ben
/
mwparserfromhell
mirror of https://github.com/earwig/mwparserfromhell
1
1
Fork 0
Code Releases 24 Activity
Browse Source

Split up C tokenizer into tag_data, tok_parse, tok_support, tokens.

tags/v0.4.1
Ben Kurtovic 9 years ago
parent
0e547aa416
commit
2005efd309
13 changed files with 3722 additions and 3368 deletions
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. +54
    -2
      mwparserfromhell/parser/ctokenizer/common.h
  2. +104
    -0
      mwparserfromhell/parser/ctokenizer/contexts.h
  3. +88
    -0
      mwparserfromhell/parser/ctokenizer/tag_data.c
  4. +43
    -0
      mwparserfromhell/parser/ctokenizer/tag_data.h
  5. +2
    -10
      mwparserfromhell/parser/ctokenizer/textbuffer.h
  6. +2750
    -0
      mwparserfromhell/parser/ctokenizer/tok_parse.c
  7. +29
    -0
      mwparserfromhell/parser/ctokenizer/tok_parse.h
  8. +362
    -0
      mwparserfromhell/parser/ctokenizer/tok_support.c
  9. +66
    -0
      mwparserfromhell/parser/ctokenizer/tok_support.h
  10. +40
    -3122
      mwparserfromhell/parser/ctokenizer/tokenizer.c
  11. +4
    -234
      mwparserfromhell/parser/ctokenizer/tokenizer.h
  12. +111
    -0
      mwparserfromhell/parser/ctokenizer/tokens.c
  13. +69
    -0
      mwparserfromhell/parser/ctokenizer/tokens.h

+ 54
- 2
mwparserfromhell/parser/ctokenizer/common.h View File

@@ -20,14 +20,18 @@ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

#ifndef PY_SSIZE_T_CLEAN
#define PY_SSIZE_T_CLEAN
#define PY_SSIZE_T_CLEAN // See: https://docs.python.org/2/c-api/arg.html
#endif

#include <Python.h>
#include <structmember.h>
#include <bytesobject.h>

/* Compatibility macros */

#if PY_MAJOR_VERSION >= 3
#define IS_PY3K
#endif
@@ -36,5 +40,53 @@ SOFTWARE.
#define uint64_t unsigned PY_LONG_LONG
#endif

#define malloc PyObject_Malloc
#define malloc PyObject_Malloc // XXX: yuck
#define free PyObject_Free

/* Error handling globals/macros */

extern int route_state; // TODO: this is NOT thread-safe!
extern uint64_t route_context;

#define BAD_ROUTE route_state
#define BAD_ROUTE_CONTEXT route_context
#define FAIL_ROUTE(context) { route_state = 1; route_context = context; }
#define RESET_ROUTE() route_state = 0

/* Shared globals */

extern char** entitydefs;

extern PyObject* EMPTY;
extern PyObject* NOARGS;
extern PyObject* definitions;

/* Structs */

struct Textbuffer {
Py_ssize_t size;
Py_UNICODE* data;
struct Textbuffer* prev;
struct Textbuffer* next;
};
typedef struct Textbuffer Textbuffer;

struct Stack {
PyObject* stack;
uint64_t context;
struct Textbuffer* textbuffer;
struct Stack* next;
};
typedef struct Stack Stack;

typedef struct {
PyObject_HEAD
PyObject* text; /* text to tokenize */
Stack* topstack; /* topmost stack */
Py_ssize_t head; /* current position in text */
Py_ssize_t length; /* length of text */
int global; /* global context */
int depth; /* stack recursion depth */
int cycles; /* total number of stack recursions */
int skip_style_tags; /* temporary fix for the sometimes broken tag parser */
} Tokenizer;

+ 104
- 0
mwparserfromhell/parser/ctokenizer/contexts.h View File

@@ -0,0 +1,104 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

/* Local contexts */

#define LC_TEMPLATE 0x0000000000000007
#define LC_TEMPLATE_NAME 0x0000000000000001
#define LC_TEMPLATE_PARAM_KEY 0x0000000000000002
#define LC_TEMPLATE_PARAM_VALUE 0x0000000000000004

#define LC_ARGUMENT 0x0000000000000018
#define LC_ARGUMENT_NAME 0x0000000000000008
#define LC_ARGUMENT_DEFAULT 0x0000000000000010

#define LC_WIKILINK 0x0000000000000060
#define LC_WIKILINK_TITLE 0x0000000000000020
#define LC_WIKILINK_TEXT 0x0000000000000040

