summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorErik Andersen1999-12-29 03:34:00 +0000
committerErik Andersen1999-12-29 03:34:00 +0000
commita89910829c73c5ed4f30eecb181d7b3087ca4f74 (patch)
tree9467006b9e64b9051ec48c6fcccd30e50c587f84
parentf664c004db12cdb3e379c1745351e225c7b92de7 (diff)
downloadbusybox-a89910829c73c5ed4f30eecb181d7b3087ca4f74.zip
busybox-a89910829c73c5ed4f30eecb181d7b3087ca4f74.tar.gz
Add the beginings of some regression testing. Add in the recently
contributed loadacm utility (for dealing with unicode fonts) -Erik
-rw-r--r--Changelog2
-rw-r--r--applets/busybox.c3
-rw-r--r--busybox.c3
-rw-r--r--busybox.def.h1
-rw-r--r--console-tools/loadacm.c441
-rw-r--r--internal.h1
-rw-r--r--loadacm.c441
-rwxr-xr-xreg_test.sh136
-rw-r--r--utility.c2
9 files changed, 1029 insertions, 1 deletions
diff --git a/Changelog b/Changelog
index 8d0b78e..8ba7369 100644
--- a/Changelog
+++ b/Changelog
@@ -15,6 +15,8 @@
special files.
* cp and mv were quite broken when moving directories. I have rewritten
them so they should now work as expected.
+ * New app: loadacm contributed by Peter Novodvorsky <petya@logic.ru>
+ for loading application character maps for working with Unicode fonts.
-Erik Andersen
diff --git a/applets/busybox.c b/applets/busybox.c
index cd4ab8f..403b140 100644
--- a/applets/busybox.c
+++ b/applets/busybox.c
@@ -232,6 +232,9 @@ static const struct Applet applets[] = {
#ifdef BB_GZIP //bin
{"gzip", gzip_main},
#endif
+#ifdef BB_LOADACM //usr/bin
+ {"loadacm", loadacm_main},
+#endif
{0}
};
diff --git a/busybox.c b/busybox.c
index cd4ab8f..403b140 100644
--- a/busybox.c
+++ b/busybox.c
@@ -232,6 +232,9 @@ static const struct Applet applets[] = {
#ifdef BB_GZIP //bin
{"gzip", gzip_main},
#endif
+#ifdef BB_LOADACM //usr/bin
+ {"loadacm", loadacm_main},
+#endif
{0}
};
diff --git a/busybox.def.h b/busybox.def.h
index 3c290f7..3982a1c 100644
--- a/busybox.def.h
+++ b/busybox.def.h
@@ -33,6 +33,7 @@
#define BB_KLOGD
//#define BB_LENGTH
#define BB_LN
+//#define BB_LOADACM
//#define BB_LOADFONT
//#define BB_LOADKMAP
#define BB_LOGGER
diff --git a/console-tools/loadacm.c b/console-tools/loadacm.c
new file mode 100644
index 0000000..491bc5c
--- /dev/null
+++ b/console-tools/loadacm.c
@@ -0,0 +1,441 @@
+/*
+ * Derived from
+ * mapscrn.c - version 0.92
+ *
+ * Was taken from console-tools and adapted by
+ * Peter Novodvorsky <petya@logic.ru>
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <memory.h>
+#include <string.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <assert.h>
+#include <errno.h>
+#include <signal.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/ioctl.h>
+#include <sys/kd.h>
+
+typedef unsigned short unicode;
+
+static long int ctoi(unsigned char *s, int *is_unicode);
+int old_screen_map_read_ascii(FILE *fp, unsigned char buf[]);
+int uni_screen_map_read_ascii(FILE *fp, unicode buf[], int* is_unicode);
+unicode utf8_to_ucs2 (char* buf);
