[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[no subject]
|
From: |
Gavin D. Smith |
|
Date: |
Fri, 20 Oct 2023 16:14:47 -0400 (EDT) |
branch: master
commit 49c596d717347c03e1d06df2d5ec05c92c94af11
Author: Gavin Smith <gavinsmith0123@gmail.com>
AuthorDate: Fri Oct 20 20:56:40 2023 +0100
* tp/Texinfo/XS/xspara.c (xspara_add_text):
Keep track of both the type of current (block of) characters to
proess and the type of the next (block of) characters.
Still only operate on one character at a time for now.
---
ChangeLog | 7 +++
tp/Texinfo/XS/xspara.c | 144 ++++++++++++++++++++++++++++++++-----------------
2 files changed, 101 insertions(+), 50 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index 682711c978..753c714283 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,10 @@
+2023-10-20 Gavin Smith <gavinsmith0123@gmail.com>
+
+ * tp/Texinfo/XS/xspara.c (xspara_add_text):
+ Keep track of both the type of current (block of) characters to
+ proess and the type of the next (block of) characters.
+ Still only operate on one character at a time for now.
+
2023-10-20 Gavin Smith <gavinsmith0123@gmail.com>
* tp/Texinfo/XS/xspara.c (xspara_add_text): Rearrange code
diff --git a/tp/Texinfo/XS/xspara.c b/tp/Texinfo/XS/xspara.c
index d15d59f546..58901d12a8 100644
--- a/tp/Texinfo/XS/xspara.c
+++ b/tp/Texinfo/XS/xspara.c
@@ -882,7 +882,8 @@ xspara_set_space_protection (int no_break,
/*****************************************************************/
enum text_class { type_NULL, type_spaces, type_regular,
- type_double_width, type_unknown };
+ type_double_width, type_EOS, type_finished,
+ type_unknown };
/* Return string to be added to paragraph contents, wrapping text. This
function relies on there being a UTF-8 locale in LC_CTYPE for mbrtowc to
@@ -890,12 +891,12 @@ enum text_class { type_NULL, type_spaces, type_regular,
TEXT
xspara_add_text (char *text, int len)
{
- char *p = text;
- wchar_t wc;
- size_t char_len;
+ char *p = text, *q = 0;
+ wchar_t wc, wc_fw;
+ size_t next_len = 0;
int width;
static TEXT result;
- enum text_class type = type_NULL;
+ enum text_class type = type_NULL, next_type = type_NULL;
dTHX;
@@ -903,7 +904,7 @@ xspara_add_text (char *text, int len)
state.end_line_count = 0;
- while (len > 0)
+ while (1)
{
if (debug)
{
@@ -915,47 +916,93 @@ xspara_add_text (char *text, int len)
state.word.end > 0 ? state.word.text : "UNDEF");
}
- /* Get the type of the next character. Set wc and char_len
- if it is not a space. */
- if (isspace ((unsigned char) *p))
- {
- type = type_spaces;
- }
- else
+ /* p is now at the beginning of the text we have left to process.
