The diff command calls an element-type-specific program (the compare method for the type) to compare the contents of two or more file elements or two or more directory elements. Typically, the files are versions of the same file element.
You can also use this command to compare ordinary text files.
diff uses the type manager mechanism to determine how to compare the specified objects. For more information, see the type_manager reference page.
Each difference is reported as one or more pairwise differences. For example, if three contributors all differ from the base contributor in a particular section, diff lists the file1-file2 difference, followed by the file1-file3 difference, followed by the file1-file4 difference.
The default file-comparison report begins with a file summary, which lists all the input files and their assignments as file 1, file 2, and so on. If no differences are detected among the files, this listing is replaced by the message Files are identical.
The remainder of the report is a series of pairwise differences, each of which is preceded by a descriptive header line:
******************************** (file summary)
<<< file 1: util.c@@/main/1
>>> file 2: util.c@@/main/3
********************************
----------[after 15]------|-------[inserted 16]------ (header)
| char *s; (difference)
|-
---------[changed 18]-----|----[changed to 19-21]---- (header)
return ctime(&clock); | s = ctime(&clock); (difference)
- | s[ strlen(s)-1 ] = '\0';
| return s;
|-
The –quiet and –diff_format options suppress the file summary. The –headers_only option suppresses the differences, listing the header lines only.
Header Lines. Each header line indicates which text lines in the input files were changed and how they were changed. The words describe the change in terms of how the first file was changed to produce the second file. Header lines can have the following formats, where each of A, B, and so on may be a single line number (for example, 46) or a range (for example, 256–290):
Insertion of one or more lines. B indicates where the inserted lines occur in the second file. A indicates the corresponding point in the first file.
Deletion of one or more lines. C indicates which lines from the first file were deleted. D indicates the corresponding point in the second file.
Deletion of one or more lines from the first file, to which there corresponds an insertion of the same lines in the second file. Typically, this indicates that a range of lines was moved from one location to another; see inserted/moved below. C indicates where the lines were deleted from the first file; B indicates where these same lines were inserted in the second file. D indicates the point in the second file that corresponds to C.
One or more lines changed in place. X indicates which lines in the first file were changed. Y indicates where the replacement lines occur in the second file.
Insertion of one or more lines in the second file, to which there corresponds a deletion of the same lines from the second file. Typically, this indicates that a range of lines was moved from one location to another; see deleted/moved above. B indicates where the lines were inserted in the second file; C indicates where these same lines were deleted from the first file; A indicates the point in the first file that corresponds to B.
Differences. diff can report a difference in several ways. When comparing files, its default behavior is to list corresponding lines side by side, and possibly truncated.
A plus sign (+) at the end of a difference line indicates that it has been truncated in the report. To see more of such lines, you can increase the report widthby using the –columns option (or on UNIX, by using the –tiny option). The minus signs (–) along the vertical separator line indicate the endpoints of the groups of differing lines. They help to distinguish empty lines in the input files from blank space in command output.
The –serial_format option causes the differences to be reported as entire lines, in above-and-below format instead of side-by-side format. For example:
-----[after 15 inserted 16]-----
> char *s;
-----[18 changed to 19-21]-----
< return ctime(&clock);
---
> s = ctime(&clock);
> s[ strlen(s)-1 ] = ' ';
> return s;
The –diff_format option causes both the headers and differences to be reported in the style of the UNIX diff utility, writing a list of the changes necessary to convert the first file being compared into the second file, as follows:
When diff compares multiple files, it adds file-identification annotations to the diff-style headers.
The –graphical option displays differences graphically.
For a comparison of directory versions, diff uses a directory-element-specific compare method, whose report format is very similar to the one described in “Side-by-Side File Comparison Report Style”.
A version of a VOB directory can contain several types of entries:
Note: Multiple VOB hard links to the same element will have the same creator and creation time, but different names.
The diff report is a series of differences, each of which focuses on one directory entry. A difference can be a simple addition or deletion; it can also involve the renaming of an existing object or the reuse of an existing name for another object. The following examples illustrate the various possibilities.
An object named obj2 was added (mkelem, mkdir, or ln) in the second version of the directory.
An object named obj5 was removed (rmname) in the second version of the directory.
---------------[ renamed ]------------ |--------[ renamed to ]-------
obj3 12-Aug.14:00 akp | obj3.new 12-Aug.14:00 akp
An object named obj3 was renamed (mv) to obj3.new in the second version of the directory.
-------------[ old object ]------------ |--------[ new object ]-------
obj4 12-Aug.14:04 akp | obj4 19-Oct.17:10 akp
In the second version of the directory, an object named obj4 was removed (rmname) and another object was created with that same name.
----------[ old link text ]---------|------[ new link text ]----
doc -> ../vob1/doc 13-Aug.08:44 akp | doc -> ../vb/doc 19-Sep.21:01 akp
(Special case of the preceding example) In the second version of the directory, a VOB symbolic link named doc was removed (rmname) and another VOB symbolic link was created with that same name.
---------------[ renamed ]------------|---[ renamed to ]------------
obj4 12-Aug.14:01 akp | obj1 12-Aug.14:01 akp
---------------[ removed ]------------|-----------------------------
obj1 12-Aug.14:00 akp |-
These two differences show that in the second version of the directory, an object named obj1 was removed and another object was renamed from obj4 to obj1.
With the exception of –predecessor and –options, diff options are the same as those of cleardiff.
Note: When comparing files of type html on UNIX systems, if the machine on which you execute diff –graphical is not the machine on which you run your HTML browser, your browser may not be able to find the pathname to the files being compared.
An error occurs if the pname does not specify a version:
Use quotes if you are specifying more than one pass-through option; diff must see them as a single command-line argument. For example, this command passes through the –quiet and –blank_ignore options:
For descriptions of the valid options , see the cleardiff reference page.
The UNIX examples in this section are written for use in csh. If you use another shell, you may need to use different quoting and escaping conventions.
The Windows examples that include wildcards or quoting are written for use in cleartool interactive mode. If you use cleartool single-command mode, you may need to change the wildcards and quoting to make your command interpreter process the command appropriately.
In cleartool single-command mode, cmd-context represents the UNIX shell or Windows command interpreter prompt, followed by the cleartool command. In cleartool interactive mode, cmd-context represents the interactive cleartool prompt.
cleardiff, diffcr, ln, merge, mkdir, mkelem, mv, rmname, type_manager, xclearcase, xcleardiff
Copyright© 2003 Rational Software. All Rights Reserved.