[u/mdw/catacomb] / hashsum.c

/* -*-c-*-
 *
 * $Id$
 *
 * Hash files using some secure hash function
 *
 * (c) 2000 Straylight/Edgeware
 */

/*----- Licensing notice --------------------------------------------------*
 *
 * This file is part of Catacomb.
 *
 * Catacomb is free software; you can redistribute it and/or modify
 * it under the terms of the GNU Library General Public License as
 * published by the Free Software Foundation; either version 2 of the
 * License, or (at your option) any later version.
 *
 * Catacomb is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU Library General Public License for more details.
 *
 * You should have received a copy of the GNU Library General Public
 * License along with Catacomb; if not, write to the Free
 * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
 * MA 02111-1307, USA.
 */

/*----- Header files ------------------------------------------------------*/

#define _FILE_OFFSET_BITS 64

#include "config.h"

#include <assert.h>
#include <ctype.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include <mLib/alloc.h>
#include <mLib/dstr.h>
#include <mLib/mdwopt.h>
#include <mLib/quis.h>
#include <mLib/report.h>
#include <mLib/sub.h>
#include <mLib/str.h>

#include <mLib/hex.h>
#include <mLib/base32.h>
#include <mLib/base64.h>

#include "ghash.h"
#include "cc.h"

#ifndef PATHSEP
#  if defined(__riscos)
#    define PATHSEP '.'
#  elif defined(__unix) || defined(unix)
#    define PATHSEP '/'
#  else
#    define PATHSEP '\\'
#  endif
#endif

/*----- Static variables --------------------------------------------------*/

#define f_binary 1u
#define f_bogus 2u
#define f_verbose 4u
#define f_check 8u
#define f_files 16u
#define f_raw 32u
#define f_oddhash 64u
#define f_escape 128u
#define f_oddenc 256u
#define f_progress 512u

/*----- Encoding and decoding ---------------------------------------------*/

/* --- Hex encoding --- */

static void puthex(const octet *buf, size_t sz, FILE *fp)
{
  while (sz) {
    fprintf(fp, "%02x", *buf++);
    sz--;
  }
}

static size_t gethex(const char *p, octet *q, size_t sz, char **pp)
{
  size_t i = 0;
  while (sz > 0 &&
	 isxdigit((unsigned char)p[0]) &&
	 isxdigit((unsigned char)p[1])) {
    char buf[3];
    buf[0] = p[0];
    buf[1] = p[1];
    buf[2] = 0;
    *q++ = strtoul(buf, 0, 16);
    sz--;
    p += 2;
    i++;
  }
  if (pp)
    *pp = (char *)p;
  return (i);
}

/* --- Base64 encoding --- */

static void putb64(const octet *buf, size_t sz, FILE *fp)
{
  base64_ctx b;
  dstr d = DSTR_INIT;

  base64_init(&b);
  b.indent = "";
  b.maxline = 0;
  base64_encode(&b, buf, sz, &d);
  base64_encode(&b, 0, 0, &d);
  dstr_write(&d, fp);
  dstr_destroy(&d);
}

static size_t getb64(const char *p, octet *q, size_t sz, char **pp)
{
  base64_ctx b;
  dstr d = DSTR_INIT;
  size_t n = strlen(p);

  base64_init(&b);
  base64_decode(&b, p, n, &d);
  if (pp) *pp = (/*unconst*/ char *)p + n;
  base64_decode(&b, 0, 0, &d);
  assert(d.len <= sz);
  memcpy(q, d.buf, sz);
  n = d.len;
  dstr_destroy(&d);
  return (n);
}

/* --- Base32 encoding --- */

static void putb32(const octet *buf, size_t sz, FILE *fp)
{
  base32_ctx b;
  dstr d = DSTR_INIT;

  base32_init(&b);
  b.indent = "";
  b.maxline = 0;
  base32_encode(&b, buf, sz, &d);
  base32_encode(&b, 0, 0, &d);
  dstr_write(&d, fp);
  dstr_destroy(&d);
}

static size_t getb32(const char *p, octet *q, size_t sz, char **pp)
{
  base32_ctx b;
  dstr d = DSTR_INIT;
  size_t n = strlen(p);

  base32_init(&b);
  base32_decode(&b, p, n, &d);
  if (pp) *pp = (/*unconst*/ char *)p + n;
  base32_decode(&b, 0, 0, &d);
  assert(d.len <= sz);
  memcpy(q, d.buf, sz);
  n = d.len;
  dstr_destroy(&d);
  return (n);
}