#define LC_EXT_LINK 0x0000000000000180
#define LC_EXT_LINK_URI 0x0000000000000080
#define LC_EXT_LINK_TITLE 0x0000000000000100

#define LC_HEADING 0x0000000000007E00
#define LC_HEADING_LEVEL_1 0x0000000000000200
#define LC_HEADING_LEVEL_2 0x0000000000000400
#define LC_HEADING_LEVEL_3 0x0000000000000800
#define LC_HEADING_LEVEL_4 0x0000000000001000
#define LC_HEADING_LEVEL_5 0x0000000000002000
#define LC_HEADING_LEVEL_6 0x0000000000004000

#define LC_TAG 0x0000000000078000
#define LC_TAG_OPEN 0x0000000000008000
#define LC_TAG_ATTR 0x0000000000010000
#define LC_TAG_BODY 0x0000000000020000
#define LC_TAG_CLOSE 0x0000000000040000

#define LC_STYLE 0x0000000000780000
#define LC_STYLE_ITALICS 0x0000000000080000
#define LC_STYLE_BOLD 0x0000000000100000
#define LC_STYLE_PASS_AGAIN 0x0000000000200000
#define LC_STYLE_SECOND_PASS 0x0000000000400000

#define LC_DLTERM 0x0000000000800000

#define LC_SAFETY_CHECK 0x000000003F000000
#define LC_HAS_TEXT 0x0000000001000000
#define LC_FAIL_ON_TEXT 0x0000000002000000
#define LC_FAIL_NEXT 0x0000000004000000
#define LC_FAIL_ON_LBRACE 0x0000000008000000
#define LC_FAIL_ON_RBRACE 0x0000000010000000
#define LC_FAIL_ON_EQUALS 0x0000000020000000

#define LC_TABLE 0x0000000FC0000000
#define LC_TABLE_CELL_LINE_CONTEXTS 0x0000000D00000000
#define LC_TABLE_OPEN 0x0000000040000000
#define LC_TABLE_CELL_OPEN 0x0000000080000000
#define LC_TABLE_CELL_STYLE 0x0000000100000000
#define LC_TABLE_ROW_OPEN 0x0000000200000000
#define LC_TABLE_TD_LINE 0x0000000400000000
#define LC_TABLE_TH_LINE 0x0000000800000000

/* Global contexts */

#define GL_HEADING 0x1

/* Aggregate contexts */

#define AGG_FAIL (LC_TEMPLATE | LC_ARGUMENT | LC_WIKILINK | LC_EXT_LINK_TITLE | LC_HEADING | LC_TAG | LC_STYLE | LC_TABLE_OPEN)
#define AGG_UNSAFE (LC_TEMPLATE_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK_TITLE | LC_TEMPLATE_PARAM_KEY | LC_ARGUMENT_NAME)
#define AGG_DOUBLE (LC_TEMPLATE_PARAM_KEY | LC_TAG_CLOSE | LC_TABLE_ROW_OPEN)
#define AGG_NO_WIKILINKS (LC_TEMPLATE_NAME | LC_ARGUMENT_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK_URI)
#define AGG_NO_EXT_LINKS (LC_TEMPLATE_NAME | LC_ARGUMENT_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK)

/* Tag contexts */

#define TAG_NAME 0x01
#define TAG_ATTR_READY 0x02
#define TAG_ATTR_NAME 0x04
#define TAG_ATTR_VALUE 0x08
#define TAG_QUOTED 0x10
#define TAG_NOTE_SPACE 0x20
#define TAG_NOTE_EQUALS 0x40
#define TAG_NOTE_QUOTE 0x80

+ 88
- 0
mwparserfromhell/parser/ctokenizer/tag_data.c View File

@@ -0,0 +1,88 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#include "tag_data.h"
#include "contexts.h"

/*
Initialize a new TagData object.
*/
TagData* TagData_new(void)
{
#define ALLOC_BUFFER(name) \
name = Textbuffer_new(); \
if (!name) { \
TagData_dealloc(self); \
return NULL; \
}

TagData *self = malloc(sizeof(TagData));
if (!self) {
PyErr_NoMemory();
return NULL;
}
self->context = TAG_NAME;
ALLOC_BUFFER(self->pad_first)
ALLOC_BUFFER(self->pad_before_eq)
ALLOC_BUFFER(self->pad_after_eq)
self->quoter = 0;
self->reset = 0;
return self;

#undef ALLOC_BUFFER
}

/*
Deallocate the given TagData object.
*/
void TagData_dealloc(TagData* self)
{
#define DEALLOC_BUFFER(name) \
if (name) \
Textbuffer_dealloc(name);

