extended wc

Makefile

@@ -138,13 +138,19 @@ UPROGS=\
 	$U/_usertests\
 	$U/_grind\
 	$U/_wc\
+	$U/_wc_test\
 	$U/_zombie\
 	$U/_logstress\
 	$U/_forphan\
 	$U/_dorphan\
 
-fs.img: mkfs/mkfs README.md $(UPROGS)
-	mkfs/mkfs fs.img README.md $(UPROGS)
+#fs.img: mkfs/mkfs README.md $(UPROGS) tests/wc
+#	mkfs/mkfs fs.img README.md $(UPROGS) tests/wc
+
+fs.img: mkfs/mkfs README.md $(UPROGS) \
+  empty.txt nlonly.txt hello.txt words.txt tabsnl.txt multi.txt longline.txt
+	mkfs/mkfs fs.img README.md $(UPROGS) \
+  empty.txt nlonly.txt hello.txt words.txt tabsnl.txt multi.txt longline.txt
 
 -include kernel/*.d user/*.d
 

docs/wc.md

@@ -0,0 +1,57 @@
+# wc — Count lines, words, characters, bytes, and max line length
+
+## Usage
+
+The `wc` program prints statistics about its input files. By default, it displays **all counts** with labels. You can use flags to select which counts to display:
+
+- `-l` : **Line count** (number of `\n` newlines)
+- `-w` : **Word count** (sequences of non-whitespace: not space, tab, carriage return, newline, vertical tab)
+- `-c` : **Character count** (counts every character, including newlines and tabs; ASCII expected)
+- `-b` : **Byte count** (same as `-c` on xv6)
+- `-L` : **Max line length** (length of the longest line, including its newline)
+
+If multiple flags are given, each selected count is displayed with a label.
+
+### Examples
+
+```sh
+$ wc README.md
+Line count: 42
+Word count: 278
+Character count: 1536
+Byte count: 1536
+Max line length: 62
+File name: README.md
+
+$ wc -l README.md
+Line count: 42
+File name: README.md
+
+$ wc -L longline.txt
+Max line length: 301
+File name: longline.txt
+```
+
+---
+
+## Testing
+
+An automated test runner `/wc_test` is provided. It validates each flag across a set of sample files bundled in the image. **The tests ensure correctness for lines, words, characters, bytes, and max line length.**
+
+### To run tests:
+
+1. Boot the OS and open the shell.
+2. Run:
+   ```
+   /wc_test
+   ```
+
+3. The test runner will output `PASS` or `FAIL` for each test case:
+   ```
+   PASS: wc -l empty.txt
+   PASS: wc -w empty.txt
+   ...
+   wc_test summary: 35 pass, 0 fail
+   ```
+
+4. If all tests pass, your implementation is

hello.txt

@@ -0,0 +1 @@
+hello

kernel/syscall.c

@@ -101,6 +101,7 @@ extern uint64 sys_unlink(void);
 extern uint64 sys_link(void);
 extern uint64 sys_mkdir(void);
 extern uint64 sys_close(void);
+extern uint64 sys_wc(void);
 
 // An array mapping syscall numbers from syscall.h
 // to the function that handles the system call.
@@ -126,6 +127,7 @@ static uint64 (*syscalls[])(void) = {
 [SYS_link]    sys_link,
 [SYS_mkdir]   sys_mkdir,
 [SYS_close]   sys_close,
+[SYS_wc]    sys_wc,
 };
 
 void

kernel/syscall.h

@@ -20,3 +20,4 @@
 #define SYS_link   19
 #define SYS_mkdir  20
 #define SYS_close  21
+#define SYS_wc 22

kernel/sysproc.c

@@ -105,3 +105,10 @@ sys_uptime(void)
   release(&tickslock);
   return xticks;
 }
+
+uint64
+sys_wc(void)
+{
+  printf("Hello from your ne wc function!\n");
+  return 0;
+}

longline.txt

@@ -0,0 +1 @@
+aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

multi.txt

@@ -0,0 +1,2 @@
+one two
+three

nlonly.txt

@@ -0,0 +1 @@
+

tabsnl.txt

@@ -0,0 +1,2 @@
+a	b c
+

user/file.txt

@@ -0,0 +1,4 @@
+hello world
+hello again
+this is a test file
+hello world

user/types.h

@@ -0,0 +1,9 @@
+typedef unsigned int   uint;
+typedef unsigned short ushort;
+typedef unsigned char  uchar;
+typedef unsigned long  ulong;
+
+typedef unsigned long long uint64;
+typedef long long          int64;
+typedef unsigned int       uint32;
+typedef int                int32;

user/user.h

@@ -24,6 +24,7 @@ int getpid(void);
 char* sys_sbrk(int,int);
 int pause(int);
 int uptime(void);
+int wc(void);
 
