[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
/*
|
|
|
|
passage.c
|
|
|
|
|
|
|
|
Text passage formatting.
|
|
|
|
|
|
|
|
Copyright (C) 2022 Bill Currie <bill@taniwha.org>
|
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or
|
|
|
|
modify it under the terms of the GNU General Public License
|
|
|
|
as published by the Free Software Foundation; either version 2
|
|
|
|
of the License, or (at your option) any later version.
|
|
|
|
|
|
|
|
This program 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 General Public License for more details.
|
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
|
|
along with this program; if not, write to:
|
|
|
|
|
|
|
|
Free Software Foundation, Inc.
|
|
|
|
59 Temple Place - Suite 330
|
|
|
|
Boston, MA 02111-1307, USA
|
|
|
|
|
|
|
|
*/
|
|
|
|
#ifdef HAVE_CONFIG_H
|
|
|
|
# include "config.h"
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef HAVE_STRING_H
|
|
|
|
# include <string.h>
|
|
|
|
#endif
|
|
|
|
#ifdef HAVE_STRINGS_H
|
|
|
|
# include <strings.h>
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#include "QF/alloc.h"
|
|
|
|
#include "QF/qtypes.h"
|
|
|
|
#include "QF/sys.h"
|
|
|
|
|
2022-12-13 04:36:51 +00:00
|
|
|
#include "QF/ecs.h"
|
2022-10-27 10:47:10 +00:00
|
|
|
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
#include "QF/ui/passage.h"
|
2022-10-27 10:47:10 +00:00
|
|
|
#include "QF/ui/view.h"
|
|
|
|
|
2022-12-16 09:12:38 +00:00
|
|
|
const component_t passage_components[passage_comp_count] = {
|
|
|
|
[passage_href] = {
|
|
|
|
.size = sizeof (hierref_t),
|
|
|
|
.name = "passage href",
|
2023-03-05 12:29:26 +00:00
|
|
|
.destroy = Hierref_DestroyComponent,
|
2022-12-16 09:12:38 +00:00
|
|
|
},
|
|
|
|
};
|
|
|
|
|
|
|
|
static const component_t passage_type_components[passage_type_count] = {
|
2022-10-27 10:47:10 +00:00
|
|
|
[passage_type_text_obj] = {
|
|
|
|
.size = sizeof (psg_text_t),
|
|
|
|
.name = "Text",
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
|
|
|
static const hierarchy_type_t passage_type = {
|
|
|
|
.num_components = passage_type_count,
|
2022-12-16 09:12:38 +00:00
|
|
|
.components = passage_type_components,
|
2022-10-27 10:47:10 +00:00
|
|
|
};
|
|
|
|
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
VISIBLE int
|
|
|
|
Passage_IsSpace (const char *text)
|
|
|
|
{
|
|
|
|
if (text[0] == ' ') {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
// 2002;EN SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2003;EM SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2004;THREE-PER-EM SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2005;FOUR-PER-EM SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2006;SIX-PER-EM SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2008;PUNCTUATION SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 2009;THIN SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
// 200A;HAIR SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
if ((byte)text[0] == 0xe2 && (byte)text[1] == 0x80
|
|
|
|
&& ((byte)text[2] >= 0x80 && (byte)text[2] < 0x90
|
|
|
|
&& ((1 << (text[2] & 0xf)) & 0x077c))) {
|
|
|
|
return 3;
|
|
|
|
}
|
|
|
|
// 205F;MEDIUM MATHEMATICAL SPACE;Zs;0;WS;<compat> 0020;;;;N;;;;;
|
|
|
|
if ((byte)text[0] == 0xe2 && (byte)text[1] == 0x81
|
|
|
|
&& (byte)text[2] == 0x9f) {
|
|
|
|
return 3;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2022-10-28 02:43:03 +00:00
|
|
|
static void
|
|
|
|
add_entity (hierarchy_t *h, uint32_t parent)
|
|
|
|
{
|
|
|
|
uint32_t i = Hierarchy_InsertHierarchy (h, 0, parent, 0);
|
|
|
|
h->ent[i] = ECS_NewEntity (h->reg);
|
2022-12-11 15:20:20 +00:00
|
|
|
hierref_t *ref = Ent_AddComponent (h->ent[i], h->href_comp, h->reg);
|
2022-10-28 02:43:03 +00:00
|
|
|
ref->hierarchy = h;
|
|
|
|
ref->index = i;
|
|
|
|
}
|
|
|
|
|
2022-11-04 04:06:20 +00:00
|
|
|
VISIBLE void
|