DEALLOC_BUFFER(self->pad_first);
DEALLOC_BUFFER(self->pad_before_eq);
DEALLOC_BUFFER(self->pad_after_eq);
free(self);

#undef DEALLOC_BUFFER
}

/*
Clear the internal buffers of the given TagData object.
*/
int TagData_reset_buffers(TagData* self)
{
#define RESET_BUFFER(name) \
Textbuffer_dealloc(name); \
name = Textbuffer_new(); \
if (!name) \
return -1;

RESET_BUFFER(self->pad_first)
RESET_BUFFER(self->pad_before_eq)
RESET_BUFFER(self->pad_after_eq)
return 0;

#undef RESET_BUFFER
}

+ 43
- 0
mwparserfromhell/parser/ctokenizer/tag_data.h View File

@@ -0,0 +1,43 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

#include "common.h"
#include "textbuffer.h"

/* Structs */

typedef struct {
uint64_t context;
Textbuffer* pad_first;
Textbuffer* pad_before_eq;
Textbuffer* pad_after_eq;
Py_UNICODE quoter;
Py_ssize_t reset;
} TagData;

/* Functions */

TagData* TagData_new(void);
void TagData_dealloc(TagData*);
int TagData_reset_buffers(TagData*);

+ 2
- 10
mwparserfromhell/parser/ctokenizer/textbuffer.h View File

@@ -20,17 +20,9 @@ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#include "common.h"

/* Structs */
#pragma once

struct Textbuffer {
Py_ssize_t size;
Py_UNICODE* data;
struct Textbuffer* prev;
struct Textbuffer* next;
};
typedef struct Textbuffer Textbuffer;
#include "common.h"

/* Functions */



+ 2750
- 0
mwparserfromhell/parser/ctokenizer/tok_parse.c
File diff suppressed because it is too large
View File


+ 29
- 0
mwparserfromhell/parser/ctokenizer/tok_parse.h View File

@@ -0,0 +1,29 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

#include "common.h"

/* Functions */

PyObject* Tokenizer_parse(Tokenizer*, uint64_t, int);

+ 362
- 0
mwparserfromhell/parser/ctokenizer/tok_support.c View File

@@ -0,0 +1,362 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#include "tok_support.h"
#include "textbuffer.h"
#include "tokens.h"

/*
Add a new token stack, context, and textbuffer to the list.
*/
int Tokenizer_push(Tokenizer* self, uint64_t context)
{
Stack* top = malloc(sizeof(Stack));

if (!top) {
PyErr_NoMemory();
return -1;
}
top->stack = PyList_New(0);
top->context = context;
top->textbuffer = Textbuffer_new();
if (!top->textbuffer)
return -1;
top->next = self->topstack;
self->topstack = top;
self->depth++;
self->cycles++;
return 0;
}

/*
Push the textbuffer onto the stack as a Text node and clear it.
*/
int Tokenizer_push_textbuffer(Tokenizer* self)
{
PyObject *text, *kwargs, *token;
Textbuffer* buffer = self->topstack->textbuffer;

if (buffer->size == 0 && !buffer->next)
return 0;
text = Textbuffer_render(buffer);
if (!text)
return -1;
kwargs = PyDict_New();
if (!kwargs) {
Py_DECREF(text);
return -1;
}
PyDict_SetItemString(kwargs, "text", text);
Py_DECREF(text);
token = PyObject_Call(Text, NOARGS, kwargs);
Py_DECREF(kwargs);
if (!token)
return -1;
if (PyList_Append(self->topstack->stack, token)) {
Py_DECREF(token);
return -1;
}
Py_DECREF(token);
Textbuffer_dealloc(buffer);
self->topstack->textbuffer = Textbuffer_new();
if (!self->topstack->textbuffer)
return -1;
return 0;
}

/*
Pop and deallocate the top token stack/context/textbuffer.
*/
void Tokenizer_delete_top_of_stack(Tokenizer* self)
{
Stack* top = self->topstack;

Py_DECREF(top->stack);
Textbuffer_dealloc(top->textbuffer);
self->topstack = top->next;
free(top);
self->depth--;
}

/*
Pop the current stack/context/textbuffer, returing the stack.
*/
PyObject* Tokenizer_pop(Tokenizer* self)
{
PyObject* stack;

if (Tokenizer_push_textbuffer(self))
return NULL;
stack = self->topstack->stack;
Py_INCREF(stack);
Tokenizer_delete_top_of_stack(self);
return stack;
}