/* --- Table --- */

typedef struct encodeops {
  const char *name;
  void (*put)(const octet *, size_t, FILE *);
  size_t (*get)(const char *, octet *, size_t, char **);
} encodeops;

static const encodeops encodingtab[] = {
  { "hex", puthex, gethex },
  { "base64", putb64, getb64 },
  { "base32", putb32, getb32 },
  { 0, 0, 0 }
};

static const encodeops *getencoding(const char *ename)
{
  const encodeops *e;

  for (e = encodingtab; e->name; e++) {
    if (strcmp(ename, e->name) == 0)
      return (e);
  }
  return (0);
}

/*----- Support functions -------------------------------------------------*/

/* --- @fhash@ --- *
 *
 * Arguments:	@const char *file@ = file name to be hashed (null for stdin)
 *		@unsigned f@ = flags to set
 *		@const gchash *gch@ = pointer to hash function to use
 *		@void *buf@ = pointer to hash output buffer
 *
 * Returns:	Zero if it worked, nonzero on error.
 *
 * Use:		Hashes a file.
 */

struct unit {
  const char *name;
  int m;
};

static void prhuman_time(FILE *fp, unsigned long n)
{
  const static struct unit utime[] = {
    { "s",  60 }, { "m", 60 }, { "h", 24 }, { "d", 0 }
  };

  unsigned long m = 0;
  const struct unit *u = utime;

  while (u[1].m && n > u[0].m*u[1].m) { n /= u->m; u++; }
  m = n / u[1].m; n %= u[0].m;
  if (m) fprintf(fp, "%3lu%s%02lu%s", m, u[1].name, n, u[0].name);
  else fprintf(fp, "    %2lu%s", n, u[0].name);
}

static void prhuman_data(FILE *fp, off_t n)
{
  const static struct unit udata[] = {
    { " ", 1024 }, { "k", 1024 }, { "M", 1024 }, { "G", 1024 },
    { "T", 1024 }, { "P", 1024 }, { "E", 1024 }, { "Z", 1024 },
    { "Y",    0 }
  };

  double x = n;
  const struct unit *u = udata;

  while (u->m && x >= u->m) { x /= u->m; u++; }
  fprintf(fp, "%6.1f%s", x, u->name);
}

static int fhash(const char *file, unsigned f, const gchash *gch, void *buf)
{
  FILE *fp;
  char fbuf[1024 * 128];
  size_t sz;
  ghash *h;
  int e;
  off_t fsz = -1, fo;
  const char *p;
  dstr d = DSTR_INIT;
  static char baton[] = "-\\|/";
  char *bp = baton;
  time_t now, last, start;
  int i, pc;

  if (!file || strcmp(file, "-") == 0)
    fp = stdin;
  else if ((fp = fopen(file, f & f_binary ? "rb" : "r")) == 0)
    return (-1);

  if (f & f_progress) {
    if ((fo = ftello(fp)) >= 0 &&
	fseeko(fp, 0, SEEK_END) >= 0 &&
	(fsz = ftello(fp),
	 fseeko(fp, fo, SEEK_SET) < 0))
      return (-1);
    if (fo != -1 && fsz != -1) fsz -= fo;
    fo = 0;
    sz = strlen(file);
    if (sz < 24)
      dstr_puts(&d, file);
    else if ((p = strchr(file + sz - 20, PATHSEP)) != 0) {
      dstr_puts(&d, "..."); dstr_puts(&d, p);
    } else {
      p = strrchr(file, PATHSEP);
      if (!p) dstr_putf(&d, "%.20s...", file);
      else dstr_putf(&d, "...%.17s...", p);
    }
    start = last = time(0);
  }

