Midterm I Flashcards

Question

What are the 7 types of regular files?

Answer 1

1. Regular files 2. Directory files 3. Buffered special files (block devices) 4. Unbuffered special files (character devices) 5. Symbolic links 6. Pipes (named pipes) 7. Sockets

Answer 2

Regular files: are files containing text, executable programs, graphics, video, and similar. These are the files we usually think of by default under the name ‘files’. In a long ls output (ls -l), these files are indicated with the character ‘-’.

Answer 3

The directories that store groups of other files are also a type of files called directory files. In a long ls output, these files are indicated with the character ‘d’.

Answer 4

are also called block devices. These files represent disk drives, USB keys and similar in a direct access. They are called buffered because a buffer in memory is used in reading and writing them. In a long ls output, these file are indicated with the character ‘b’.

Answer 5

are also called character devices, and represent devices such as terminal and keyboard. Both, buffered and unbuffered special files are called device files. In a long ls output, these files are indicated with the character ‘c’.

Answer 6

are references to other files. Using them, we can access existing files through a different pathname (path). They are similar to shortcuts in Windows. In a long ls output, these files are indicated with the character ‘l’.

Answer 7

are used for inter-process communication, and they are also called named pipes. This file type should not be confused with pipes created between processes in a compound shell command. In a long ls output, these files are indicated with the character ‘p’.

Answer 8

are used for inter-process communication, similarly to pipes. One of the differences is that sockes are fully duplex-capable; i.e., processes can send data both ways, unlike pipes. Beside data, processes can also send file descriptors using sockets. In a long ls output, the socket files are indicated with the character ‘s’.

Answer 9

Each file has a name – Files can have the same name if they are in different directories – Example: see bin in the previous figure – To distinguish files with the same name, we use pathnames – Pathname (or path) is a sequence of directories, finishing with a file name – Directories are separated using character slash (/) – Example: /users/faculty/vlado/csci2132/lab1/HelloWorld.java

Answer 10

– Absolute path starts from root (initial slash /), examples: /usr/bin /users/faculty/vlado/csci2132/lab1/HelloWorld.java

Answer 11

``` Relative path starts from the current directory; examples (if the current directory is ‘vlado’): csci2132 csci2132/lab1/HelloWorld.java ./csci2132/lab1/HelloWorld.java ../../visitor ```

Answer 12

basename /home/ed/file.txt basename /home/ed/file.txt .txt dirname /home/ed/file.txt

Answer 13

– ls paths — list directory contents – pwd — print working directory – cd path — change directory – mkdir dirs — make directory(ies) – mkdir -p paths – rmdir dirs — remove empty directory(ies) – mv path1 path2 — move or rename directory or file – mv -i path1 path2 – rm paths — remove files but can remove directories with option -r; useful to consider -f and -i

Answer 14

– cat files — showing textual file(s) content – more files — showing textual file content, paged – head files — showing textual file content, first part – tail files — showing textual file content, last part – vi, emacs, pico, nano — file editors – wc files — word count

Answer 15

– Used to protect files and processes between different users – Every user has a unique username, which is a text string – Try command: whoami – The system uses numeric userid, which we will call just userID (username is for string id) – Try command: id -u

Answer 16

Every UNIX user is a member of a group – A user can be member of multiple groups, but one is effective for a process – Each group has a unique groupname and groupID – Command to list groups user is member of: groups – Command for more complete information: id – Each process, including shell, has one effective userID and groupID – Each file is owned by one user and one group: file owner and file group

Answer 17

user group other

Answer 18

– Command: ls -l – Note: a few more useful ls options: -a -t -r – Example: $ echo test > tmpfile.txt $ ls -l tmpfile.txt -rw-r--r-- 1 vlado csfac 5 Sep 13 11:21 file.txt

Answer 19

javac HelloWorld.java

Answer 20

java HelloWorld

Answer 21

The command chmod is used to change file permissions, or in other words, the file mode bits. Its syntax is: chmod mode files where mode specifies the new mode-bit pattern for the files given in the argument files. One way is to give the pattern explicitly, as in: chmod 664 file.txt where the number 664 represents in the octal base the new bit-mode pattern for the file file.txt. If we translate the octal number 664 in binary, it is 110 110 100, so the new permission for the file are rw-rw-r--, or read, write, and not execute for user, read, write, and not execute for group, and read, not write, and not execute for the others

Answer 22