/*
Pop the current stack/context/textbuffer, returing the stack. We will also
replace the underlying stack's context with the current stack's.
*/
PyObject* Tokenizer_pop_keeping_context(Tokenizer* self)
{
PyObject* stack;
uint64_t context;

if (Tokenizer_push_textbuffer(self))
return NULL;
stack = self->topstack->stack;
Py_INCREF(stack);
context = self->topstack->context;
Tokenizer_delete_top_of_stack(self);
self->topstack->context = context;
return stack;
}

/*
Fail the current tokenization route. Discards the current
stack/context/textbuffer and sets the BAD_ROUTE flag.
*/
void* Tokenizer_fail_route(Tokenizer* self)
{
uint64_t context = self->topstack->context;
PyObject* stack = Tokenizer_pop(self);

Py_XDECREF(stack);
FAIL_ROUTE(context);
return NULL;
}

/*
Write a token to the current token stack.
*/
int Tokenizer_emit_token(Tokenizer* self, PyObject* token, int first)
{
PyObject* instance;

if (Tokenizer_push_textbuffer(self))
return -1;
instance = PyObject_CallObject(token, NULL);
if (!instance)
return -1;
if (first ? PyList_Insert(self->topstack->stack, 0, instance) :
PyList_Append(self->topstack->stack, instance)) {
Py_DECREF(instance);
return -1;
}
Py_DECREF(instance);
return 0;
}

/*
Write a token to the current token stack, with kwargs. Steals a reference
to kwargs.
*/
int Tokenizer_emit_token_kwargs(Tokenizer* self, PyObject* token,
PyObject* kwargs, int first)
{
PyObject* instance;

if (Tokenizer_push_textbuffer(self)) {
Py_DECREF(kwargs);
return -1;
}
instance = PyObject_Call(token, NOARGS, kwargs);
if (!instance) {
Py_DECREF(kwargs);
return -1;
}
if (first ? PyList_Insert(self->topstack->stack, 0, instance):
PyList_Append(self->topstack->stack, instance)) {
Py_DECREF(instance);
Py_DECREF(kwargs);
return -1;
}
Py_DECREF(instance);
Py_DECREF(kwargs);
return 0;
}

/*
Write a Unicode codepoint to the current textbuffer.
*/
int Tokenizer_emit_char(Tokenizer* self, Py_UNICODE code)
{
return Textbuffer_write(&(self->topstack->textbuffer), code);
}

/*
Write a string of text to the current textbuffer.
*/
int Tokenizer_emit_text(Tokenizer* self, const char* text)
{
int i = 0;

while (text[i]) {
if (Tokenizer_emit_char(self, text[i]))
return -1;
i++;
}
return 0;
}

/*
Write the contents of another textbuffer to the current textbuffer,
deallocating it in the process.
*/
int
Tokenizer_emit_textbuffer(Tokenizer* self, Textbuffer* buffer, int reverse)
{
Textbuffer *original = buffer;
Py_ssize_t i;

if (reverse) {
do {
for (i = buffer->size - 1; i >= 0; i--) {
if (Tokenizer_emit_char(self, buffer->data[i])) {
Textbuffer_dealloc(original);
return -1;
}
}
} while ((buffer = buffer->next));
}
else {
while (buffer->next)
buffer = buffer->next;
do {
for (i = 0; i < buffer->size; i++) {
if (Tokenizer_emit_char(self, buffer->data[i])) {
Textbuffer_dealloc(original);
return -1;
}
}
} while ((buffer = buffer->prev));
}
Textbuffer_dealloc(original);
return 0;
}

/*
Write a series of tokens to the current stack at once.
*/
int Tokenizer_emit_all(Tokenizer* self, PyObject* tokenlist)
{
int pushed = 0;
PyObject *stack, *token, *left, *right, *text;
Textbuffer* buffer;
Py_ssize_t size;

if (PyList_GET_SIZE(tokenlist) > 0) {
token = PyList_GET_ITEM(tokenlist, 0);
switch (PyObject_IsInstance(token, Text)) {
case 0:
break;
case 1: {
pushed = 1;
buffer = self->topstack->textbuffer;
if (buffer->size == 0 && !buffer->next)
break;
left = Textbuffer_render(buffer);
if (!left)
return -1;
right = PyObject_GetAttrString(token, "text");
if (!right)
return -1;
text = PyUnicode_Concat(left, right);
Py_DECREF(left);
Py_DECREF(right);
if (!text)
return -1;
if (PyObject_SetAttrString(token, "text", text)) {
Py_DECREF(text);
return -1;
}
Py_DECREF(text);
Textbuffer_dealloc(buffer);
self->topstack->textbuffer = Textbuffer_new();
if (!self->topstack->textbuffer)
return -1;
break;
}
case -1:
return -1;
}
}
if (!pushed) {
if (Tokenizer_push_textbuffer(self))
return -1;
}
stack = self->topstack->stack;
size = PyList_GET_SIZE(stack);
if (PyList_SetSlice(stack, size, size, tokenlist))
return -1;
return 0;
}