|
|
|
Passage_ParseText (passage_t *passage, const char *text)
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
{
|
2022-11-04 04:06:20 +00:00
|
|
|
if (passage->hierarchy) {//FIXME just empty hierarchy
|
|
|
|
Hierarchy_Delete (passage->hierarchy);
|
|
|
|
}
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
if (!*text) {
|
2022-11-04 04:06:20 +00:00
|
|
|
return;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
2022-11-04 04:06:20 +00:00
|
|
|
passage->text = text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
|
|
|
|
unsigned num_paragraphs = 1;
|
2022-10-27 10:47:10 +00:00
|
|
|
unsigned num_text_objects = 1;
|
2022-11-04 06:26:01 +00:00
|
|
|
psg_text_t root_text = {};
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
int parsing_space = Passage_IsSpace (text);
|
|
|
|
for (const char *c = text; *c; c++) {
|
|
|
|
int size;
|
|
|
|
if ((size = Passage_IsSpace (c))) {
|
|
|
|
if (!parsing_space) {
|
2022-10-27 10:47:10 +00:00
|
|
|
num_text_objects++;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
parsing_space = 1;
|
|
|
|
c += size - 1;
|
|
|
|
} else if (*c == '\n') {
|
|
|
|
if (c[1]) {
|
|
|
|
num_paragraphs++;
|
2022-10-27 10:47:10 +00:00
|
|
|
num_text_objects += !Passage_IsSpace (c + 1);
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
if (parsing_space) {
|
2022-10-27 10:47:10 +00:00
|
|
|
num_text_objects++;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
parsing_space = 0;
|
|
|
|
}
|
2022-11-04 06:26:01 +00:00
|
|
|
root_text.size = c - text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
2022-12-16 09:12:38 +00:00
|
|
|
passage->hierarchy = Hierarchy_New (passage->reg,
|
|
|
|
passage->comp_base + passage_href,
|
2022-12-11 15:20:20 +00:00
|
|
|
&passage_type, 0);
|
2022-10-27 10:47:10 +00:00
|
|
|
Hierarchy_Reserve (passage->hierarchy,
|
|
|
|
1 + num_paragraphs + num_text_objects);
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
#if 0
|
|
|
|
printf ("num_paragraphs %d, num_text_objects %d\n", num_paragraphs,
|
|
|
|
passage->num_text_objects);
|
|
|
|
#endif
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, nullent);
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
for (unsigned i = 0; i < num_paragraphs; i++) {
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, 0);
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
num_paragraphs = 0;
|
2022-10-27 10:47:10 +00:00
|
|
|
hierarchy_t *h = passage->hierarchy;
|
|
|
|
psg_text_t *passage_obj = h->components[passage_type_text_obj];
|
|
|
|
psg_text_t *par_obj = &passage_obj[h->childIndex[0]];
|
|
|
|
psg_text_t *text_obj = &passage_obj[h->childIndex[1]];
|
|
|
|
*par_obj = *text_obj = (psg_text_t) { };
|
|
|
|
|
2022-11-04 06:26:01 +00:00
|
|
|
*passage_obj = root_text;
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, par_obj - passage_obj);
|
2022-10-27 10:47:10 +00:00
|
|
|
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
parsing_space = Passage_IsSpace (text);
|
|
|
|
for (const char *c = text; *c; c++) {
|
|
|
|
int size;
|
|
|
|
if ((size = Passage_IsSpace (c))) {
|
|
|
|
if (!parsing_space) {
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, par_obj - passage_obj);
|
2022-10-27 10:47:10 +00:00
|
|
|
text_obj->size = c - text - text_obj->text;
|
|
|
|
(++text_obj)->text = c - text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
parsing_space = 1;
|
|
|
|
c += size - 1;
|
|
|
|
} else if (*c == '\n') {
|
2022-10-27 10:47:10 +00:00
|
|
|
text_obj->size = c - text - text_obj->text;
|
|
|
|
par_obj->size = c - text - par_obj->text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
if (c[1]) {
|
2022-10-27 10:47:10 +00:00
|
|
|
(++par_obj)->text = c + 1 - text;
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, par_obj - passage_obj);
|
2022-10-27 10:47:10 +00:00
|
|
|
(++text_obj)->text = c + 1 - text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
