ref.cc

早期freebsd实现

    const token_info *ti = lookup_token(token_start, ptr);
    if (need_period) {
      if ((ptr - token_start == 1 && *token_start == ' ')
	  || (ptr - token_start == 2 && token_start[0] == '\\'
	      && token_start[1] == ' '))
	continue;
      if (ti->is_upper())
	result += period_before_initial;
      else
	result += period_before_other;
      need_period = 0;
    }
    result.append(token_start, ptr - token_start);
    if (ti->is_upper()) {
      const char *lower_ptr = ptr;
      int first_token = 1;
      for (;;) {
	token_start = ptr;
	if (!get_token(&ptr, last_name_start))
	  break;
	if ((ptr - token_start == 1 && *token_start == ' ')
	    || (ptr - token_start == 2 && token_start[0] == '\\'
		&& token_start[1] == ' '))
	  break;
	ti = lookup_token(token_start, ptr);
	if (ti->is_hyphen()) {
	  const char *ptr1 = ptr;
	  if (get_token(&ptr1, last_name_start)) {
	    ti = lookup_token(ptr, ptr1);
	    if (ti->is_upper()) {
	      result += period_before_hyphen;
	      result.append(token_start, ptr1 - token_start);
	      ptr = ptr1;
	    }
	  }
	}
	else if (ti->is_upper()) {
	  // MacDougal -> MacD.
	  result.append(lower_ptr, ptr - lower_ptr);
	  lower_ptr = ptr;
	  first_token = 1;
	}
	else if (first_token && ti->is_accent()) {
	  result.append(token_start, ptr - token_start);
	  lower_ptr = ptr;
	}
	first_token = 0;
      }
      need_period = 1;
    }
  }
  if (need_period)
    result += period_before_last_name;
  result.append(last_name_start, end - last_name_start);
}

static void abbreviate_names(string &result)
{
  string str;
  str.move(result);
  const char *ptr = str.contents();
  const char *end = ptr + str.length();
  while (ptr < end) {
    const char *name_end = (char *)memchr(ptr, FIELD_SEPARATOR, end - ptr);
    if (name_end == 0)
      name_end = end;
    abbreviate_name(ptr, name_end, result);
    if (name_end >= end)
      break;
    ptr = name_end + 1;
    result += FIELD_SEPARATOR;
  }
}

void reverse_name(const char *ptr, const char *name_end, string &result)
{
  const char *last_name_end;
  const char *last_name_start = find_last_name(ptr, name_end, &last_name_end);
  result.append(last_name_start, last_name_end - last_name_start);
  while (last_name_start > ptr
	 && (last_name_start[-1] == ' ' || last_name_start[-1] == '\n'))
    last_name_start--;
  if (last_name_start > ptr) {
    result += ", ";
    result.append(ptr, last_name_start - ptr);
  }
  if (last_name_end < name_end)
    result.append(last_name_end, name_end - last_name_end);
}

void reverse_names(string &result, int n)
{
  if (n <= 0)
    return;
  string str;
  str.move(result);
  const char *ptr = str.contents();
  const char *end = ptr + str.length();
  while (ptr < end) {
    if (--n < 0) {
      result.append(ptr, end - ptr);
      break;
    }
    const char *name_end = (char *)memchr(ptr, FIELD_SEPARATOR, end - ptr);
    if (name_end == 0)
      name_end = end;
    reverse_name(ptr, name_end, result);
    if (name_end >= end)
      break;
    ptr = name_end + 1;
    result += FIELD_SEPARATOR;
  }
}

// Return number of field separators.