+ next_type is set to the type of the next block, or is null. */
+
+ type = next_type;
+
+ q = p;
+ q += next_len; len -= next_len; /* Skip over the last character
+ processed. */
+
+ /* Set q to the end of the block. Set next_len to the length of
+ the following character, and next_type to the type of
+ the block after. */
+ while (1)
{
- /* Set wc and char_len */
- if (!PRINTABLE_ASCII(*p))
+ if (len <= 0)
{
- char_len = mbrtowc (&wc, p, len, NULL);
+ next_type = type_finished;
}
- else
+ else if (isspace ((unsigned char) *q))
{
- /* Functonally the same as mbrtowc but (tested) slightly
- quicker. */
- char_len = 1;
- wc = btowc (*p);
+ next_type = type_spaces;
+ next_len = 1;
}
-
- if ((long) char_len == 0)
- break; /* Null character. Shouldn't happen. */
- else if ((long) char_len < 0)
+ else if (*q == '\b')
{
- p++; len--; /* Invalid. Just try to keep going. */
- continue;
+ /* Code to say that a following full stop (or question or
+ exclamation mark) may be an end of sentence. */
+ next_type = type_EOS;
+ next_len = 1;
}
-
- /* Note: width == 0 includes accent characters which should not
- properly increase the column count. This is not what the pure
- Perl code does, though. */
- width = wcwidth (wc);
- if (width == 1 || width == 0)
- type = type_regular;
- else if (width == 2)
- type = type_double_width;
else
- type = type_unknown;
+ {
+ /* Set wc and next_len */
+ if (!PRINTABLE_ASCII(*q))
+ {
+ next_len = mbrtowc (&wc, q, len, NULL);
+ }
+ else
+ {
+ /* Functionally the same as mbrtowc but (tested) slightly
+ quicker. */
+ next_len = 1;
+ wc = btowc (*q);
+ }
+
+ if ((long) next_len == 0)
+ break; /* Null character. Shouldn't happen. */
+ else if ((long) next_len < 0)
+ {
+ q++; len--; /* Invalid. Just try to keep going. */
+ continue;
+ }
+
+ /* Note: width == 0 includes accent characters which should not
+ properly increase the column count. This is not what the pure
+ Perl code does, though. */
+ width = wcwidth (wc);
+ if (width == 1 || width == 0)
+ next_type = type_regular;
+ else if (width == 2)
+ {
+ next_type = type_double_width;
+ wc_fw = wc; /* final value is used below */
+ }
+ else
+ next_type = type_unknown;
+ }
+
+ /* TODO: test just one character at a time to start. then
+ we can gradually work on the various blocks of
+ code to operate on multiple characters. */
+ if (1 || next_type != type || next_type == type_finished)
+ break;
+
+ q += next_len; len -= next_len;
}
+ /* For the very start of the string. */
+ if (type == type_NULL)
+ continue;
+
+ /* Now type is the type of the block we are about to operate on, and
+ next_type the one after it. p is the beginning of the span and q
+ is the end. */
+
+ if (type == type_finished)
+ break;
/*************** Whitespace character. *********************/
if (type == type_spaces)
{
@@ -1055,23 +1102,21 @@ xspara_add_text (char *text, int len)
xspara__end_line ();
text_append (&result, "\n");
}
- p++; len--;
state.last_letter = ' ';
- continue;
}
/*************** Double width character. *********************/
- if (type == type_double_width)
+ else if (type == type_double_width)
{
if (debug)
fprintf (stderr, "FULLWIDTH\n");
- text_append_n (&state.word, p, char_len);
+ text_append_n (&state.word, p, q - p);
state.word_counter += 2;
/* fullwidth latin letters can be upper case, so it is important to
use the actual characters here. */
- state.last_letter = wc;
+ state.last_letter = wc_fw;
/* We allow a line break in between Chinese characters even if
there was no space between them, unlike single-width
@@ -1090,20 +1135,18 @@ xspara_add_text (char *text, int len)
}
state.end_sentence = -2;
}
- else if (wc == L'\b')
+ else if (type == type_EOS)
{
- /* Code to say that a following full stop (or question or
- exclamation mark) may be an end of sentence. */
xspara_allow_end_sentence ();
}
/*************** Word character ******************************/
else if (type == type_regular)
{
static char added_word[8]; /* long enough for one UTF-8 character */
- memcpy (added_word, p, char_len);
- added_word[char_len] = '\0';
+ memcpy (added_word, p, q - p);
+ added_word[q - p] = '\0';
- xspara__add_next (&result, added_word, char_len, 0);
+ xspara__add_next (&result, added_word, q - p, 0);
/* Now check if it is considered as an end of sentence, and
set state.end_sentence if it is. */
@@ -1142,9 +1185,10 @@ xspara_add_text (char *text, int len)
/* Not printable, possibly a tab, or a combining character.
Add it to the pending word without increasing the column
count. */
- text_append_n (&state.word, p, char_len);
+ text_append_n (&state.word, p, q - p);
}
- p += char_len; len -= char_len;
+
+ p = q;
}
return result;