  h = GH_INIT(gch);
  while ((sz = fread(fbuf, 1, sizeof(fbuf), fp)) > 0) {
    GH_HASH(h, fbuf, sz);
    if (f & f_progress) {
      fo += sz;
      now = time(0);
      if (difftime(now, last) < 1) continue;
      last = now;
      fprintf(stderr, "\r%-24s", d.buf);
      fprintf(stderr, "%c ", *bp++); if (!*bp) bp = baton;
      prhuman_data(stderr, fo);
      if (fsz >= fo) {
	fputc('/', stderr);
	prhuman_data(stderr, fsz);
	fputs(" [", stderr);
	pc = (fo*16 + fsz/2)/fsz;
	for (i = 0; i < pc; i++) fputc('.', stderr);
	for (; i < 16; i++) fputc(' ', stderr);
	fprintf(stderr, "] %3d%%", (int)((fo*100 + 50)/fsz));
	fprintf(stderr, "  ETA ");
	prhuman_time(stderr, difftime(now, start)*(fsz - fo)/fo);
      }
    }
  }
  GH_DONE(h, buf);
  GH_DESTROY(h);
  if (f & f_progress) fprintf(stderr, "\r%78s\r", "");
  e = ferror(fp);
  if (file)
    fclose(fp);
  return (e ? -1 : 0);
}

/* --- @gethash@ --- *
 *
 * Arguments:	@const char *name@ = pointer to name string
 *
 * Returns:	Pointer to appropriate hash class.
 *
 * Use:		Chooses a hash function by name.
 */

static const gchash *gethash(const char *name)
{
  const gchash *const *g, *gg = 0;
  size_t sz = strlen(name);
  for (g = ghashtab; *g; g++) {
    if (strncmp(name, (*g)->name, sz) == 0) {
      if ((*g)->name[sz] == 0) {
	gg = *g;
	break;
      } else if (gg)
	return (0);
      else
	gg = *g;
    }
  }
  return (gg);
}

/* --- @getstring@ --- *
 *
 * Arguments:	@FILE *fp@ = stream from which to read
 *		@const char *p@ = string to read from instead
 *		@dstr *d@ = destination string
 *		@unsigned raw@ = raw or cooked read
 *
 * Returns:	Zero if OK, nonzero on end-of-file.
 *
 * Use:		Reads a filename (or something similar) from a stream.
 */

static int getstring(FILE *fp, const char *p, dstr *d, unsigned raw)
{
  int ch;
  int q = 0;

  /* --- Raw: just read exactly what's written up to a null byte --- */

#define NEXTCH (fp ? getc(fp) : (unsigned char)*p++)
#define EOFCH (fp ? EOF : 0)

  if (raw) {
    if ((ch = NEXTCH) == EOFCH)
      return (EOF);
    for (;;) {
      if (!ch)
	break;
      DPUTC(d, ch);
      if ((ch = NEXTCH) == EOFCH)
	break;
    }
    DPUTZ(d);
    return (0);
  }

  /* --- Skip as far as whitespace --- *
   *
   * Also skip past comments.
   */

again:
  ch = NEXTCH;
  while (isspace(ch))
    ch = NEXTCH;
  if (ch == '#') {
    do ch = NEXTCH; while (ch != '\n' && ch != EOFCH);
    goto again;
  }
  if (ch == EOFCH)
    return (EOF);

  /* --- If the character is a quote then read a quoted string --- */

  switch (ch) {
    case '`':
      ch = '\'';
    case '\'':
    case '\"':
      q = ch;
      ch = NEXTCH;
      break;
  }

  /* --- Now read all sorts of interesting things --- */

  for (;;) {

    /* --- Handle an escaped thing --- */

    if (ch == '\\') {
      ch = NEXTCH;
      if (ch == EOFCH)
	break;
      switch (ch) {
	case 'a': ch = '\a'; break;
	case 'b': ch = '\b'; break;
	case 'f': ch = '\f'; break;
	case 'n': ch = '\n'; break;
	case 'r': ch = '\r'; break;
	case 't': ch = '\t'; break;
	case 'v': ch = '\v'; break;
      }
      DPUTC(d, ch);
      ch = NEXTCH;
      continue;
    }