/*
Pop the current stack, write text, and then write the stack. 'text' is a
NULL-terminated array of chars.
*/
int Tokenizer_emit_text_then_stack(Tokenizer* self, const char* text)
{
PyObject* stack = Tokenizer_pop(self);

if (Tokenizer_emit_text(self, text)) {
Py_DECREF(stack);
return -1;
}
if (stack) {
if (PyList_GET_SIZE(stack) > 0) {
if (Tokenizer_emit_all(self, stack)) {
Py_DECREF(stack);
return -1;
}
}
Py_DECREF(stack);
}
self->head--;
return 0;
}

/*
Read the value at a relative point in the wikicode, forwards.
*/
PyObject* Tokenizer_read(Tokenizer* self, Py_ssize_t delta)
{
Py_ssize_t index = self->head + delta;

if (index >= self->length)
return EMPTY;
return PyList_GET_ITEM(self->text, index);
}

/*
Read the value at a relative point in the wikicode, backwards.
*/
PyObject* Tokenizer_read_backwards(Tokenizer* self, Py_ssize_t delta)
{
Py_ssize_t index;

if (delta > self->head)
return EMPTY;
index = self->head - delta;
return PyList_GET_ITEM(self->text, index);
}

+ 66
- 0
mwparserfromhell/parser/ctokenizer/tok_support.h View File

@@ -0,0 +1,66 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

#include "common.h"

/* Functions */

int Tokenizer_push(Tokenizer*, uint64_t);
int Tokenizer_push_textbuffer(Tokenizer*);
void Tokenizer_delete_top_of_stack(Tokenizer*);
PyObject* Tokenizer_pop(Tokenizer*);
PyObject* Tokenizer_pop_keeping_context(Tokenizer*);
void* Tokenizer_fail_route(Tokenizer*);

int Tokenizer_emit_token(Tokenizer*, PyObject*, int);
int Tokenizer_emit_token_kwargs(Tokenizer*, PyObject*, PyObject*, int);
int Tokenizer_emit_char(Tokenizer*, Py_UNICODE);
int Tokenizer_emit_text(Tokenizer*, const char*);
int Tokenizer_emit_textbuffer(Tokenizer*, Textbuffer*, int);
int Tokenizer_emit_all(Tokenizer*, PyObject*);
int Tokenizer_emit_text_then_stack(Tokenizer*, const char*);

PyObject* Tokenizer_read(Tokenizer*, Py_ssize_t);
PyObject* Tokenizer_read_backwards(Tokenizer*, Py_ssize_t);

/* Macros */

#define MAX_DEPTH 40
#define MAX_CYCLES 100000

#define Tokenizer_READ(self, delta) \
(*PyUnicode_AS_UNICODE(Tokenizer_read(self, delta)))
#define Tokenizer_READ_BACKWARDS(self, delta) \
(*PyUnicode_AS_UNICODE(Tokenizer_read_backwards(self, delta)))
#define Tokenizer_CAN_RECURSE(self) \
(self->depth < MAX_DEPTH && self->cycles < MAX_CYCLES)

#define Tokenizer_emit(self, token) \
Tokenizer_emit_token(self, token, 0)
#define Tokenizer_emit_first(self, token) \
Tokenizer_emit_token(self, token, 1)
#define Tokenizer_emit_kwargs(self, token, kwargs) \
Tokenizer_emit_token_kwargs(self, token, kwargs, 0)
#define Tokenizer_emit_first_kwargs(self, token, kwargs) \
Tokenizer_emit_token_kwargs(self, token, kwargs, 1)

+ 40
- 3122
mwparserfromhell/parser/ctokenizer/tokenizer.c
File diff suppressed because it is too large
View File


+ 4
- 234
mwparserfromhell/parser/ctokenizer/tokenizer.h View File

@@ -20,256 +20,27 @@ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#include <math.h>
#pragma once

#include "common.h"
#include "textbuffer.h"

#define DIGITS "0123456789"
#define HEXDIGITS "0123456789abcdefABCDEF"
#define ALPHANUM "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"

static const char MARKERS[] = {
'{', '}', '[', ']', '<', '>', '|', '=', '&', '\'', '#', '*', ';', ':', '/',
'-', '!', '\n', '\0'};