int join_fields(string &f)
{
  const char *ptr = f.contents();
  int len = f.length();
  int nfield_seps = 0;
  for (int j = 0; j < len; j++)
    if (ptr[j] == FIELD_SEPARATOR)
      nfield_seps++;
  if (nfield_seps == 0)
    return 0;
  string temp;
  int field_seps_left = nfield_seps;
  for (j = 0; j < len; j++) {
    if (ptr[j] == FIELD_SEPARATOR) {
      if (nfield_seps == 1)
	temp += join_authors_exactly_two;
      else if (--field_seps_left == 0)
	temp += join_authors_last_two;
      else
	temp += join_authors_default;
    }
    else
      temp += ptr[j];
  }
  f = temp;
  return nfield_seps;
}

void uppercase(const char *start, const char *end, string &result)
{
  for (;;) {
    const char *token_start = start;
    if (!get_token(&start, end))
      break;
    const token_info *ti = lookup_token(token_start, start);
    ti->upper_case(token_start, start, result);
  }
}

void lowercase(const char *start, const char *end, string &result)
{
  for (;;) {
    const char *token_start = start;
    if (!get_token(&start, end))
      break;
    const token_info *ti = lookup_token(token_start, start);
    ti->lower_case(token_start, start, result);
  }
}

void capitalize(const char *ptr, const char *end, string &result)
{
  int in_small_point_size = 0;
  for (;;) {
    const char *start = ptr;
    if (!get_token(&ptr, end))
      break;
    const token_info *ti = lookup_token(start, ptr);
    const char *char_end = ptr;
    int is_lower = ti->is_lower();
    if ((is_lower || ti->is_upper()) && get_token(&ptr, end)) {
      const token_info *ti2 = lookup_token(char_end, ptr);
      if (!ti2->is_accent())
	ptr = char_end;
    }
    if (is_lower) {
      if (!in_small_point_size) {
	result += "\\s-2";
	in_small_point_size = 1;
      }
      ti->upper_case(start, char_end, result);
      result.append(char_end, ptr - char_end);
    }
    else {
      if (in_small_point_size) {
	result += "\\s+2";
	in_small_point_size = 0;
      }
      result.append(start, ptr - start);
    }
  }
  if (in_small_point_size)
    result += "\\s+2";
}

void capitalize_field(string &str)
{
  string temp;
  capitalize(str.contents(), str.contents() + str.length(), temp);
  str.move(temp);
}

int is_terminated(const char *ptr, const char *end)
{
  const char *last_token = end;
  for (;;) {
    const char *p = ptr;
    if (!get_token(&ptr, end))
      break;
    last_token = p;
  }
  return end - last_token == 1
    && (*last_token == '.' || *last_token == '!' || *last_token == '?');
}

void reference::output(FILE *fp)
{
  fputs(".]-\n", fp);
  for (int i = 0; i < 256; i++)
    if (field_index[i] != NULL_FIELD_INDEX && i != annotation_field) {
      string &f = field[field_index[i]];
      if (!csdigit(i)) {
	int j = reverse_fields.search(i);
	if (j >= 0) {
	  int n;
	  int len = reverse_fields.length();
	  if (++j < len && csdigit(reverse_fields[j])) {
	    n = reverse_fields[j] - '0';
	    for (++j; j < len && csdigit(reverse_fields[j]); j++)
	      // should check for overflow
	      n = n*10 + reverse_fields[j] - '0';
	  }
	  else 
	    n = INT_MAX;
	  reverse_names(f, n);
	}
      }
      int is_multiple = join_fields(f) > 0;
      if (capitalize_fields.search(i) >= 0)
	capitalize_field(f);
      if (memchr(f.contents(), '\n', f.length()) == 0) {
	fprintf(fp, ".ds [%c ", i);
	if (f[0] == ' ' || f[0] == '\\' || f[0] == '"')
	  putc('"', fp);
	put_string(f, fp);
	putc('\n', fp);
      }
      else {
	fprintf(fp, ".de [%c\n", i);
	put_string(f, fp);
	fputs("..\n", fp);
      }
      if (i == 'P') {
	int multiple_pages = 0;
	if (f.length() > 0 && memchr(f.contents(), '-', f.length()) != 0)
	  multiple_pages = 1;
	fprintf(fp, ".nr [P %d\n", multiple_pages);
      }
      else if (i == 'E')
	fprintf(fp, ".nr [E %d\n", is_multiple);
    }
  for (const char *p = "TAO"; *p; p++) {
    int fi = field_index[(unsigned char)*p];
    if (fi != NULL_FIELD_INDEX) {
      string &f = field[fi];
      fprintf(fp, ".nr [%c %d\n", *p,
	      is_terminated(f.contents(), f.contents() + f.length()));
    }
  }
  int t = classify();
  fprintf(fp, ".][ %d %s\n", t, reference_types[t]);
  if (annotation_macro.length() > 0 && annotation_field >= 0
      && field_index[annotation_field] != NULL_FIELD_INDEX) {
    putc('.', fp);
    put_string(annotation_macro, fp);
    putc('\n', fp);
    put_string(field[field_index[annotation_field]], fp);
  }
}