``` – chown newuser file.txt – chown -R newuser files dirs – chgrp newgroup file.txt – chgrp -R newgroup files dirs -R is used for directory recursive change ```

Answer 23

``` – newgrp newgroup – changes into newgroup (logs into new group) – su newuser – changes effective user – needs to be superuser (root user) ```

Answer 24

– Remember what we learned about: stdin, stdout, stderr – Redirecting the standard output of a program into a file: command > filename – Creates a file (filename) if it does not exist – Example: ls lab1 > listing – Important: ‘>’ redirection deletes previous file contents – To append a file with new content use ‘>>’ – Example: ls lab1 >> listing – Creates a file (‘listing’) if it does not exist, as well

Answer 25

Redirects the standard input from a file into a process – Useful in testing – Syntax: command < filename – Example: sort < names.txt – sorts names in a file names.txt and prints out – Example 2: sort < names.txt > names-sorted.txt – Example 3: mail csusername < HelloWorld.java – Example 4: mail full@email < HelloWorld.java

Answer 26

``` – Standard error output is not redirected by > – Syntax (bash specific): command 2> filename – Cannot be a space between ‘2’ and ‘>’ – Example: rm x 2> error – If file x does not exist, it will produce an error message – >> can be used to append output: command 2>> filename ```

Answer 27

– Used to specify multiple filenames – Makes use of “wildcards”; i.e., metacharacters expanded by the shell – Some wildcard types: – ?: matches any single character – *: matches any string, including empty string – [...]: matches any single character in the set – [!...]: any character except characters from the set

Answer 28

– [0-9]: any digit between 0 and 9 – [a-zA-Z]: any English alphabet character – [unix]: matches either ‘u’, ‘n’, ‘i’, or ‘x’ – ls ˜/csci2132/lab1/*.java — list Java files – ls *.???? — list all files with 4-character extension – ls lab[1-9] — list all files with the name consisting of word lab and a digit from 1 to 9 – ls [!0-9]* — list all files which name does not start with a digit – cp lab1.bk/*.java lab1/ — copy Java files from one directory to another

Answer 29

Every file has a unique inode number. A directory is a file containing a table of pairs (inode number, filename), where each pair corresponds to a file in the directory. There is one system-wide inode table, and inode numbers are indices into it. In this way, the execution of commands such as ls -l that retrieve information regarding files in a directory can be efficient. This is a nice model that is effective and logical—a good example to follow when we design complex systems.

Answer 30

To reveal file inode numbers, we can use the option -i in command ls. For example, the command: ls -lid tmp may give an output like: 84492732 drwx------ 2 ... tmp where the first number (84492732) is the inode number of the directory tmp, which is a file. The option -d is used in the command ls to avoid printing the contents of the directory tmp.

Answer 31

– Programs cannot tell difference between ‘original’ and ‘target’ – Deleting a file just removes a link (unlink system call) – Only when number of links is 0, the space is freed – Limitations: – same file system – no directories (exceptions possible on some systems) A file is deleted when no hard links to its inode exist. In the above example, if we remove ˜/csci2132/lab1/HelloWorld.java, we can still use ˜/csci2132/HelloWorld.java to access the file. To check the hard link count of a given file, use the ls -l command

Answer 32

There are some restrictions about creating hard links. First, hard links cannot span different disk drives. Second, directories cannot have hard links, as this would generally make directory structure not a tree any more, and this could possibly create an infinite recursion for the programs that traverse a directory structure. Additionally, it would be ambiguous which would be a parent directory (..) of a hard-linked directory. Other than these, there are no really design issues of why hard links on directories would not be allowed, so some systems may allow them in exceptional situations (e.g., some newer Mac OS systems in the Time Machine application).

Answer 33

The following command can be used to create a hard link: | ln target linkname

Answer 34

A soft link is a file that has the absolute or relative pathname of another file as its data part. When we create a soft link, we create a new inode. This is equivalent to a file ‘shortcut’ in Windows. Strictly speaking, a soft link is not a link to a file, it is a link to a pathname. The pathname in a soft link could not exist any more, and the link will still be there, which could give us a wrong impression that the actual file exists, while it does not.

