Reducing duplicate code and simplifying well formed string copying

by adding a new class String_copier. This is a pre-requisite patch for MDEV-6566 and MDEV-6572, to avoid adding more similar code.

Author: Alexander Barkov

sql/field.cc

@@ -6262,10 +6262,9 @@ bool Field_datetimef::get_date(MYSQL_TIME *ltime, ulonglong fuzzydate)
 
  SYNOPSIS
  check_string_copy_error()
- field - Field
- well_formed_error_pos - where not well formed data was first met
- cannot_convert_error_pos - where a not-convertable character was first met
- end - end of the string
+ copier - the conversion status
+ end - the very end of the source string
+ that was just copied
  cs - character set of the string
 
  NOTES
@@ -6282,30 +6281,25 @@ bool Field_datetimef::get_date(MYSQL_TIME *ltime, ulonglong fuzzydate)
  TRUE - If an error happened
 */
 
-static bool
-check_string_copy_error(Field_str *field,
- const char *well_formed_error_pos,
- const char *cannot_convert_error_pos,
- const char *end,
- CHARSET_INFO *cs)
+bool
+Field_longstr::check_string_copy_error(const String_copier *copier,
+ const char *end,
+ CHARSET_INFO *cs)
 {
  const char *pos;
  char tmp[32];
- THD *thd;
 
- thd= field->get_thd();
-
- if (!(pos= well_formed_error_pos) &&
- !(pos= cannot_convert_error_pos))
+ if (!(pos= copier->most_important_error_pos()))
  return FALSE;
 
  convert_to_printable(tmp, sizeof(tmp), pos, (end - pos), cs, 6);
 
+ THD *thd= get_thd();
  push_warning_printf(thd,
  Sql_condition::WARN_LEVEL_WARN,
  ER_TRUNCATED_WRONG_VALUE_FOR_FIELD,
  ER(ER_TRUNCATED_WRONG_VALUE_FOR_FIELD),
- "string", tmp, field->field_name,
+ "string", tmp, field_name,
  thd->get_stmt_da()->current_row_for_warning());
  return TRUE;
 }
@@ -6362,32 +6356,23 @@ int Field_string::store(const char *from,uint length,CHARSET_INFO *cs)
 {
  ASSERT_COLUMN_MARKED_FOR_WRITE_OR_COMPUTED;
  uint copy_length;
- const char *well_formed_error_pos;
- const char *cannot_convert_error_pos;
- const char *from_end_pos;
+ String_copier copier;
 
  /* See the comment for Field_long::store(long long) */
  DBUG_ASSERT(!table || table->in_use == current_thd);
 
- copy_length= well_formed_copy_nchars(field_charset,
+ copy_length= copier.well_formed_copy(field_charset,
  (char*) ptr, field_length,
  cs, from, length,
- field_length / field_charset->mbmaxlen,
- &well_formed_error_pos,
- &cannot_convert_error_pos,
- &from_end_pos);
+ field_length / field_charset->mbmaxlen);
 
  /* Append spaces if the string was shorter than the field. */
  if (copy_length < field_length)
  field_charset->cset->fill(field_charset,(char*) ptr+copy_length,
  field_length-copy_length,
  field_charset->pad_char);
 
- if (check_string_copy_error(this, well_formed_error_pos,
- cannot_convert_error_pos, from + length, cs))
- return 2;
-
- return report_if_important_data(from_end_pos, from + length, FALSE);
+ return check_conversion_status(&copier, from + length, cs, false);
 }
 
 
@@ -6874,29 +6859,19 @@ int Field_varstring::store(const char *from,uint length,CHARSET_INFO *cs)
 {
  ASSERT_COLUMN_MARKED_FOR_WRITE_OR_COMPUTED;
  uint copy_length;
- const char *well_formed_error_pos;
- const char *cannot_convert_error_pos;
- const char *from_end_pos;
+ String_copier copier;
 
- copy_length= well_formed_copy_nchars(field_charset,
+ copy_length= copier.well_formed_copy(field_charset,
  (char*) ptr + length_bytes,
  field_length,
  cs, from, length,
- field_length / field_charset->mbmaxlen,
- &well_formed_error_pos,
- &cannot_convert_error_pos,
- &from_end_pos);
-
+ field_length / field_charset->mbmaxlen);
  if (length_bytes == 1)
  *ptr= (uchar) copy_length;
  else
  int2store(ptr, copy_length);
 
- if (check_string_copy_error(this, well_formed_error_pos,
- cannot_convert_error_pos, from + length, cs))
- return 2;
-
- return report_if_important_data(from_end_pos, from + length, TRUE);
+ return check_conversion_status(&copier, from + length, cs, true);
 }
 