    /* --- If it's a quote or some other end marker then stop --- */

    if (ch == q)
      break;
    if (!q && isspace(ch))
      break;

    /* --- Otherwise contribute and continue --- */

    DPUTC(d, ch);
    if ((ch = NEXTCH) == EOFCH)
      break;
  }

  /* --- Done --- */

  DPUTZ(d);
  return (0);

#undef NEXTCH
#undef EOFCH
}

/* --- @putstring@ --- *
 *
 * Arguments:	@FILE *fp@ = stream to write on
 *		@const char *p@ = pointer to text
 *		@unsigned raw@ = whether the string is to be written raw
 *
 * Returns:	---
 *
 * Use:		Emits a string to a stream.
 */

static void putstring(FILE *fp, const char *p, unsigned raw)
{
  size_t sz = strlen(p);
  unsigned qq;
  const char *q;

  /* --- Just write the string null terminated if raw --- */

  if (raw) {
    fwrite(p, 1, sz + 1, fp);
    return;
  }

  /* --- Check for any dodgy characters --- */

  qq = 0;
  for (q = p; *q; q++) {
    if (isspace((unsigned char)*q)) {
      qq = '\"';
      break;
    }
  }

  if (qq)
    putc(qq, fp);

  /* --- Emit the string --- */

  for (q = p; *q; q++) {
    switch (*q) {
      case '\a': fputc('\\', fp); fputc('a', fp); break;
      case '\b': fputc('\\', fp); fputc('b', fp); break;
      case '\f': fputc('\\', fp); fputc('f', fp); break;
      case '\n': fputc('\\', fp); fputc('n', fp); break;
      case '\r': fputc('\\', fp); fputc('r', fp); break;
      case '\t': fputc('\\', fp); fputc('t', fp); break;
      case '\v': fputc('\\', fp); fputc('v', fp); break;
      case '`': fputc('\\', fp); fputc('`', fp); break;
      case '\'': fputc('\\', fp); fputc('\'', fp); break;
      case '\"': fputc('\\', fp); fputc('\"', fp); break;
      case '#': fputc('\\', fp); fputc('#', fp); break;
      default:
	putc(*q, fp);
	break;
    }
  }

  /* --- Done --- */

  if (qq)
    putc(qq, fp);
}

/*----- Guts --------------------------------------------------------------*/

static int checkhash(const char *file, unsigned f,
		     const gchash *gch, const encodeops *e)
{
  int rc;
  FILE *fp;
  dstr d = DSTR_INIT;
  dstr dd = DSTR_INIT;
  unsigned long n = 0, nfail = 0;
  octet *buf = xmalloc(2 * gch->hashsz);

  if (!file || strcmp(file, "-") == 0)
    fp = stdin;
  else if ((fp = fopen(file, f & f_raw ? "r" : "rb")) == 0) {
    moan("couldn't open `%s': %s", file, strerror(errno));
    return (EXIT_FAILURE);
  }

  while (DRESET(&d), dstr_putline(&d, fp) != EOF) {
    char *p = d.buf;
    char *q;
    unsigned ff = f;

    /* --- Handle a directive --- */

    if (*p == '#') {
      p++;
      if ((q = str_getword(&p)) == 0)
	continue;
      if (strcmp(q, "hash") == 0) {
	const gchash *g;
	if ((q = str_getword(&p)) == 0)
	  continue;
	if ((g = gethash(q)) == 0)
	  continue;
	gch = g;
	xfree(buf);
	buf = xmalloc(2 * gch->hashsz);
      } else if (strcmp(q, "encoding") == 0) {
	const encodeops *ee;
	if ((q = str_getword(&p)) == 0)
	  continue;
	if ((ee = getencoding(q)) == 0)
	  continue;
	e = ee;
      } else if (strcmp(q, "escape") == 0)
	f |= f_escape;
      continue;
    }