+int screen_map_load(int fd, FILE *fp);
+
+int loadacm_main(int argc, char **argv)
+{
+ int fd;
+
+ fd = open("/dev/tty", O_RDWR);
+ if (fd < 0) {
+ fprintf(stderr, "Error opening /dev/tty1: %s\n", strerror(errno));
+ return 1;
+ }
+
+ if (screen_map_load(fd, stdin))
+ {
+ fprintf(stderr, "Error loading acm: %s\n", strerror(errno));
+ return 1;
+ }
+
+ write(fd, "\033(K", 3);
+
+ return 0;
+}
+
+int screen_map_load(int fd, FILE *fp)
+{
+ struct stat stbuf;
+ unicode wbuf[E_TABSZ];
+ unsigned char buf[E_TABSZ];
+ int parse_failed = 0;
+ int is_unicode;
+
+ if (fstat(fp->_fileno, &stbuf))
+ perror("Cannot stat map file"), exit(1);
+
+ /* first try a UTF screen-map: either ASCII (no restriction) or binary (regular file) */
+ if (!(parse_failed = (-1 == uni_screen_map_read_ascii(fp,wbuf,&is_unicode))) ||
+ (S_ISREG(stbuf.st_mode) &&
+ (stbuf.st_size == (sizeof(unicode) * E_TABSZ)))) /* test for binary UTF map by size */
+ {
+ if (parse_failed)
+ {
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ fprintf (stderr, "16bit screen-map MUST be a regular file.\n"), exit (1);
+ else
+ perror ("fseek failed reading binary 16bit screen-map"), exit (1);
+ }
+
+ if (fread(wbuf, sizeof(unicode) * E_TABSZ, 1, fp) != 1)
+ perror("Cannot read [new] map from file"), exit(1);
+#if 0
+ else
+ fprintf(stderr, "Input screen-map is binary.\n");
+#endif
+ }
+
+ /* if it was effectively a 16-bit ASCII, OK, else try to read as 8-bit map */
+ /* same if it was binary, ie. if parse_failed */
+ if (parse_failed || is_unicode)
+ {
+ if (ioctl(fd,PIO_UNISCRNMAP,wbuf))
+ perror("PIO_UNISCRNMAP ioctl"), exit(1);
+ else
+ return 0;
+ }
+ }
+
+ /* rewind... */
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ fprintf (stderr, "Assuming 8bit screen-map - MUST be a regular file.\n"), exit (1);
+ else
+ perror ("fseek failed assuming 8bit screen-map"), exit (1);
+ }
+
+ /* ... and try an old 8-bit screen-map */
+ if (!(parse_failed = (-1 == old_screen_map_read_ascii(fp,buf))) ||
+ (S_ISREG(stbuf.st_mode) &&
+ (stbuf.st_size == E_TABSZ))) /* test for binary old 8-bit map by size */
+ {
+ if (parse_failed)
+ {
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ /* should not - it succedeed above */
+ fprintf (stderr, "fseek() returned ESPIPE !\n"), exit (1);
+ else
+ perror ("fseek for binary 8bit screen-map"), exit (1);
+ }
+
+ if (fread(buf,E_TABSZ,1,fp) != 1)
+ perror("Cannot read [old] map from file"), exit(1);
+#if 0
+ else
+ fprintf(stderr, "Input screen-map is binary.\n");
+#endif
+ }
+
+ if (ioctl(fd,PIO_SCRNMAP,buf))
+ perror("PIO_SCRNMAP ioctl"), exit(1);
+ else
+ return 0;
+ }
+ else
+ {
+ fprintf(stderr, "Error parsing symbolic map\n");
+ exit(1);
+ }
+}
+
+
+/*
+ * - reads `fp' as a 16-bit ASCII SFM file.
+ * - returns -1 on error.
+ * - returns it in `unicode' in an E_TABSZ-elements array.
+ * - sets `*is_unicode' flagiff there were any non-8-bit
+ * (ie. real 16-bit) mapping.
