History
Joe Steffen began writing cscope in the early 1980s, as an aid to his programming work on a PDP-11. The tool became very popular within Bell Labs, as fellow employees requested more features and improvements, and was used on some large projects with as many as a million lines of source code. [2]
The tool was later made part of the AT&T Unix distribution. That distribution then went through several ownership changes: to Unix System Laboratories, then Novell, then The Santa Cruz Operation (SCO). [2]
Cscope itself is implemented in C. [2] The text interface is built on top of curses. [3] Cscope is well-integrated with vim and Emacs. [3]
In April 2000, SCO announced that it was contributing cscope to open source under a BSD license, while still maintaining it. The move came as SCO was staging several initiatives in the open source and Linux space, [4] even as it was competition from those spaces that would soon lead to SCO's demise. [5]
Petr Sorfa, a software engineer at SCO, became the maintainer of cscope and did the work for the first open source release of it, including porting it to the GNU gcc environment. [6] Sorfa subsequently continued as a maintainer of it, through an association with SourceForge. [7]
Cscope is still used by developers today, some of whom are most accustomed to using vi or other text-based editors, instead of GUI editors. The functions in cscope are available to varying degrees in modern graphical text editors.
Cscope's maintainers later became Hans-Bernhard Bröker and Neil Horman. [8]
The Linux kernel's top-level makefile has a target for building for cscope. [9] Indeed, an O'Reilly Media book about Linux networking published in 2006 recommends use of cscope for browsing the Linux kernel source code. [10] Similarly, a Packt Publishing book about Linux kernel programming published in 2024 has multiple mentions of using cscope in connection with understanding aspects of the kernel source code. [9]
Analysts have also found cscope useful for software audit reviews. According to a 2007 book published by John Wiley & Sons, "Cscope can be invaluable for auditing and is used by many security researchers on a regular basis." [3]
Usage
Cscope is used in two phases. First, a developer builds the cscope database of the source code. The developer can often use find or other Unix tools to get the list of filenames needed to index into a file called cscope.files. The developer then builds a database using the command cscope -b -q -k. The k flag is intended to build a database for an operating system or C library source code. It will not look in /usr/include. Second, the developer can now search those files using the command cscope -d. An index must be rebuilt whenever changes are made to indexed files. [11]
The menus within the cscope user interface allows the developer to search for functions, variables, and preprocessor directives including macros. [12]
There are a number of other options to the cscope command that can tailor its behavior. [12] The cscope database file is known for sometimes becoming large and requiring sufficient disk space. [10]
As created, cscope was created to search content within C code, but it can also be used (with some caveats) for C++ and Java code. [8] Cscope can also process lex and yacc files. [12]
While cscope normally runs on Unix and Linux systems, it has also been built in a version for Microsoft Windows. [3]
This page is based on this
Wikipedia article Text is available under the
CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.