#define NUM_MARKERS 19
#define MAX_DEPTH 40
#define MAX_CYCLES 100000
#define MAX_BRACES 255
#define MAX_ENTITY_SIZE 8

static int route_state = 0;
static uint64_t route_context = 0;
#define BAD_ROUTE route_state
#define BAD_ROUTE_CONTEXT route_context
#define FAIL_ROUTE(context) route_state = 1; route_context = context
#define RESET_ROUTE() route_state = 0

static char** entitydefs;

static PyObject* EMPTY;
static PyObject* NOARGS;
static PyObject* ParserError;
static PyObject* definitions;


/* Tokens: */

static PyObject* Text;

static PyObject* TemplateOpen;
static PyObject* TemplateParamSeparator;
static PyObject* TemplateParamEquals;
static PyObject* TemplateClose;

static PyObject* ArgumentOpen;
static PyObject* ArgumentSeparator;
static PyObject* ArgumentClose;

static PyObject* WikilinkOpen;
static PyObject* WikilinkSeparator;
static PyObject* WikilinkClose;

static PyObject* ExternalLinkOpen;
static PyObject* ExternalLinkSeparator;
static PyObject* ExternalLinkClose;

static PyObject* HTMLEntityStart;
static PyObject* HTMLEntityNumeric;
static PyObject* HTMLEntityHex;
static PyObject* HTMLEntityEnd;
static PyObject* HeadingStart;
static PyObject* HeadingEnd;

static PyObject* CommentStart;
static PyObject* CommentEnd;

static PyObject* TagOpenOpen;
static PyObject* TagAttrStart;
static PyObject* TagAttrEquals;
static PyObject* TagAttrQuote;
static PyObject* TagCloseOpen;
static PyObject* TagCloseSelfclose;
static PyObject* TagOpenClose;
static PyObject* TagCloseClose;


/* Local contexts: */

#define LC_TEMPLATE 0x0000000000000007
#define LC_TEMPLATE_NAME 0x0000000000000001
#define LC_TEMPLATE_PARAM_KEY 0x0000000000000002
#define LC_TEMPLATE_PARAM_VALUE 0x0000000000000004

#define LC_ARGUMENT 0x0000000000000018
#define LC_ARGUMENT_NAME 0x0000000000000008
#define LC_ARGUMENT_DEFAULT 0x0000000000000010

#define LC_WIKILINK 0x0000000000000060
#define LC_WIKILINK_TITLE 0x0000000000000020
#define LC_WIKILINK_TEXT 0x0000000000000040

#define LC_EXT_LINK 0x0000000000000180
#define LC_EXT_LINK_URI 0x0000000000000080
#define LC_EXT_LINK_TITLE 0x0000000000000100

#define LC_HEADING 0x0000000000007E00
#define LC_HEADING_LEVEL_1 0x0000000000000200
#define LC_HEADING_LEVEL_2 0x0000000000000400
#define LC_HEADING_LEVEL_3 0x0000000000000800
#define LC_HEADING_LEVEL_4 0x0000000000001000
#define LC_HEADING_LEVEL_5 0x0000000000002000
#define LC_HEADING_LEVEL_6 0x0000000000004000

#define LC_TAG 0x0000000000078000
#define LC_TAG_OPEN 0x0000000000008000
#define LC_TAG_ATTR 0x0000000000010000
#define LC_TAG_BODY 0x0000000000020000
#define LC_TAG_CLOSE 0x0000000000040000

#define LC_STYLE 0x0000000000780000
#define LC_STYLE_ITALICS 0x0000000000080000
#define LC_STYLE_BOLD 0x0000000000100000
#define LC_STYLE_PASS_AGAIN 0x0000000000200000
#define LC_STYLE_SECOND_PASS 0x0000000000400000

#define LC_DLTERM 0x0000000000800000

#define LC_SAFETY_CHECK 0x000000003F000000
#define LC_HAS_TEXT 0x0000000001000000
#define LC_FAIL_ON_TEXT 0x0000000002000000
#define LC_FAIL_NEXT 0x0000000004000000
#define LC_FAIL_ON_LBRACE 0x0000000008000000
#define LC_FAIL_ON_RBRACE 0x0000000010000000
#define LC_FAIL_ON_EQUALS 0x0000000020000000

#define LC_TABLE 0x0000000FC0000000
#define LC_TABLE_CELL_LINE_CONTEXTS 0x0000000D00000000
#define LC_TABLE_OPEN 0x0000000040000000
#define LC_TABLE_CELL_OPEN 0x0000000080000000
#define LC_TABLE_CELL_STYLE 0x0000000100000000
#define LC_TABLE_ROW_OPEN 0x0000000200000000
#define LC_TABLE_TD_LINE 0x0000000400000000
#define LC_TABLE_TH_LINE 0x0000000800000000