Answer 35

ln -s target linkname

Answer 36

Some advantages of hard links are: They are more efficient to use: faster to access and they do not require additional inodes. We do not need to worry about which file is the ‘master’ file in order not to delete it by mistake. Hard links are particularly useful in some situations, such as keeping incremental backups of a file directory structure. Some programs do not treat symbolic links as ‘real’ files, which may or may not be desirable. For example, in a recursive copy, the cp command will copy symbolic links to regular files as symbolic links.

Answer 37

Answer 38

Regular expressions are used to address the problem of supporting fast and flexible text search: regular expression search can be done using a DFA (you will learn this in a 3rd year course) which is fast, and they can be used to specify a range of patterns and are thus flexible. A regular expression is a sequence of normal and special characters specifying a pattern to match against strings. It is used in many programs such as grep, sed, and perl. Again, we need learn metacharacters that can be used to specify matching rules. Pay attention to the difference between these characters and wildcards in filename substitution.

Answer 39

– .: Matches any single character. For example, some strings that match a.b are: acb, a b, a-b, ... – []: Matches any of the single characters enclosed in brackets. We can use - to specify a range, and ˆ to compliment the set (e.g. [ˆ ]). Other special characters, including ., *, ˆ (note that ˆ has other special meanings outside []), $ and \ lose their special meanings inside []. – *: 0 or more occurrences of the character that precedes it. For example, a* can be used to specify the following set of strings: empty string, a, aa, aaa, aaaa, ... – ˆ: Matches the beginning of a line. – $: Matches the end of a line. ˆ and $ are useful if we want a pattern to occur as the prefix/suffix of a line. – \: Inhibits the meaning of any metacharacters.

Answer 40

A filter is a program that gets most of its data from stdin and writes its main results to stdout. As pipes use one program’s stdout as another’s stdin, it is natural that filters are often used as elements of pipelines.

Answer 41

– grep reads a file or stdin and outputs lines matching a regular expression – grep syntax grep [options] BRE [file(s)] If we enter the following command grep ’\$[0-9][0-9]*\.[0-9][0-9]’ price The following command will find all the dictionary words that start with a or b and end with b: grep ’ˆ[ab].*b$’ /usr/share/dict/linux.words

Answer 42

There are a few reasons. First, for the example of listing java files, one ls command is enough with wildcards, and we need not use pipes, so this is simpler. Second, wildcards can also be used with many other commands such as cp to perform operations on a set of files, while regular expressions can only work with some commands such as grep.

Answer 43

– -n: Output lines preceded by line numbers – -i: Ignores case – -v: Output lines that don’t match – -w: Restricts matching to whole words only

Answer 44

– grep : the standard grep – grep -F (or fgrep) : searching for fixed strings – grep -E (or egrep) : support for extended regular expressions

Answer 45

– Include matacharacters: ? + | ( ) { – These metacharacters can be used in BRE with a backslash; e.g., \? – Back-referencing; e.g., (...)\1 – Further extension: PCRE — Perl-Compatible Regular Expressions The extended regular expressions include the use of parentheses, and also the so-called back-referencing expressions ‘\n’, such as \1, \2, \3, and so on, which match substrings captured with parentheses. As an example, the extended regular expression: ([ab])cd\1 would match strings: acda and bcdb, since in the first case ‘a’ is captured in the first set of parentheses, and ‘b’ is captured in the second case. As another example, we can use these kinds of expressions to find words that have the first letter the same as the last one, and the second letter the same as the second last letter. If we use a dictionary file, in which each word is placed in one line, we could use the following regular expression: ˆ(.)(.).*\2\1$ The back-referencing (backslash-n or \n) expressions, like \1 and \2 can be used several times in the same expression; for example, the expression: ˆ(.)(.).*\2\2\1\1$ would match the word ‘abcdebbaa’.

Answer 46

(i) Control-a: Moving to the beginning of a line; | (ii) Control-e: Moving to the end of a line.

