Deploy the new <ctype.h> and `foocmp' macros from mLib.

[catacomb] / progs / cc-hash.c

/* -*-c-*-
 *
 * Common functions for hashing utilities
 *
 * (c) 2011 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 <ctype.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include <sys/types.h>
#include <dirent.h>

#include <mLib/alloc.h>
#include <mLib/dstr.h>
#include <mLib/macros.h>
#include <mLib/report.h>
#include <mLib/str.h>

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

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

/*----- 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(p[0]) && ISXDIGIT(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 putbase64(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 getbase64(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 putbase32(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 getbase32(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 --- */

const encodeops encodingtab[] = {
#define TAB(tag, name) { #name, put##name, get##name },
  ENCODINGS(TAB)
#undef TAB
  { 0, 0, 0 }
};

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

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

/*----- File hashing ------------------------------------------------------*/

enum {
  FHETY_DIR,
  FHETY_FILE
};

typedef struct fhent {
  struct fhent *next;
  unsigned ty;
  struct fhent *sub;
  char name[1];
} fhdir;

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

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)) {
      if ((*g)->name[sz] == 0) {
	gg = *g;
	break;
      } else if (gg)
	return (0);
      else
	gg = *g;
    }
  }
  return (gg);
}

/* --- @describefile@ --- *
 *
 * Arguments:	@const struct stat *st@ = pointer to file state
 *
 * Returns:	A snappy one-word description of the file.
 */

const char *describefile(const struct stat *st)
{
  switch (st->st_mode & S_IFMT) {
    case S_IFBLK: return ("block-special");
    case S_IFCHR: return ("char-special");
    case S_IFIFO: return ("fifo");
    case S_IFREG: return ("file");
    case S_IFLNK: return ("symlink");
    case S_IFDIR: return ("directory");
    case S_IFSOCK: return ("socket");
    default: return ("unknown");
  }
}

/* --- @fhash_init@ ---*
 *
 * Arguments:	@fhashstate *fh@ = pointer to fhash state to initialize
 *		@const gchash *gch@ = hash class to set
 *		@unsigned f@ initial flags to set
 *
 * Returns:	---
 *
 * Use:		Initializes an @fhashstate@ structure.
 */

void fhash_init(fhashstate *fh, const gchash *gch, unsigned f)
  { fh->f = f; fh->gch = gch; fh->ents = 0; }

/* --- @fhash_free@ --- *
 *
 * Arguments:	@fhashstate *fh@ = pointer to fhash state to free
 *
 * Returns:	---
 *
 * Use:		Frees an fhash state.
 */

static void freefhents(struct fhent *fhe)
{
  struct fhent *ffhe;

  for (; fhe; fhe = ffhe) {
    ffhe = fhe->next;
    freefhents(fhe->sub);
    xfree(fhe);
  }
}

void fhash_free(fhashstate *fh)
  { freefhents(fh->ents); }

/* --- @fhash@ --- *
 *
 * Arguments:	@fhashstate *fh@ = pointer to fhash state
 *		@const char *file@ = file name to be hashed (null for stdin)
 *		@void *buf@ = pointer to hash output buffer
 *
 * Returns:	Zero if it worked, nonzero on error.
 *
 * Use:		Hashes a file.
 */

int fhash(fhashstate *fh, const char *file, void *buf)
{
  FILE *fp;
  char fbuf[1024 * 128];
  size_t sz;
  ghash *h;
  int rc = 0;
  struct fhent *fhe, **ffhe;
  const char *p, *q;
  size_t n;
  fprogress ff;