+ *
+ * FIXME: ignores everything after second word
+ */
+int uni_screen_map_read_ascii(FILE *fp, unicode buf[], int *is_unicode)
+{
+ char buffer[256]; /* line buffer reading file */
+ char *p, *q; /* 1st + 2nd words in line */
+ int in, on; /* the same, as numbers */
+ int tmp_is_unicode; /* tmp for is_unicode calculation */
+ int i; /* loop index - result holder */
+ int ret_code = 0; /* return code */
+ sigset_t sigset, old_sigset;
+
+ assert (is_unicode);
+
+ *is_unicode = 0;
+
+ /* first 128 codes defaults to ASCII */
+ for (i=0; i<128; i++) buf[i] = i;
+ /* remaining defaults to replacement char (usually E_TABSZ = 256) */
+ for ( ; i<E_TABSZ; i++) buf[i] = 0xfffd;
+
+ /* block SIGCHLD */
+ sigemptyset (&sigset);
+ sigaddset (&sigset, SIGCHLD);
+ sigprocmask (SIG_BLOCK, &sigset, &old_sigset);
+
+ do
+ {
+ if (NULL == fgets(buffer, sizeof(buffer),fp))
+ {
+ if (feof (fp))
+ break;
+ else
+ {
+ perror ("uni_screen_map_read_ascii() can't read line");
+ exit (2);
+ }
+ }
+
+ /* get "charset-relative charcode", stripping leading spaces */
+ p = strtok(buffer," \t\n");
+
+ /* skip empty lines and comments */
+ if (!p || *p == '#')
+ continue;
+
+ /* get unicode mapping */
+ q = strtok(NULL," \t\n");
+ if (q)
+ {
+ in = ctoi(p, NULL);
+ if (in < 0 || in > 255)
+ {
+ ret_code = -1;
+ break;
+ }
+
+ on = ctoi(q, &tmp_is_unicode);
+ if (in < 0 && on > 65535)
+ {
+ ret_code = -1;
+ break;
+ }
+
+ *is_unicode |= tmp_is_unicode;
+ buf[in] = on;
+ }
+ else
+ {
+ ret_code = -1;
+ break;
+ }
+ }
+ while (1); /* terminated by break on feof() */
+
+ /* restore sig mask */
+ sigprocmask (SIG_SETMASK, &old_sigset, NULL);
+
+ return ret_code;
+}
+
+
+int old_screen_map_read_ascii(FILE *fp, unsigned char buf[])
+{
+ char buffer[256];
+ int in, on;
+ char *p, *q;
+
+ for (in=0; in<256; in++) buf[in]=in;
+
+ while (fgets(buffer,sizeof(buffer)-1,fp))
+ {
+ p = strtok(buffer," \t\n");
+
+ if (!p || *p == '#')
+ continue;
+
+ q = strtok(NULL," \t\n#");
+ if (q)
+ {
+ in = ctoi(p, NULL);
+ if (in < 0 || in > 255) return -1;
+
+ on = ctoi(q, NULL);
+ if (in < 0 && on > 255) return -1;
+
+ buf[in] = on;
+ }
+ else return -1;
+ }
+
+ return(0);
+}
+
+
+/*
+ * - converts a string into an int.
+ * - supports dec and hex bytes, hex UCS2, single-quoted byte and UTF8 chars.
+ * - returns the converted value
+ * - if `is_unicode != NULL', use it to tell whether it was unicode
+ *
+ * CAVEAT: will report valid UTF mappings using only 1 byte as 8-bit ones.
+ */
+long int ctoi(unsigned char *s, int *is_unicode)
+{
+ int i;
+ size_t ls;
+
+ ls = strlen(s);
+ if (is_unicode) *is_unicode = 0;
+
+ /* hex-specified UCS2 */
+ if ((strncmp(s,"U+",2) == 0) &&
+ (strspn(s+2,"0123456789abcdefABCDEF") == ls-2))
+ {
+ sscanf(s+2,"%x",&i);
+ if (is_unicode) *is_unicode = 1;
+ }
+
+ /* hex-specified byte */
+ else if ((ls <= 4) && (strncmp(s,"0x",2) == 0) &&
+ (strspn(s+2,"0123456789abcdefABCDEF") == ls-2))
+ sscanf(s+2,"%x",&i);
+
+ /* oct-specified number (byte) */
+ else if ((*s == '0') &&
+ (strspn(s,"01234567") == ls))
+ sscanf(s,"%o",&i);
+
+ /* dec-specified number (byte) */
+ else if (strspn(s,"0123456789") == ls)
+ sscanf(s,"%d",&i);
+
+ /* single-byte quoted char */