 // ulib.c
 int stat(const char*, struct stat*);

user/usys.pl

@@ -42,3 +42,4 @@ sub entry {
 entry("sbrk");
 entry("pause");
 entry("uptime");
+entry("wc");

user/wc.c

@@ -1,54 +1,103 @@
-#include "kernel/types.h"
-#include "kernel/stat.h"
-#include "kernel/fcntl.h"
-#include "user/user.h"
+#include "types.h"
+#include "user.h"
 
-char buf[512];
+#define BUF_SIZE 512
 
-void
-wc(int fd, char *name)
-{
-  int i, n;
-  int l, w, c, inword;
+int is_word_char(char c) {
+  return !(c == ' ' || c == '\n' || c == '\r' || c == '\t' || c == '\v');
+}
+
+void wc_stats(int fd, char *name, int show_lines, int show_words, int show_chars, int show_bytes, int show_max_line) {
+  int lines = 0, words = 0, chars = 0, bytes = 0;
+  int max_line_len = 0, current_line_len = 0, inword = 0;
+  char buf[BUF_SIZE];
+  int n;
 
-  l = w = c = 0;
-  inword = 0;
-  while((n = read(fd, buf, sizeof(buf))) > 0){
-    for(i=0; i<n; i++){
-      c++;
-      if(buf[i] == '\n')
-        l++;
-      if(strchr(" \r\t\n\v", buf[i]))
+  while ((n = read(fd, buf, sizeof(buf))) > 0) {
+    bytes += n;
+    for (int i = 0; i < n; i++) {
+      char c = buf[i];
+      chars++;
+      current_line_len++;
+      if (c == '\n') {
+        lines++;
+        if (current_line_len > max_line_len)
+          max_line_len = current_line_len;
+        current_line_len = 0;
+      }
+      if (is_word_char(c)) {
+        if (!inword) {
+          words++;
+          inword = 1;
+        }
+      } else {
         inword = 0;
-      else if(!inword){
-        w++;
-        inword = 1;
       }
     }
   }
-  if(n < 0){
+  // Check for last line without newline
+  if (current_line_len > 0 && current_line_len > max_line_len)
+    max_line_len = current_line_len;
+
+  if (n < 0) {
     printf("wc: read error\n");
     exit(1);
   }
-  printf("%d %d %d %s\n", l, w, c, name);
-}
 
-int
-main(int argc, char *argv[])
-{
-  int fd, i;
+  int flag_used = show_lines || show_words || show_chars || show_bytes || show_max_line;
+  if (flag_used) {
+    if (show_lines) {
+      printf("Line count: %d\t(Each '\\n' in the file is counted as a line)\n", lines);
+    }
+    if (show_words) {
+      printf("Word count: %d\t(A word is a sequence of non-whitespace characters)\n", words);
+    }
+    if (show_chars) {
+      printf("Character count: %d\t(Counts every character including newlines and tabs)\n", chars);
+    }
+    if (show_bytes) {
+      printf("Byte count: %d\t(Total bytes read from the file)\n", bytes);
+    }
+    if (show_max_line) {
+      printf("Max line length: %d\t(The longest line, including its newline, in characters)\n", max_line_len);
+    }
+    printf("File name: %s\n", name);
+  } else {
+    printf("Line count: %d\t(Each '\\n' in the file is counted as a line)\n", lines);
+    printf("Word count: %d\t(A word is a sequence of non-whitespace characters)\n", words);
+    printf("Character count: %d\t(Counts every character including newlines and tabs)\n", chars);
+    printf("Byte count: %d\t(Total bytes read from the file)\n", bytes);
+    printf("Max line length: %d\t(The longest line, including its newline, in characters)\n", max_line_len);
+    printf("File name: %s\n", name);
+  }
+}
 
-  if(argc <= 1){
-    wc(0, "");
+int main(int argc, char *argv[]) {
+  int show_lines = 0, show_words = 0, show_chars = 0, show_bytes = 0, show_max_line = 0, file_start = 1;
+  int i;
+  for (i = 1; i < argc; i++) {
+    if (argv[i][0] == '-') {
+      for (int j = 1; argv[i][j]; j++) {
+        if (argv[i][j] == 'l') show_lines = 1;
+        else if (argv[i][j] == 'w') show_words = 1;
+        else if (argv[i][j] == 'c') show_chars = 1;
+        else if (argv[i][j] == 'b') show_bytes = 1;
+        else if (argv[i][j] == 'L') show_max_line = 1;
+      }
+      file_start++;
+    }
+  }
+  if (argc == file_start) {
+    wc_stats(0, "", show_lines, show_words, show_chars, show_bytes, show_max_line);
     exit(0);
   }
-
-  for(i = 1; i < argc; i++){
-    if((fd = open(argv[i], O_RDONLY)) < 0){
+  for (i = file_start; i < argc; i++) {
+    int fd = open(argv[i], 0);
+    if (fd < 0) {
       printf("wc: cannot open %s\n", argv[i]);
-      exit(1);
+      continue;
     }
-    wc(fd, argv[i]);
+    wc_stats(fd, argv[i], show_lines, show_words, show_chars, show_bytes, show_max_line);
     close(fd);
   }
   exit(0);