 
@@ -7351,9 +7326,8 @@ int Field_blob::store(const char *from,uint length,CHARSET_INFO *cs)
 {
  ASSERT_COLUMN_MARKED_FOR_WRITE_OR_COMPUTED;
  uint copy_length, new_length;
- const char *well_formed_error_pos;
- const char *cannot_convert_error_pos;
- const char *from_end_pos, *tmp;
+ String_copier copier;
+ const char *tmp;
  char buff[STRING_BUFFER_USUAL_SIZE];
  String tmpstr(buff,sizeof(buff), &my_charset_bin);
 
@@ -7401,28 +7375,14 @@ int Field_blob::store(const char *from,uint length,CHARSET_INFO *cs)
  bmove(ptr + packlength, (uchar*) &tmp, sizeof(char*));
  return 0;
  }
- /*
- "length" is OK as "nchars" argument to well_formed_copy_nchars as this
- is never used to limit the length of the data. The cut of long data
- is done with the new_length value.
- */
- copy_length= well_formed_copy_nchars(field_charset,
+ copy_length= copier.well_formed_copy(field_charset,
  (char*) value.ptr(), new_length,
- cs, from, length,
- length,
- &well_formed_error_pos,
- &cannot_convert_error_pos,
- &from_end_pos);
-
+ cs, from, length);
  Field_blob::store_length(copy_length);
  tmp= value.ptr();
  bmove(ptr+packlength,(uchar*) &tmp,sizeof(char*));
 
- if (check_string_copy_error(this, well_formed_error_pos,
- cannot_convert_error_pos, from + length, cs))
- return 2;
-
- return report_if_important_data(from_end_pos, from + length, TRUE);
+ return check_conversion_status(&copier, from + length, cs, true);
 
 oom_error:
  /* Fatal OOM error */

sql/field.h

@@ -1132,6 +1132,17 @@ class Field_longstr :public Field_str
 protected:
  int report_if_important_data(const char *ptr, const char *end,
  bool count_spaces);
+ bool check_string_copy_error(const String_copier *copier,
+ const char *end, CHARSET_INFO *cs);
+ int check_conversion_status(const String_copier *copier,
+ const char *end, CHARSET_INFO *cs,
+ bool count_spaces)
+ {
+ if (check_string_copy_error(copier, end, cs))
+ return 2;
+ return report_if_important_data(copier->source_end_pos(),
+ end, count_spaces);
+ }
 public:
  Field_longstr(uchar *ptr_arg, uint32 len_arg, uchar *null_ptr_arg,
  uchar null_bit_arg, utype unireg_check_arg,

sql/sql_class.cc

@@ -2986,9 +2986,7 @@ int select_export::send_data(List<Item> &items)
  if (res && !my_charset_same(write_cs, res->charset()) &&
  !my_charset_same(write_cs, &my_charset_bin))
  {
- const char *well_formed_error_pos;
- const char *cannot_convert_error_pos;
- const char *from_end_pos;
+ String_copier copier;
  const char *error_pos;
  uint32 bytes;
  uint64 estimated_bytes=
@@ -3001,16 +2999,11 @@ int select_export::send_data(List<Item> &items)
  goto err;
  }
 
- bytes= well_formed_copy_nchars(write_cs, (char *) cvt_str.ptr(),
+ bytes= copier.well_formed_copy(write_cs, (char *) cvt_str.ptr(),
  cvt_str.alloced_length(),
- res->charset(), res->ptr(), res->length(),
- UINT_MAX32, // copy all input chars,
- // i.e. ignore nchars parameter
- &well_formed_error_pos,
- &cannot_convert_error_pos,
- &from_end_pos);
- error_pos= well_formed_error_pos ? well_formed_error_pos
- : cannot_convert_error_pos;
+ res->charset(),
+ res->ptr(), res->length());
+ error_pos= copier.most_important_error_pos();
  if (error_pos)
  {
  char printable_buff[32];
@@ -3023,7 +3016,7 @@ int select_export::send_data(List<Item> &items)
  "string", printable_buff,
  item->name, static_cast<long>(row_count));
  }
- else if (from_end_pos < res->ptr() + res->length())
+ else if (copier.source_end_pos() < res->ptr() + res->length())
  { 
  /*
  result is longer than UINT_MAX32 and doesn't fit into String

sql/sql_signal.cc

@@ -195,16 +195,9 @@ static bool assign_fixed_string(MEM_ROOT *mem_root,
  dst_str= (char*) alloc_root(mem_root, dst_len + 1);
  if (dst_str)
  {
- const char* well_formed_error_pos;
- const char* cannot_convert_error_pos;
- const char* from_end_pos;
-
- dst_copied= well_formed_copy_nchars(dst_cs, dst_str, dst_len,
- src_cs, src_str, src_len,
- numchars,
- & well_formed_error_pos,
- & cannot_convert_error_pos,
- & from_end_pos);
+ dst_copied= String_copier().well_formed_copy(dst_cs, dst_str, dst_len,
+ src_cs, src_str, src_len,
+ numchars);
  DBUG_ASSERT(dst_copied <= dst_len);
  dst_len= dst_copied; /* In case the copy truncated the data */
  dst_str[dst_copied]= '\0';