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

  if (fh->f & FHF_PROGRESS) {
    if (fprogress_init(&ff, file ? file : "<stdin>", fp)) return (-1);
  }

  if (fh->f & FHF_JUNK) {
    p = file;
    if (STRNCMP(p, ==, "./", 2)) p += 2;
    q = p;
    ffhe = &fh->ents;
    for (;;) {
      if (*q == '/' || *q == 0) {
	n = q - p;
	for (; *ffhe; ffhe = &(*ffhe)->next) {
	  fhe = *ffhe;
	  if (STRNCMP(p, ==, fhe->name, n) && fhe->name[n] == 0)
	    goto found;
	}
	fhe = xmalloc(offsetof(struct fhent, name) + n + 1);
	fhe->next = 0;
	fhe->ty = *q == '/' ? FHETY_DIR : FHETY_FILE;
	fhe->sub = 0;
	*ffhe = fhe;
	memcpy(fhe->name, p, n); fhe->name[n] = 0;
      found:
	if (!*q) break;
	while (*++q == '/');
	p = q;
	ffhe = &fhe->sub;
      } else
	q++;
    }
  }

  h = GH_INIT(fh->gch);
  while ((sz = fread(fbuf, 1, sizeof(fbuf), fp)) > 0) {
    GH_HASH(h, fbuf, sz);
    if (fh->f & FHF_PROGRESS) fprogress_update(&ff, sz);
  }
  if (ferror(fp)) rc = -1;
  if (fp != stdin) fclose(fp);
  if (fh->f & FHF_PROGRESS) fprogress_done(&ff);
  GH_DONE(h, buf);
  GH_DESTROY(h);
  return (rc);
}

/* --- @fhash_junk@ --- *
 *
 * Arguments:	@fhashstate *fh@ = pointer to fhash state
 *		@void (*func)(const char *, const struct stat *, void *)@
 *		@void *p@ = pointer to pass to function
 *
 * Returns:	Positive if any junk was found, negative on error, zero if
 *		everything was fine.
 *
 * Use:		Reports junk files in any directories covered by the hash
 *		state.
 */

struct fhjunk {
  int (*func)(const char *, const struct stat *, void *);
  void *p;
  dstr *d;
};

static int fhjunk(struct fhjunk *fhj, struct fhent *ents)
{
  DIR *dp;
  int rc = 0, rrc;
  struct stat st;
  struct dirent *d;
  const char *dname;
  size_t n = fhj->d->len;
  struct fhent *fhe;

  dname = n ? fhj->d->buf : ".";
  if ((dp = opendir(dname)) == 0) {
    moan("failed to open directory `%s': %s", dname, strerror(errno));
    rc = -1;
    goto subs;
  }
  if (n) {
    dstr_putc(fhj->d, '/');
    n++;
  }
  while (errno = 0, (d = readdir(dp)) != 0) {
    if (STRCMP(d->d_name, ==, ".") || STRCMP(d->d_name, ==, ".."))
      continue;
    for (fhe = ents; fhe; fhe = fhe->next) {
      if (STRCMP(d->d_name, ==, fhe->name)) goto found;
    }
    fhj->d->len = n;
    dstr_puts(fhj->d, d->d_name);
    if (!lstat(fhj->d->buf, &st)) {
      if (!rc) rc = 1;
      rrc = fhj->func(fhj->d->buf, &st, fhj->p);
    } else {
      rc = -1;
      rrc = fhj->func(fhj->d->buf, 0, fhj->p);
    }
    if (rrc < 0) rc = -1;
  found:;
  }
  closedir(dp);
  if (errno) {
    moan("failed to read directory `%s': %s", dname, strerror(errno));
    rc = -1;
  }

subs:
  for (fhe = ents; fhe; fhe = fhe->next) {
    if (fhe->ty == FHETY_DIR) {
      fhj->d->len = n;
      dstr_puts(fhj->d, fhe->name);
      rrc = fhjunk(fhj, fhe->sub);
      if (rrc < 0) rc = -1;
      else if (!rc) rc = rrc;
    }
  }

  return (rc);
}

int fhash_junk(fhashstate *fh,
	       int (*func)(const char *, const struct stat *, void *),
	       void *p)
{
  dstr d = DSTR_INIT;
  struct fhjunk fhj;
  int rc;

  fhj.func = func;
  fhj.p = p;
  fhj.d = &d;
  rc = fhjunk(&fhj, fh->ents);
  dstr_destroy(&d);
  return (rc);
}

/* --- @hfparse@ --- *
 *
 * Arguments:	@hfpctx *hfp@ = pointer to the context structure
 *
 * Returns:	A code indicating what happened.
 *
 * Use:		Parses a line from the input file.
 */

int hfparse(hfpctx *hfp)
{
  char *p, *q;
  const gchash *gch;
  const encodeops *ee;
  dstr *d = hfp->dline;
  size_t hsz;