+ else if ((strlen(s) == 3) && (s[0] == '\'') && (s[2] == '\''))
+ i=s[1];
+
+ /* multi-byte UTF8 quoted char */
+ else if ((s[0] == '\'') && (s[ls-1] == '\''))
+ {
+ s[ls-1] = 0; /* ensure we'll not "parse UTF too far" */
+ i = utf8_to_ucs2(s+1);
+ if (is_unicode) *is_unicode = 1;
+ }
+ else
+ return(-1);
+
+ return(i);
+}
+
+
+void saveoldmap(int fd, char *omfil)
+{
+ FILE *fp;
+ char buf[E_TABSZ];
+#ifdef GIO_UNISCRNMAP
+ unicode xbuf[E_TABSZ];
+ int is_old_map = 0;
+
+ if (ioctl(fd,GIO_UNISCRNMAP,xbuf))
+ {
+ perror("GIO_UNISCRNMAP ioctl error");
+#endif
+ if (ioctl(fd,GIO_SCRNMAP,buf))
+ {
+ perror("GIO_SCRNMAP ioctl error");
+ exit(1);
+ }
+ else
+ is_old_map = 1;
+#ifdef GIO_UNISCRNMAP
+ }
+#endif
+
+ if ((fp = fopen(omfil, "w")) == NULL)
+ {
+ perror(omfil);
+ exit(1);
+ }
+
+#ifdef GIO_UNISCRNMAP
+ if (is_old_map)
+ {
+#endif
+ if (fwrite(buf,E_TABSZ,1,fp) != 1)
+ {
+ perror("Error writing map to file");
+ exit(1);
+ }
+#ifdef GIO_UNISCRNMAP
+ }
+ else
+ if (fwrite(xbuf, sizeof(unicode) * E_TABSZ,1,fp) != 1)
+ {
+ perror("Error writing map to file");
+ exit(1);
+ }
+#endif
+
+ fclose(fp);
+}
+
+unicode utf8_to_ucs2 (char* buf)
+{
+ int utf_count = 0;
+ long utf_char;
+ unicode tc;
+ unsigned char c;
+
+ do
+ {
+ c = *buf;
+ buf++;
+
+ /* if byte should be part of multi-byte sequence */
+ if(c & 0x80)
+ {
+ /* if we have already started to parse a UTF8 sequence */
+ if (utf_count > 0 && (c & 0xc0) == 0x80)
+ {
+ utf_char = (utf_char << 6) | (c & 0x3f);
+ utf_count--;
+ if (utf_count == 0)
+ tc = utf_char;
+ else
+ continue;
+ }
+ else /* Possibly 1st char of a UTF8 sequence */
+ {
+ if ((c & 0xe0) == 0xc0)
+ {
+ utf_count = 1;
+ utf_char = (c & 0x1f);
+ }
+ else if ((c & 0xf0) == 0xe0)
+ {
+ utf_count = 2;
+ utf_char = (c & 0x0f);
+ }
+ else if ((c & 0xf8) == 0xf0)
+ {
+ utf_count = 3;
+ utf_char = (c & 0x07);
+ }
+ else if ((c & 0xfc) == 0xf8)
+ {
+ utf_count = 4;
+ utf_char = (c & 0x03);
+ }
+ else if ((c & 0xfe) == 0xfc)
+ {
+ utf_count = 5;
+ utf_char = (c & 0x01);
+ }
+ else
+ utf_count = 0;
+ continue;
+ }
+ }
+ else /* not part of multi-byte sequence - treat as ASCII
+ * this makes incomplete sequences to be ignored
+ */
+ {
+ tc = c;
+ utf_count = 0;
+ }
+ }
+ while (utf_count);
+
+ return tc;
+}
+
diff --git a/internal.h b/internal.h
index 05e3904..1e42982 100644
--- a/internal.h
+++ b/internal.h
@@ -130,6 +130,7 @@ extern int update_main(int argc, char** argv);
extern int uname_main(int argc, char** argv);
extern int gunzip_main (int argc, char** argv);
extern int gzip_main(int argc, char** argv);
+extern int loadacm_main(int argc, char** argv);
const char *modeString(int mode);
diff --git a/loadacm.c b/loadacm.c
new file mode 100644
index 0000000..491bc5c
--- /dev/null
+++ b/loadacm.c
@@ -0,0 +1,441 @@
+/*
+ * Derived from
+ * mapscrn.c - version 0.92
+ *
+ * Was taken from console-tools and adapted by
+ * Peter Novodvorsky <petya@logic.ru>
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <memory.h>
+#include <string.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <assert.h>
+#include <errno.h>
+#include <signal.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/ioctl.h>
+#include <sys/kd.h>
+