parsing_space = Passage_IsSpace (c + 1);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
if (parsing_space) {
|
2022-10-28 02:43:03 +00:00
|
|
|
add_entity (passage->hierarchy, par_obj - passage_obj);
|
2022-10-27 10:47:10 +00:00
|
|
|
text_obj->size = c - text - text_obj->text;
|
|
|
|
(++text_obj)->text = c - text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
parsing_space = 0;
|
|
|
|
if (!c[1]) {
|
2022-10-27 10:47:10 +00:00
|
|
|
text_obj->size = c + 1 - text - text_obj->text;
|
|
|
|
par_obj->size = c + 1 - text - par_obj->text;
|
|
|
|
passage_obj->size = c + 1 - text - passage_obj->text;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#if 0
|
2022-10-28 02:43:03 +00:00
|
|
|
for (uint32_t i = 0; i < h->num_objects; i++) {
|
2022-10-27 10:47:10 +00:00
|
|
|
psg_text_t *to = &passage_obj[i];
|
2022-10-28 02:43:03 +00:00
|
|
|
uint32_t ent = h->ent[i];
|
2022-12-11 15:20:20 +00:00
|
|
|
hierref_t *ref = Ent_GetComponent (ent, h->href_comp, reg);
|
2022-10-28 02:43:03 +00:00
|
|
|
printf ("%3d %8x %3d %4d %4d '%.*s'\n", i, ent, ref->index,
|
2022-10-27 10:47:10 +00:00
|
|
|
to->text, to->size, to->size, text + to->text);
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
}
|
|
|
|
#endif
|
2022-11-04 04:06:20 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
VISIBLE passage_t *
|
2022-12-16 09:12:38 +00:00
|
|
|
Passage_New (ecs_system_t passage_sys)
|
2022-11-04 04:06:20 +00:00
|
|
|
{
|
|
|
|
passage_t *passage = malloc (sizeof (passage_t));
|
|
|
|
passage->text = 0;
|
2022-12-16 09:12:38 +00:00
|
|
|
passage->reg = passage_sys.reg;
|
|
|
|
passage->comp_base = passage_sys.base;
|
2022-11-04 04:06:20 +00:00
|
|
|
passage->hierarchy = 0;
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
return passage;
|
|
|
|
}
|
|
|
|
|
|
|
|
VISIBLE void
|
|
|
|
Passage_Delete (passage_t *passage)
|
|
|
|
{
|
2022-12-09 06:20:10 +00:00
|
|
|
if (passage->hierarchy) {
|
|
|
|
Hierarchy_Delete (passage->hierarchy);
|
|
|
|
}
|
[ui] Add a sub-system for parsing text passages
A passage object has a list of all the text objects in the given string,
where the objects represent either white space or "words", as well as a
view_t object representing the entire passage, with paragraphs split
into child views of the passage view, and each paragraph has a child
view for every text/space object in the paragraph.
Paragraphs are split by '\n' (not included in any object).
White space is grouped into clumps such that multiple adjacent spaces
form a single object. The standard ASCII space (0x20) and all of the
Unicode characters marked "WS;<compat> 0020" are counted as white space.
Unless a white space object is the first in the paragraph, its view is
marked for suppression by the view flow code.
Contiguous non-white space characters are grouped into single objects,
and their views are not suppressed.
All text object views (both white space and "word") have their data
pointer set to the psg_text_t object representing the text for that
view. This should be suitable for simple text-mode unattributed display.
More advanced rendering would probably want to create suitable objects
and set the view data pointers to those objects.
No assumption is made about text direction.
Passage and paragraph views need to have their primary axis sizes set
appropriately, as well as their resize flags. Their xlen and ylen are
both set to 10, and xpos,ypos is 0,0. Paragraph views need their
setgeometry pointer set to the appropriate view_flow_* function.
However, they are set up to have their secondary axis set automatically
when flowed.
Text object views are set up for automatic flowing: grav_flow, 0,0 for
xpos,ypos. However, xlen and ylen are also both 0, so need to be set by
the renderer before attempting to flow the text.
2022-09-30 10:05:18 +00:00
|
|
|
free (passage);
|
|
|
|
}
|