void reference::print_sort_key_comment(FILE *fp)
{
  fputs(".\\\"", fp);
  put_string(sort_key, fp);
  putc('\n', fp);
}

const char *find_year(const char *start, const char *end, const char **endp)
{
  for (;;) {
    while (start < end && !csdigit(*start))
      start++;
    const char *ptr = start;
    if (start == end)
      break;
    while (ptr < end && csdigit(*ptr))
      ptr++;
    if (ptr - start == 4 || ptr - start == 3
	|| (ptr - start == 2
	    && (start[0] >= '4' || (start[0] == '3' && start[1] >= '2')))) {
      *endp = ptr;
      return start;
    }
    start = ptr;
  }
  return 0;
}

static const char *find_day(const char *start, const char *end,
			    const char **endp)
{
  for (;;) {
    while (start < end && !csdigit(*start))
      start++;
    const char *ptr = start;
    if (start == end)
      break;
    while (ptr < end && csdigit(*ptr))
      ptr++;
    if ((ptr - start == 1 && start[0] != '0')
	|| (ptr - start == 2 &&
	    (start[0] == '1'
	     || start[0] == '2'
	     || (start[0] == '3' && start[1] <= '1')
	     || (start[0] == '0' && start[1] != '0')))) {
      *endp = ptr;
      return start;
    }
    start = ptr;
  }
  return 0;
}

static int find_month(const char *start, const char *end)
{
  static const char *months[] = {
    "january",
    "february",
    "march",
    "april",
    "may",
    "june",
    "july",
    "august",
    "september",
    "october",
    "november",
    "december",
  };
  for (;;) {
    while (start < end && !csalpha(*start))
      start++;
    const char *ptr = start;
    if (start == end)
      break;
    while (ptr < end && csalpha(*ptr))
      ptr++;
    if (ptr - start >= 3) {
      for (int i = 0; i < sizeof(months)/sizeof(months[0]); i++) {
	const char *q = months[i];
	const char *p = start;
	for (; p < ptr; p++, q++)
	  if (cmlower(*p) != *q)
	    break;
	if (p >= ptr)
	  return i;
      }
    }
    start = ptr;
  }
  return -1;
}

int reference::contains_field(char c) const
{
  return field_index[(unsigned char)c] != NULL_FIELD_INDEX;
}

int reference::classify()
{
  if (contains_field('J'))
    return JOURNAL_ARTICLE;
  if (contains_field('B'))
    return ARTICLE_IN_BOOK;
  if (contains_field('G'))
    return TECH_REPORT;
  if (contains_field('R'))
    return TECH_REPORT;
  if (contains_field('I'))
    return BOOK;
  if (contains_field('M'))
    return BELL_TM;
  return OTHER;
}

const char *reference::get_year(const char **endp) const
{
  if (field_index['D'] != NULL_FIELD_INDEX) {
    string &date = field[field_index['D']];
    const char *start = date.contents();
    const char *end = start + date.length();
    return find_year(start, end, endp);
  }
  else
    return 0;
}