    /* --- Otherwise it's a hex thing --- */

    q = p;
    while (*p && *p != ' ')
      p++;
    if (!*p)
      continue;
    *p++ = 0;
    if (e->get(q, buf, gch->hashsz, 0) < gch->hashsz)
      continue;
    if (*p == '*')
      ff |= f_binary;
    else if (*p != ' ')
      continue;
    p++;

    if (f & f_escape) {
      DRESET(&dd);
      getstring(0, p, &dd, 0);
      p = dd.buf;
    }

    if (fhash(p, ff, gch, buf + gch->hashsz)) {
      moan("couldn't read `%s': %s", p, strerror(errno));
      rc = EXIT_FAILURE;
      continue;
    }
    if (memcmp(buf, buf + gch->hashsz, gch->hashsz) != 0) {
      if (ff & f_verbose)
	fprintf(stderr, "FAIL %s\n", p);
      else
	moan("%s check failed for `%s'", gch->name, p);
      nfail++;
      rc = EXIT_FAILURE;
    } else {
      if (ff & f_verbose)
	fprintf(stderr, "OK %s\n", p);
    }
    n++;
  }

  dstr_destroy(&d);
  dstr_destroy(&dd);
  xfree(buf);
  if ((f & f_verbose) && nfail)
    moan("%lu of %lu file(s) failed %s check", nfail, n, gch->name);
  else if (!n)
    moan("no files checked");
  return (0);
}

static int dohash(const char *file, unsigned f,
		  const gchash *gch, const encodeops *e)
{
  int rc = 0;
  octet *p = xmalloc(gch->hashsz);

  if (fhash(file, f, gch, p)) {
    moan("couldn't read `%s': %s", file ? file : "<stdin>", strerror(errno));
    rc = EXIT_FAILURE;
  } else {
    e->put(p, gch->hashsz, stdout);
    if (file) {
      fputc(' ', stdout);
      fputc(f & f_binary ? '*' : ' ', stdout);
      if (f & f_escape)
	putstring(stdout, file, 0);
      else
	fputs(file, stdout);
    }
    fputc('\n', stdout);
  }

  xfree(p);
  return (rc);
}

static int dofile(const char *file, unsigned f,
		  const gchash *gch, const encodeops *e)
{
  return (f & f_check ? checkhash : dohash)(file, f, gch, e);
}

static int hashfiles(const char *file, unsigned f,
		     const gchash *gch, const encodeops *e)
{
  FILE *fp;
  dstr d = DSTR_INIT;
  int rc = 0;
  int rrc;

  if (!file || strcmp(file, "-") == 0)
    fp = stdin;
  else if ((fp = fopen(file, f & f_raw ? "r" : "rb")) == 0) {
    moan("couldn't open `%s': %s", file, strerror(errno));
    return (EXIT_FAILURE);
  }

  for (;;) {
    DRESET(&d);
    if (getstring(fp, 0, &d, f & f_raw))
      break;
    if ((rrc = dofile(d.buf, f, gch, e)) != 0)
      rc = rrc;
  }

  return (rc);
}

static int hashsum(const char *file, unsigned f,
		   const gchash *gch, const encodeops *e)
{
  return (f & f_files ? hashfiles : dofile)(file, f, gch, e);
}

/*----- Main driver -------------------------------------------------------*/

void version(FILE *fp)
{
  pquis(fp, "$, Catacomb version " VERSION "\n");
}

static void usage(FILE *fp)
{
  pquis(fp, "Usage: $ [-f0ebcv] [-a ALGORITHM] [-E ENC] [FILES...]\n");
}

static void help(FILE *fp, const gchash *gch)
{
  version(fp);
  fputc('\n', fp);
  usage(fp);
  pquis(fp, "\n\
Generates or checks message digests on files.  Options available:\n\
\n\
-h, --help		Display this help message.\n\
-V, --version		Display program's version number.\n\
-u, --usage		Display a terse usage message.\n\
-l, --list [ITEM...]	Show known hash functions and/or encodings.\n\
\n\
-a, --algorithm=ALG	Use the message digest algorithm ALG.\n\
-E, --encoding=ENC	Represent hashes using encoding ENC.\n\
\n\
-f, --files		Read a list of file names from standard input.\n\
-0, --null		File names are null terminated, not plain text.\n\
\n\
-e, --escape		Escape funny characters in filenames.\n\
-c, --check		Check message digests rather than emitting them.\n\
-b, --binary		When reading files, treat them as binary.\n\
-v, --verbose		Be verbose when checking digests.\n\
\n\
For a list of hashing algorithms and encodings, type `$ --list'.\n\
");
  if (gch)
    fprintf(fp, "The default message digest algorithm is %s.\n", gch->name);
}