+typedef unsigned short unicode;
+
+static long int ctoi(unsigned char *s, int *is_unicode);
+int old_screen_map_read_ascii(FILE *fp, unsigned char buf[]);
+int uni_screen_map_read_ascii(FILE *fp, unicode buf[], int* is_unicode);
+unicode utf8_to_ucs2 (char* buf);
+int screen_map_load(int fd, FILE *fp);
+
+int loadacm_main(int argc, char **argv)
+{
+ int fd;
+
+ fd = open("/dev/tty", O_RDWR);
+ if (fd < 0) {
+ fprintf(stderr, "Error opening /dev/tty1: %s\n", strerror(errno));
+ return 1;
+ }
+
+ if (screen_map_load(fd, stdin))
+ {
+ fprintf(stderr, "Error loading acm: %s\n", strerror(errno));
+ return 1;
+ }
+
+ write(fd, "\033(K", 3);
+
+ return 0;
+}
+
+int screen_map_load(int fd, FILE *fp)
+{
+ struct stat stbuf;
+ unicode wbuf[E_TABSZ];
+ unsigned char buf[E_TABSZ];
+ int parse_failed = 0;
+ int is_unicode;
+
+ if (fstat(fp->_fileno, &stbuf))
+ perror("Cannot stat map file"), exit(1);
+
+ /* first try a UTF screen-map: either ASCII (no restriction) or binary (regular file) */
+ if (!(parse_failed = (-1 == uni_screen_map_read_ascii(fp,wbuf,&is_unicode))) ||
+ (S_ISREG(stbuf.st_mode) &&
+ (stbuf.st_size == (sizeof(unicode) * E_TABSZ)))) /* test for binary UTF map by size */
+ {
+ if (parse_failed)
+ {
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ fprintf (stderr, "16bit screen-map MUST be a regular file.\n"), exit (1);
+ else
+ perror ("fseek failed reading binary 16bit screen-map"), exit (1);
+ }
+
+ if (fread(wbuf, sizeof(unicode) * E_TABSZ, 1, fp) != 1)
+ perror("Cannot read [new] map from file"), exit(1);
+#if 0
+ else
+ fprintf(stderr, "Input screen-map is binary.\n");
+#endif
+ }
+
+ /* if it was effectively a 16-bit ASCII, OK, else try to read as 8-bit map */
+ /* same if it was binary, ie. if parse_failed */
+ if (parse_failed || is_unicode)
+ {
+ if (ioctl(fd,PIO_UNISCRNMAP,wbuf))
+ perror("PIO_UNISCRNMAP ioctl"), exit(1);
+ else
+ return 0;
+ }
+ }
+
+ /* rewind... */
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ fprintf (stderr, "Assuming 8bit screen-map - MUST be a regular file.\n"), exit (1);
+ else
+ perror ("fseek failed assuming 8bit screen-map"), exit (1);
+ }
+
+ /* ... and try an old 8-bit screen-map */
+ if (!(parse_failed = (-1 == old_screen_map_read_ascii(fp,buf))) ||
+ (S_ISREG(stbuf.st_mode) &&
+ (stbuf.st_size == E_TABSZ))) /* test for binary old 8-bit map by size */
+ {
+ if (parse_failed)
+ {
+ if (-1 == fseek (fp, 0, SEEK_SET))
+ {
+ if (errno == ESPIPE)
+ /* should not - it succedeed above */
+ fprintf (stderr, "fseek() returned ESPIPE !\n"), exit (1);
+ else
+ perror ("fseek for binary 8bit screen-map"), exit (1);
+ }
+
+ if (fread(buf,E_TABSZ,1,fp) != 1)
+ perror("Cannot read [old] map from file"), exit(1);
+#if 0
+ else
+ fprintf(stderr, "Input screen-map is binary.\n");
+#endif
+ }
+
+ if (ioctl(fd,PIO_SCRNMAP,buf))
+ perror("PIO_SCRNMAP ioctl"), exit(1);
+ else
+ return 0;
+ }
+ else
+ {
+ fprintf(stderr, "Error parsing symbolic map\n");
+ exit(1);
+ }
+}
+
+
+/*
+ * - reads `fp' as a 16-bit ASCII SFM file.
+ * - returns -1 on error.
+ * - returns it in `unicode' in an E_TABSZ-elements array.
+ * - sets `*is_unicode' flagiff there were any non-8-bit
+ * (ie. real 16-bit) mapping.