Answer 47

(b) Press Control-@ to mark the first character of the block of text that you wish to copy. Emacs will probably display the following message at the bottom of the screen “Mark set”. Once a mark is set, all text from that position to the current position of the cursor is known as the current region in emacs. A note: Control-@ key combination is usually denoted as C-@ in Emacs. (c) Move your cursor to the end of the third line. (d) Press ESC w (Esc and then w) to copy the above block of text, i.e., the region, into the working memory. On a Unix keyboard this combination can be obtained with Meta-w, or you should be able to obtain it on a PC keyboard with Alt-w. In Emacs documents it is usually referred to as M-w. (e) Move your cursor to the line below the last line of your program. (f) C-y to paste the above block of text.

Answer 48

emacs hot key for undo is C-x u

Answer 49

There are advantages of separating the shell from the kernel. First, if something goes wrong in the shell, it will not bring down the system. Other programs can still be running. Second, shells can be replaced without rewriting the kernel. Thus, it is easy to add new functionality by writing a new shell. This is good software development practice.

Answer 50

``` – Built-in commands – Scripts – Variables (local and environment) – Redirection – Wildcards – Pipes – Sequences (conditional and unconditional) – Subshells – Background processing – Command substitution ```

Answer 51

– Bourne shell (/bin/sh): This shell replaced the original Thompson shell (which was the first UNIX shell). – Korn shell (/bin/ksh) – C shell (/bin/csh) – TC shell (/bin/tcsh) – Bash shell (/bin/bash) – Note: Use cat /etc/shells to see shells available

Answer 52

The Bash shell, or the Bourne Again shell, provides backward compatibility with the Bourne shell. It also includes most useful features of C/Korn shells. It is licensed under GPL, and is thus open-source. Most Unix systems have it.

Answer 53

The built-in commands are commands that a shell recognizes and executes internally. They are called built-in commands or internal commands. The code for these commands is executed within the shell program. This program is already in memory when shell is running so it is fast to execute them. However, it is difficult to replace these commands with new versions as they are part of the shell. Therefore, these commands typically execute simple tasks that we use frequently, or tasks which cannot be easily executed from a another process, or ‘child’ process in this case. Some examples of internal commands are cd, logout, and echo, although echo can also be an external command

Answer 54

External commands are executable programs that are stored in the directory hierarchy, separate from the shell. Some examples are: ls, grep, sort, cut, and uniq. (uniq and cut are covered in a lab.)

Answer 55

– The following are usually common variables among different shells: – SHELL is the full pathname of the login shell – HOME is the full pathname of the home directory – PATH is the list of directories searched for a command – USER is the username – MAIL is the full pathname to the mailbox – TERM is the type of the terminal

Answer 56

A program is a piece of code which is inactive and stored in a file

Answer 57

e a process is a running program which is | active. A process is instantiated from a program

Answer 58

– Code, which stores the program – Data, which stores static data, i.e., data whose values are maintained throughout the execution – Heap, which is used for dynamic allocation – Stack, which is used to store temporary data

Answer 59

The active part of the process is the thread of execution. We know that the code section contains the program (instructions). The process runs by executing the program instructions sequentially. The sequential execution of program instructions forms a thread of execution.

Answer 60

– PCB information includes: – Process Identification: PID (process identifier, unique nonnegative integer) – The present position in the thread of execution – Resources allocated to the process (e.g., memory, open files) – Process ownership (user and group) – Process state (running, sleeping, pre-empted, created, zombie)

Answer 61

Processes can only be created by other processes. When a user runs a program, the shell (a process) creates the process of the program. We now describe how a parent process creates a child process. A parent process first ‘forks’ itself into two processes, itself and a child process. The ‘fork’ operation is a system call. Remember that system calls provide an interface for programmers to use kernel functionality. At this point, a new process has been created. However, the new process is a copy (clone) of the old process. In particular, it executes the same code from the same position, but it has its own thread. The new thread (child) is distinguished from the old thread (parent) by the return code of the fork call. Since we usually want the child process to run a different program, the child process calls another system call ‘exec’, to run a new program that replaces the child’s original program (which was the same as its parent’s program). After this, the new program begins executing with the child control block.

Answer 62

``` – Shell facility for: – Starting processes in the background – Changing processes between background and foreground mode – Suspending and resuming processes – Terminating processes – Displaying a list of current processes – Foreground and background processes ```

Answer 63

A foreground process is the process that controls the keyboard of a terminal

Answer 64

A background process cannot read from the terminal, but can output to the screen. To create a background process, we enter a command line with a metacharacter & (ampersand) at the end. If we know that the background process might print a lot of output and error messages, then it might be preferable to redirect stdout and stderr.