#define LISTS(LI)							\
  LI("Lists", list, listtab[i].name, listtab[i].name)			\
  LI("Hash functions", hash, ghashtab[i], ghashtab[i]->name)		\
  LI("Encodings", enc, encodingtab[i].name, encodingtab[i].name)

MAKELISTTAB(listtab, LISTS)

int main(int argc, char *argv[])
{
  unsigned f = 0;
  const gchash *gch = 0;
  const encodeops *e = &encodingtab[0];
  int rc;

  /* --- Initialization --- */

  ego(argv[0]);
  sub_init();

  /* --- Choose a hash function from the name --- */

  {
    char *q = xstrdup(QUIS);
    size_t len = strlen(q);
    if (len > 3 && strcmp(q + len - 3, "sum") == 0) {
      q[len - 3] = 0;
      gch = gethash(q);
    }
    if (!gch)
      gch = gethash("md5");
    xfree(q);
  }

  /* --- Read options --- */

  for (;;) {
    static struct option opts[] = {
      { "help",		0,		0,	'h' },
      { "verbose",	0,		0,	'V' },
      { "usage",	0,		0,	'u' },

      { "algorithm",	OPTF_ARGREQ,	0,	'a' },
      { "hash",		OPTF_ARGREQ,	0,	'a' },
      { "encoding",	OPTF_ARGREQ,	0,	'E' },
      { "list",		0,		0,	'l' },

      { "files",	0,		0,	'f' },
      { "find",		0,		0,	'f' },
      { "null",		0,		0,	'0' },

      { "escape",	0,		0,	'e' },
      { "check",	0,		0,	'c' },
      { "binary",	0,		0,	'b' },
      { "verbose",	0,		0,	'v' },
      { "progress",	0,		0,	'p' },

      { 0,		0,		0,	0 }
    };
    int i = mdwopt(argc, argv, "hVu a:E:l f0 ecbvp", opts, 0, 0, 0);
    if (i < 0)
      break;

    switch (i) {
      case 'h':
	help(stdout, gch);
	exit(0);
      case 'V':
	version(stdout);
	exit(0);
      case 'u':
	usage(stdout);
	exit(0);
      case 'l':
	exit(displaylists(listtab, argv + optind));
      case 'a':
	if ((gch = gethash(optarg)) == 0)
	  die(EXIT_FAILURE, "unknown hash algorithm `%s'", optarg);
	f |= f_oddhash;
	break;
      case 'E':
	if ((e = getencoding(optarg)) == 0)
	  die(EXIT_FAILURE, "unknown encoding `%s'", optarg);
	f |= f_oddenc;
	break;
      case 'f':
	f |= f_files;
	break;
      case '0':
	f |= f_raw;
	break;
      case 'e':
	f |= f_escape;
	break;
      case 'c':
	f |= f_check;
	break;
      case 'b':
	f |= f_binary;
	break;
      case 'v':
	f |= f_verbose;
	break;
      case 'p':
	f |= f_progress;
	break;
      default:
	f |= f_bogus;
	break;
    }
  }

  if (f & f_bogus) {
    usage(stderr);
    exit(EXIT_FAILURE);
  }
  argv += optind;
  argc -= optind;

  /* --- Generate output --- */

  if (!(f & f_check) && (argc || (f & f_files))) {
    if (f & f_oddhash) printf("#hash %s\n", gch->name);
    if (f & f_oddenc) printf("#encoding %s\n", e->name);
    if (f & f_escape) fputs("#escape\n", stdout);
  }
  if (!argc)
    rc = hashsum(0, f, gch, e);
  else {
    int i;
    int rrc;

    rc = 0;
    for (i = 0; i < argc; i++) {
      if ((rrc = hashsum(argv[i], f, gch, e)) != 0)
	rc = rrc;
    }
  }

  return (rc);
}

/*----- That's all, folks -------------------------------------------------*/