+ *
+ * FIXME: ignores everything after second word
+ */
+int uni_screen_map_read_ascii(FILE *fp, unicode buf[], int *is_unicode)
+{
+ char buffer[256]; /* line buffer reading file */
+ char *p, *q; /* 1st + 2nd words in line */
+ int in, on; /* the same, as numbers */
+ int tmp_is_unicode; /* tmp for is_unicode calculation */
+ int i; /* loop index - result holder */
+ int ret_code = 0; /* return code */
+ sigset_t sigset, old_sigset;
+
+ assert (is_unicode);
+
+ *is_unicode = 0;
+
+ /* first 128 codes defaults to ASCII */
+ for (i=0; i<128; i++) buf[i] = i;
+ /* remaining defaults to replacement char (usually E_TABSZ = 256) */
+ for ( ; i<E_TABSZ; i++) buf[i] = 0xfffd;
+
+ /* block SIGCHLD */
+ sigemptyset (&sigset);
+ sigaddset (&sigset, SIGCHLD);
+ sigprocmask (SIG_BLOCK, &sigset, &old_sigset);
+
+ do
+ {
+ if (NULL == fgets(buffer, sizeof(buffer),fp))
+ {
+ if (feof (fp))
+ break;
+ else
+ {
+ perror ("uni_screen_map_read_ascii() can't read line");
+ exit (2);
+ }
+ }
+
+ /* get "charset-relative charcode", stripping leading spaces */
+ p = strtok(buffer," \t\n");
+
+ /* skip empty lines and comments */
+ if (!p || *p == '#')
+ continue;
+
+ /* get unicode mapping */
+ q = strtok(NULL," \t\n");
+ if (q)
+ {
+ in = ctoi(p, NULL);
+ if (in < 0 || in > 255)
+ {
+ ret_code = -1;
+ break;
+ }
+
+ on = ctoi(q, &tmp_is_unicode);
+ if (in < 0 && on > 65535)
+ {
+ ret_code = -1;
+ break;
+ }
+
+ *is_unicode |= tmp_is_unicode;
+ buf[in] = on;
+ }
+ else
+ {
+ ret_code = -1;
+ break;
+ }
+ }
+ while (1); /* terminated by break on feof() */
+
+ /* restore sig mask */
+ sigprocmask (SIG_SETMASK, &old_sigset, NULL);
+
+ return ret_code;
+}
+
+
+int old_screen_map_read_ascii(FILE *fp, unsigned char buf[])
+{
+ char buffer[256];
+ int in, on;
+ char *p, *q;
+
+ for (in=0; in<256; in++) buf[in]=in;
+
+ while (fgets(buffer,sizeof(buffer)-1,fp))
+ {
+ p = strtok(buffer," \t\n");
+
+ if (!p || *p == '#')
+ continue;
+
+ q = strtok(NULL," \t\n#");
+ if (q)
+ {
+ in = ctoi(p, NULL);
+ if (in < 0 || in > 255) return -1;
+
+ on = ctoi(q, NULL);
+ if (in < 0 && on > 255) return -1;
+
+ buf[in] = on;
+ }
+ else return -1;
+ }
+
+ return(0);
+}
+
+
+/*
+ * - converts a string into an int.
+ * - supports dec and hex bytes, hex UCS2, single-quoted byte and UTF8 chars.
+ * - returns the converted value
+ * - if `is_unicode != NULL', use it to tell whether it was unicode
+ *
+ * CAVEAT: will report valid UTF mappings using only 1 byte as 8-bit ones.