/* Global contexts: */

#define GL_HEADING 0x1

/* Aggregate contexts: */

#define AGG_FAIL (LC_TEMPLATE | LC_ARGUMENT | LC_WIKILINK | LC_EXT_LINK_TITLE | LC_HEADING | LC_TAG | LC_STYLE | LC_TABLE_OPEN)
#define AGG_UNSAFE (LC_TEMPLATE_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK_TITLE | LC_TEMPLATE_PARAM_KEY | LC_ARGUMENT_NAME)
#define AGG_DOUBLE (LC_TEMPLATE_PARAM_KEY | LC_TAG_CLOSE | LC_TABLE_ROW_OPEN)
#define AGG_NO_WIKILINKS (LC_TEMPLATE_NAME | LC_ARGUMENT_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK_URI)
#define AGG_NO_EXT_LINKS (LC_TEMPLATE_NAME | LC_ARGUMENT_NAME | LC_WIKILINK_TITLE | LC_EXT_LINK)

/* Tag contexts: */

#define TAG_NAME 0x01
#define TAG_ATTR_READY 0x02
#define TAG_ATTR_NAME 0x04
#define TAG_ATTR_VALUE 0x08
#define TAG_QUOTED 0x10
#define TAG_NOTE_SPACE 0x20
#define TAG_NOTE_EQUALS 0x40
#define TAG_NOTE_QUOTE 0x80


/* Miscellaneous structs: */

struct Stack {
PyObject* stack;
uint64_t context;
struct Textbuffer* textbuffer;
struct Stack* next;
};

typedef struct {
PyObject* title;
int level;
} HeadingData;

typedef struct {
uint64_t context;
struct Textbuffer* pad_first;
struct Textbuffer* pad_before_eq;
struct Textbuffer* pad_after_eq;
Py_UNICODE quoter;
Py_ssize_t reset;
} TagData;

typedef struct Stack Stack;


/* Tokenizer object definition: */

typedef struct {
PyObject_HEAD
PyObject* text; /* text to tokenize */
Stack* topstack; /* topmost stack */
Py_ssize_t head; /* current position in text */
Py_ssize_t length; /* length of text */
int global; /* global context */
int depth; /* stack recursion depth */
int cycles; /* total number of stack recursions */
int skip_style_tags; /* temporary fix for the sometimes broken tag parser */
} Tokenizer;


/* Macros related to Tokenizer functions: */

#define Tokenizer_READ(self, delta) (*PyUnicode_AS_UNICODE(Tokenizer_read(self, delta)))
#define Tokenizer_READ_BACKWARDS(self, delta) \
(*PyUnicode_AS_UNICODE(Tokenizer_read_backwards(self, delta)))
#define Tokenizer_CAN_RECURSE(self) (self->depth < MAX_DEPTH && self->cycles < MAX_CYCLES)

#define Tokenizer_emit(self, token) Tokenizer_emit_token(self, token, 0)
#define Tokenizer_emit_first(self, token) Tokenizer_emit_token(self, token, 1)
#define Tokenizer_emit_kwargs(self, token, kwargs) Tokenizer_emit_token_kwargs(self, token, kwargs, 0)
#define Tokenizer_emit_first_kwargs(self, token, kwargs) Tokenizer_emit_token_kwargs(self, token, kwargs, 1)


/* Macros for accessing definitions: */

#define GET_HTML_TAG(markup) (markup == ':' ? "dd" : markup == ';' ? "dt" : "li")
#define IS_PARSABLE(tag) (call_def_func("is_parsable", tag, NULL, NULL))
#define IS_SINGLE(tag) (call_def_func("is_single", tag, NULL, NULL))
#define IS_SINGLE_ONLY(tag) (call_def_func("is_single_only", tag, NULL, NULL))
#define IS_SCHEME(scheme, slashes, reverse) \
(call_def_func("is_scheme", scheme, slashes ? Py_True : Py_False, reverse ? Py_True : Py_False))


/* Function prototypes: */

static TagData* TagData_new(void);
static void TagData_dealloc(TagData*);
/* Functions */

static PyObject* Tokenizer_new(PyTypeObject*, PyObject*, PyObject*);
static void Tokenizer_dealloc(Tokenizer*);
static int Tokenizer_init(Tokenizer*, PyObject*, PyObject*);
static int Tokenizer_parse_entity(Tokenizer*);
static int Tokenizer_parse_comment(Tokenizer*);
static int Tokenizer_handle_dl_term(Tokenizer*);
static int Tokenizer_parse_tag(Tokenizer*);
static PyObject* Tokenizer_parse(Tokenizer*, uint64_t, int);
static PyObject* Tokenizer_tokenize(Tokenizer*, PyObject*);

static int load_exceptions(void);