Answer 65

– Print jobs: jobs – Print processes: ps – Running job in background: bg or bg %jobID – Bringing job to foreground: fg or fg %jobID – Terminating a job or process: kill, kill %jobID, or kill PID

Answer 66

– C is originally invented as a language for writing an operating system and other system software by Denis Ritchie – C optimizes for machine efficiency at the expense of increased implementation and debugging time – A central difficulty in C programming: programmers must do their own memory management – C assumes that you know what you are doing

Answer 67

``` #include — includes information about C’s standard I/O library. int main() { — is a start of the “main” function. printf — is a function from the standard I/O library to produce formatted output. return 0; — return 0 as a return code from the function, which will be the exit code of the program. ```

Answer 68

– Three steps: – Preprocessing (by a preprocessor): modifies the program by following preprocessor directives – Compiling (by a compiler): translates modified code into object code (machine instructions) – Linking (by a linker): combines object code and additional code and produces an executable program – gcc automatically executes these three steps

Answer 69

``` – Commands intended for the preprocessor – Always begin line with # – One line long – No semicolon at the end – Example: #include — the content of stdio.h to be included into the program before compiling – stdio.h: an (actual) header file: – /usr/include/stdio.h ```

Answer 70

– Building blocks from which C programs are constructed – A function is a group of statements given a name – Library functions: functions provided as a part of the C implementation; e.g., printf – Main function: the function that is called automatically when the program is executed – int main() or int main(void) means that main returns an integer value, and does not take any parameters – Nested functions not allowed by standard, but gcc allows them

Answer 71

``` – A command to be executed when the program runs – Must end with a semicolon – Examples: printf("hello, world\n"); return 0; ```

Answer 72

(a) Open hello.c using emacs. (b) Enter ‘M-x compile’ in emacs. This can be done by pressing ‘Esc x compile’ or ‘Alt-x compile’. If we change our mind, we can press C-g (i.e., Ctrl-g) to abort the function. Going back to ‘verb M-x compile,’, we then press Enter. (c) You will be asked to enter the compile command. Use backspace key to delete the default command entirely, and then enter the gcc -o hello hello.c command given earlier. (d) This will split the current emacs window into two, and the bottom window will show the message generated by the compiler. (e) To go back to the single window mode so that we can see more lines of source code, enter C-x 1 (Control-x, then 1 (digit one)). (f) You can invoke any shell command without exiting emacs, and this includes running the compiled code. To do this, enter M-! (e.g., Esc key followed by the ! key). Then you can enter any shell command. You can now enter ./hello and you will see the result in emacs. (g) Another way to run a program immediately after compilation is to enter M-x compile as before, followed by Enter, and then enter the line: gcc -o hello hello.c && ./hello So, we are entering two commands: compilation, and run the program after. We will see later what is the meaning of &&. A brief explanation for now is that after running the first command (gcc), the second command is run (./hello) only if the first one was successful.

Answer 73

pine mydalemail@dal.ca -attach hello.c

Answer 74

– Macro definition (preprocessor directive): #define PI 3.14159f – or simply #define PI 3.14159 – Preprocessor will replace each occurrence of token PI with the number – A macro definition: – does not define a variable – is oblivious about the content of the replacement – Macro replacement can be any sequence of tokens

Answer 75

– % character starts the specification, and it is mandatory. – flags are optional characters, and there could be zero or more flags, such as + in %+d, specifying a mandatory sign when printing a number, or - specifying left justification. – minimal width is an optional positive integer specifying minimal width of the printout, as in %10d, which may be useful in aligning numbers in a table. – .precision is an optional period (.) followed by an integer (negative integer has the same effect as zero), which is the number of decimal places to be printed for floating point numbers, but also minimal number of digits for integers, and maximal number of characters for strings. – length modifier is an optional character or two, such as l used for integers to denote a long integer, or it can be used with ‘f’ as ‘lf’ to denote double instead of float. However, this particular modification is not necessary. – conversion specifier is the final mandatory character, denoting how the value of the variable is to be converted to the printed string representation. The most frequently used ones are: d for an integer, f for a double or a float, c for a character, s for a string, and % for a literal percent sign.

Midterm I Flashcards

(99 cards)