+ */
+long int ctoi(unsigned char *s, int *is_unicode)
+{
+ int i;
+ size_t ls;
+
+ ls = strlen(s);
+ if (is_unicode) *is_unicode = 0;
+
+ /* hex-specified UCS2 */
+ if ((strncmp(s,"U+",2) == 0) &&
+ (strspn(s+2,"0123456789abcdefABCDEF") == ls-2))
+ {
+ sscanf(s+2,"%x",&i);
+ if (is_unicode) *is_unicode = 1;
+ }
+
+ /* hex-specified byte */
+ else if ((ls <= 4) && (strncmp(s,"0x",2) == 0) &&
+ (strspn(s+2,"0123456789abcdefABCDEF") == ls-2))
+ sscanf(s+2,"%x",&i);
+
+ /* oct-specified number (byte) */
+ else if ((*s == '0') &&
+ (strspn(s,"01234567") == ls))
+ sscanf(s,"%o",&i);
+
+ /* dec-specified number (byte) */
+ else if (strspn(s,"0123456789") == ls)
+ sscanf(s,"%d",&i);
+
+ /* single-byte quoted char */
+ else if ((strlen(s) == 3) && (s[0] == '\'') && (s[2] == '\''))
+ i=s[1];
+
+ /* multi-byte UTF8 quoted char */
+ else if ((s[0] == '\'') && (s[ls-1] == '\''))
+ {
+ s[ls-1] = 0; /* ensure we'll not "parse UTF too far" */
+ i = utf8_to_ucs2(s+1);
+ if (is_unicode) *is_unicode = 1;
+ }
+ else
+ return(-1);
+
+ return(i);
+}
+
+
+void saveoldmap(int fd, char *omfil)
+{
+ FILE *fp;
+ char buf[E_TABSZ];
+#ifdef GIO_UNISCRNMAP
+ unicode xbuf[E_TABSZ];
+ int is_old_map = 0;
+
+ if (ioctl(fd,GIO_UNISCRNMAP,xbuf))
+ {
+ perror("GIO_UNISCRNMAP ioctl error");
+#endif
+ if (ioctl(fd,GIO_SCRNMAP,buf))
+ {
+ perror("GIO_SCRNMAP ioctl error");
+ exit(1);
+ }
+ else
+ is_old_map = 1;
+#ifdef GIO_UNISCRNMAP
+ }
+#endif
+
+ if ((fp = fopen(omfil, "w")) == NULL)
+ {
+ perror(omfil);
+ exit(1);
+ }
+
+#ifdef GIO_UNISCRNMAP
+ if (is_old_map)
+ {
+#endif
+ if (fwrite(buf,E_TABSZ,1,fp) != 1)
+ {
+ perror("Error writing map to file");
+ exit(1);
+ }
+#ifdef GIO_UNISCRNMAP
+ }
+ else
+ if (fwrite(xbuf, sizeof(unicode) * E_TABSZ,1,fp) != 1)
+ {
+ perror("Error writing map to file");
+ exit(1);
+ }
+#endif
+
+ fclose(fp);
+}
+
+unicode utf8_to_ucs2 (char* buf)
+{
+ int utf_count = 0;
+ long utf_char;
+ unicode tc;
+ unsigned char c;
+
+ do
+ {
+ c = *buf;
+ buf++;
+
+ /* if byte should be part of multi-byte sequence */
+ if(c & 0x80)
+ {
+ /* if we have already started to parse a UTF8 sequence */
+ if (utf_count > 0 && (c & 0xc0) == 0x80)
+ {
+ utf_char = (utf_char << 6) | (c & 0x3f);
+ utf_count--;
+ if (utf_count == 0)
+ tc = utf_char;
+ else
+ continue;
+ }
+ else /* Possibly 1st char of a UTF8 sequence */
+ {
+ if ((c & 0xe0) == 0xc0)
+ {
+ utf_count = 1;
+ utf_char = (c & 0x1f);
+ }
+ else if ((c & 0xf0) == 0xe0)
+ {
+ utf_count = 2;
+ utf_char = (c & 0x0f);
+ }
+ else if ((c & 0xf8) == 0xf0)
+ {
+ utf_count = 3;
+ utf_char = (c & 0x07);
+ }
+ else if ((c & 0xfc) == 0xf8)
+ {
+ utf_count = 4;
+ utf_char = (c & 0x03);
+ }
+ else if ((c & 0xfe) == 0xfc)
+ {
+ utf_count = 5;
+ utf_char = (c & 0x01);
+ }
+ else
+ utf_count = 0;
+ continue;
+ }
+ }
+ else /* not part of multi-byte sequence - treat as ASCII
+ * this makes incomplete sequences to be ignored
+ */
+ {
+ tc = c;
+ utf_count = 0;
+ }
+ }
+ while (utf_count);
+
+ return tc;
+}
+
diff --git a/reg_test.sh b/reg_test.sh
new file mode 100755
index 0000000..0a30ae3
--- /dev/null
+++ b/reg_test.sh
@@ -0,0 +1,136 @@
+#!/bin/sh
+
+
+
+rm -rf testdir
+./busybox cp tar.c testdir
+
+if ! eval diff -u tar.c testdir ; then
+ echo " "
+ echo "Bummer. File copy failed."