const char *reference::get_field(unsigned char c, const char **endp) const
{
  if (field_index[c] != NULL_FIELD_INDEX) {
    string &f = field[field_index[c]];
    const char *start = f.contents();
    *endp = start + f.length();
    return start;
  }
  else
    return 0;
}

const char *reference::get_date(const char **endp) const
{
  return get_field('D', endp);
}

const char *nth_field(int i, const char *start, const char **endp)
{
  while (--i >= 0) {
    start = (char *)memchr(start, FIELD_SEPARATOR, *endp - start);
    if (!start)
      return 0;
    start++;
  }
  const char *e = (char *)memchr(start, FIELD_SEPARATOR, *endp - start);
  if (e)
    *endp = e;
  return start;
}

const char *reference::get_author(int i, const char **endp) const
{
  for (const char *f = AUTHOR_FIELDS; *f != '\0'; f++) {
    const char *start = get_field(*f, endp);
    if (start) {
      if (strchr(MULTI_FIELD_NAMES, *f) != 0)
	return nth_field(i, start, endp);
      else if (i == 0)
	return start;
      else
	return 0;
    }
  }
  return 0;
}

const char *reference::get_author_last_name(int i, const char **endp) const
{
  for (const char *f = AUTHOR_FIELDS; *f != '\0'; f++) {
    const char *start = get_field(*f, endp);
    if (start) {
      if (strchr(MULTI_FIELD_NAMES, *f) != 0) {
	start = nth_field(i, start, endp);
	if (!start)
	  return 0;
      }
      if (*f == 'A')
	return find_last_name(start, *endp, endp);
      else
	return start;
    }
  }
  return 0;
}

void reference::set_date(string &d)
{
  if (d.length() == 0)
    delete_field('D');
  else
    insert_field('D', d);
}

int same_year(const reference &r1, const reference &r2)
{
  const char *ye1;
  const char *ys1 = r1.get_year(&ye1);
  const char *ye2;
  const char *ys2 = r2.get_year(&ye2);
  if (ys1 == 0) {
    if (ys2 == 0)
      return same_date(r1, r2);
    else
      return 0;
  }
  else if (ys2 == 0)
    return 0;
  else if (ye1 - ys1 != ye2 - ys2)
    return 0;
  else
    return memcmp(ys1, ys2, ye1 - ys1) == 0;
}

int same_date(const reference &r1, const reference &r2)
{
  const char *e1;
  const char *s1 = r1.get_date(&e1);
  const char *e2;
  const char *s2 = r2.get_date(&e2);
  if (s1 == 0)
    return s2 == 0;
  else if (s2 == 0)
    return 0;
  else if (e1 - s1 != e2 - s2)
    return 0;
  else
    return memcmp(s1, s2, e1 - s1) == 0;
}

const char *reference::get_sort_field(int i, int si, int ssi,
				      const char **endp) const
{
  const char *start = sort_key.contents();
  const char *end = start + sort_key.length();
  if (i < 0) {
    *endp = end;
    return start;
  }
  while (--i >= 0) {
    start = (char *)memchr(start, SORT_SEP, end - start);
    if (!start)
      return 0;
    start++;
  }
  const char *e = (char *)memchr(start, SORT_SEP, end - start);
  if (e)
    end = e;
  if (si < 0) {
    *endp = end;
    return start;
  }
  while (--si >= 0) {
    start = (char *)memchr(start, SORT_SUB_SEP, end - start);
    if (!start)
      return 0;
    start++;
  }
  e = (char *)memchr(start, SORT_SUB_SEP, end - start);
  if (e)
    end = e;
  if (ssi < 0) {
    *endp = end;
    return start;
  }
  while (--ssi >= 0) {
    start = (char *)memchr(start, SORT_SUB_SUB_SEP, end - start);
    if (!start)
      return 0;
    start++;
  }
  e = (char *)memchr(start, SORT_SUB_SUB_SEP, end - start);
  if (e)
    end = e;
  *endp = end;
  return start;
}