]rd5 -> floating point row output (float = 8Bytes) with offset 5
- [n]rd6 -> floating point row output (float = 8Bytes) with offset 6
- [n]rd7 -> floating point row output (float = 8Bytes) with offset 7
- Visualisation
with %08.4f
of with: o's:0|1 number of digits:x number of digit behind the point:y
- add a formel for example for zooming
ini-file:
---------
- add error message for seach option
0 -> error from inifile
1 -> error from commandline
- ignore ini file when the -ii comes on the commandline!
- add the slash for ini file options: -wsinf \" -wsinf ' ' -wsinf " "
- check again: del_comments, read_iniFile, get_noElements: reduce code
Screenshots and desciption images:
----------------------------------
- describe with images : -bc -bcl -fl whats happening then ?
Bugs:
-----
- Is there a netscape bug at 1000 characters in one row with the table lines
- xterm bug ??Problem with -otxt and -txtwc if linelenght to long -> linebreak
- maybe: Start and stop at the right place use an array with the number of
files start[] stop[]
Statistic:
----------
- with detail statistic not only counting them: make a list with each pos in
the file
- maybe with a link
a name="reg01_file01_01" (no of the reg, no of the file, no of the existing
- move the reg info loading out of stat to something before
stat_html sould only show what WAS fould
Color:
------
- option for the info row text background color
- pos-color with a little grew at md mode: -g, -ng, -gf faktor
get maxcolor from r, g or b and add grew then .. divide throught maxcolor
- Is there any nice filling char for the empty spaces at the diff mode
Diff md:
----------
- Diff md: Change info row from %s to %c for the -c mode and tag splitting
- Diff md: Use another algorithmen, maybe switching by the user
- Diff md: Color of b, d mode if the value is not existing
- diff 2 files then diff a new file with the 2 existing and look at items
try that in therory
Nice to have:
----------
- vertical display mode -vm
ini.mode = 1 = vertical 2= horizontal
Example 1:
File 1 | File 2
--------------+----------
8a 10101111 | 7b 11001010
Example 2: (-v2)
File 1 | File 2
--------------+----------
8 1 | 7 1
a 0 | b 1
1 | 0
0 | 0
1 | 1
1 | 0
1 | 1
1 | 0
for (a=0; a<8; a++) pos_col (a, ...)
hex_col (a, ...)
{
if (a>1) return
}
bin_col (a, ...)
{
if (a>7) return
}
- Filename "/bla/bla/mwtext.drw" to "/bla/....mwtext.drw"
maybe: files with identical pathes could be shorted and writting only
once -pfsh
- Filename list bevor start of prog
- Filenames each time of row use -fner -nfner
- add a ini file in commandline ? -ini xxx
- commandline option : do Not load the Ini File in your Home .multi_diff -nifh
Rest:
-----
- If a regtag is on the end of the line: right splitting
- Different visualisations for the diffing (bold, italiq)
- Option -searchd "" give position and counted element as static
with links to the places inside the file
- show only diff places. Hide everything else ?? -shod
- Flexibel eof char -ceof
- Position row:
add a startvalue. With this the values is not able to get wider and wider
- Instead of -c xx use -cb for the number of blocks the programm should show
- check html code with validate web
Not usefull:
------------
- check dobble commandline expressions (no thats not usefull)
- add bgimage ??
|
-