+ exit 0
+else
+ echo "Cool. File copy is ok."
+fi
+echo " "
+
+rm -rf testdir
+mkdir -p testdir/foo
+./busybox cp tar.c testdir/foo
+
+if ! eval diff -u tar.c testdir/foo/tar.c ; then
+ echo " "
+ echo "Bummer. File copy to a directory failed."
+ exit 0
+else
+ echo "Cool. File copy to a directory is ok."
+fi
+echo " "
+
+
+rm -rf testdir
+mkdir -p testdir/foo
+./busybox cp tar.c testdir/foo/
+
+if ! eval diff -u tar.c testdir/foo/tar.c ; then
+ echo " "
+ echo "Bummer. File copy to a directory w/ a '/' failed."
+ exit 0
+else
+ echo "Cool. File copy to a directory w/ a '/' is ok."
+fi
+echo " "
+
+
+rm -rf testdir X11
+cp -a /etc/X11 .
+./busybox cp -a X11 testdir
+
+if ! eval diff -ur X11 testdir ; then
+ echo " "
+ echo "Bummer. Local dir copy failed."
+ exit 0
+else
+ echo "Cool. Local dir copy is ok."
+fi
+echo " "
+
+rm -rf testdir X11
+cp -a /etc/X11 .
+./busybox cp -a X11 testdir/
+
+if ! eval diff -ur X11 testdir ; then
+ echo " "
+ echo "Bummer. Local dir copy w/ a '/' failed."
+ exit 0
+else
+ echo "Cool. Local dir copy w/ a '/' is ok."
+fi
+echo " "
+
+rm -rf testdir X11
+cp -a /etc/X11 .
+./busybox cp -a X11/ testdir
+
+if ! eval diff -ur X11 testdir ; then
+ echo " "
+ echo "Bummer. Local dir copy w/ a src '/' failed."
+ exit 0
+else
+ echo "Cool. Local dir copy w/ a src '/' is ok."
+fi
+echo " "
+
+rm -rf testdir X11
+cp -a /etc/X11 .
+./busybox cp -a X11/ testdir/
+
+if ! eval diff -ur X11 testdir ; then
+ echo " "
+ echo "Bummer. Local dir copy w/ 2x '/'s failed."
+ exit 0
+else
+ echo "Cool. Local dir copy w/ 2x '/'s is ok."
+fi
+echo " "
+
+rm -rf testdir X11
+./busybox cp -a /etc/X11 testdir
+if ! eval diff -ur /etc/X11 testdir ; then
+ echo " "
+ echo "Bummer. Remote dir copy failed."
+ exit 0
+else
+ echo "Cool. Remote dir copy is ok."
+fi
+echo " "
+
+
+rm -rf testdir X11
+mkdir -p testdir/foo
+
+./busybox cp -a /etc/X11 testdir/foo
+if ! eval diff -ur /etc/X11 testdir/foo ; then
+ echo " "
+ echo "Bummer. Remote dir copy to a directory failed."
+ exit 0
+else
+ echo "Cool. Remote dir copy to a directory is ok."
+fi
+echo " "
+
+
+rm -rf testdir X11
+mkdir -p testdir/foo
+
+./busybox cp -a /etc/X11 testdir/foo/
+if ! eval diff -ur /etc/X11 testdir/foo ; then
+ echo " "
+ echo "Bummer. Remote dir copy to a directory w/ a '/' failed."
+ exit 0
+else
+ echo "Cool. Remote dir copy to a directory w/ a '/' is ok."
+fi
+
+rm -rf testdir
+
diff --git a/utility.c b/utility.c
index d042b99..2a840f7 100644
--- a/utility.c
+++ b/utility.c
@@ -479,7 +479,7 @@ recursiveAction(const char *fileName, int recurse, int followLinks, int depthFir
-#if defined (BB_TAR) || defined (BB_MKDIR) || defined (BB_CP)
+#if defined (BB_TAR) || defined (BB_MKDIR)
/*
* Attempt to create the directories along the specified path, except for
* the final component. The mode is given for the final directory only,