  /* --- Fetch the input line and get ready to parse --- */

  DRESET(d);
  if (dstr_putline(d, hfp->fp) == EOF) return (HF_EOF);
  p = d->buf;

  /* --- Parse magic comments --- */

  if (*p == '#') {
    p++;
    if ((q = str_getword(&p)) == 0) return (HF_BAD);
    if (STRCMP(q, ==, "hash")) {
      if ((q = str_getword(&p)) == 0) return (HF_BAD);
      if ((gch = gethash(q)) == 0) return (HF_BAD);
      hfp->gch = gch;
      return (HF_HASH);
    } else if (STRCMP(q, ==, "encoding")) {
      if ((q = str_getword(&p)) == 0) return (HF_BAD);
      if ((ee = getencoding(q)) == 0) return (HF_BAD);
      hfp->ee = ee;
      return (HF_ENC);
    } else if (STRCMP(q, ==, "escape")) {
      hfp->f |= HFF_ESCAPE;
      return (HF_ESC);
    }
    return (HF_BAD);
  }

  /* --- Otherwise it's a file line --- */

  q = p;
  while (*p && *p != ' ') p++;
  if (!*p) return (HF_BAD);
  *p++ = 0;
  hsz = hfp->gch->hashsz;
  if (hfp->ee->get(q, hfp->hbuf, hsz, 0) < hsz) return (HF_BAD);
  switch (*p) {
    case '*': hfp->f |= FHF_BINARY; break;
    case ' ': hfp->f &= ~FHF_BINARY; break;
    default: return (HF_BAD);
  }
  p++;

  DRESET(hfp->dfile);
  if (hfp->f & HFF_ESCAPE)
    getstring(&p, hfp->dfile, GSF_STRING);
  else {
    dstr_putm(hfp->dfile, p, d->len - (p - d->buf));
    dstr_putz(hfp->dfile);
  }

  return (HF_FILE);
}

/*----- String I/O --------------------------------------------------------*/

/* --- @getstring@ --- *
 *
 * Arguments:	@void *in@ = input source
 *		@dstr *d@ = destination string
 *		@unsigned f@ = input flags
 *
 * Returns:	Zero if OK, nonzero on end-of-file.
 *
 * Use:		Reads a filename (or something similar) from a stream.
 */

static int nextch_file(void *in)
  { FILE *fp = in; return (getc(fp)); }

static int nextch_string(void *in)
  { const unsigned char **p = in; return (*(*p)++); }

int getstring(void *in, dstr *d, unsigned f)
{
  int ch;
  int eofch = (f & GSF_STRING) ? 0 : EOF;
  int (*nextch)(void *) = (f & GSF_STRING) ? nextch_string : nextch_file;
  int q = 0;

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

  if (f & GSF_RAW) {
    if ((ch = nextch(in)) == eofch)
      return (EOF);
    for (;;) {
      if (!ch)
	break;
      DPUTC(d, ch);
      if ((ch = nextch(in)) == eofch)
	break;
    }
    DPUTZ(d);
    return (0);
  }

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

again:
  ch = nextch(in);
  while (ISSPACE(ch))
    ch = nextch(in);
  if (ch == '#') {
    do ch = nextch(in); 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(in);
      break;
  }

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

  for (;;) {

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

    if (ch == '\\') {
      ch = nextch(in);
      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(in);
      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(in)) == eofch)
      break;
  }

  /* --- Done --- */

  DPUTZ(d);
  return (0);
}

/* --- @putstring@ --- *
 *
 * Arguments:	@FILE *fp@ = stream to write on
 *		@const char *p@ = pointer to text
 *		@unsigned f@ = output flags
 *
 * Returns:	---
 *
 * Use:		Emits a string to a stream.
 */

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

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

  if (f & GSF_RAW) {
    fwrite(p, 1, sz + 1, fp);
    return;
  }

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

  qq = 0;
  for (q = p; *q; q++) {
    if (ISSPACE(*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;
      default:
	putc(*q, fp);
	break;
    }
  }

  /* --- Done --- */

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

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