/* Macros for Python 2/3 compatibility: */
/* Compatibility macros */

#ifdef IS_PY3K
#define NEW_INT_FUNC PyLong_FromSsize_t
#define IMPORT_NAME_FUNC PyUnicode_FromString
#define CREATE_MODULE PyModule_Create(&module_def);
#define ENTITYDEFS_MODULE "html.entities"
#define INIT_FUNC_NAME PyInit__tokenizer
#define INIT_ERROR return NULL
#else
#define NEW_INT_FUNC PyInt_FromSsize_t
#define IMPORT_NAME_FUNC PyBytes_FromString
#define CREATE_MODULE Py_InitModule("_tokenizer", NULL);
#define ENTITYDEFS_MODULE "htmlentitydefs"
@@ -277,8 +48,7 @@ static int load_exceptions(void);
#define INIT_ERROR return
#endif


/* More structs for creating the Tokenizer type: */
/* Structs */

static PyMethodDef Tokenizer_methods[] = {
{"tokenize", (PyCFunction) Tokenizer_tokenize, METH_VARARGS,


+ 111
- 0
mwparserfromhell/parser/ctokenizer/tokens.c View File

@@ -0,0 +1,111 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#include "tokens.h"

/* Globals */

PyObject* Text;

PyObject* TemplateOpen;
PyObject* TemplateParamSeparator;
PyObject* TemplateParamEquals;
PyObject* TemplateClose;

PyObject* ArgumentOpen;
PyObject* ArgumentSeparator;
PyObject* ArgumentClose;

PyObject* WikilinkOpen;
PyObject* WikilinkSeparator;
PyObject* WikilinkClose;

PyObject* ExternalLinkOpen;
PyObject* ExternalLinkSeparator;
PyObject* ExternalLinkClose;

PyObject* HTMLEntityStart;
PyObject* HTMLEntityNumeric;
PyObject* HTMLEntityHex;
PyObject* HTMLEntityEnd;
PyObject* HeadingStart;
PyObject* HeadingEnd;

PyObject* CommentStart;
PyObject* CommentEnd;

PyObject* TagOpenOpen;
PyObject* TagAttrStart;
PyObject* TagAttrEquals;
PyObject* TagAttrQuote;
PyObject* TagCloseOpen;
PyObject* TagCloseSelfclose;
PyObject* TagOpenClose;
PyObject* TagCloseClose;

/*
Load individual tokens into globals from the given Python module object.
*/
void load_tokens_from_module(PyObject* module)
{
Text = PyObject_GetAttrString(module, "Text");

TemplateOpen = PyObject_GetAttrString(module, "TemplateOpen");
TemplateParamSeparator = PyObject_GetAttrString(module,
"TemplateParamSeparator");
TemplateParamEquals = PyObject_GetAttrString(module,
"TemplateParamEquals");
TemplateClose = PyObject_GetAttrString(module, "TemplateClose");

ArgumentOpen = PyObject_GetAttrString(module, "ArgumentOpen");
ArgumentSeparator = PyObject_GetAttrString(module, "ArgumentSeparator");
ArgumentClose = PyObject_GetAttrString(module, "ArgumentClose");

WikilinkOpen = PyObject_GetAttrString(module, "WikilinkOpen");
WikilinkSeparator = PyObject_GetAttrString(module, "WikilinkSeparator");
WikilinkClose = PyObject_GetAttrString(module, "WikilinkClose");

ExternalLinkOpen = PyObject_GetAttrString(module, "ExternalLinkOpen");
ExternalLinkSeparator = PyObject_GetAttrString(module,
"ExternalLinkSeparator");
ExternalLinkClose = PyObject_GetAttrString(module, "ExternalLinkClose");

HTMLEntityStart = PyObject_GetAttrString(module, "HTMLEntityStart");
HTMLEntityNumeric = PyObject_GetAttrString(module, "HTMLEntityNumeric");
HTMLEntityHex = PyObject_GetAttrString(module, "HTMLEntityHex");
HTMLEntityEnd = PyObject_GetAttrString(module, "HTMLEntityEnd");

HeadingStart = PyObject_GetAttrString(module, "HeadingStart");
HeadingEnd = PyObject_GetAttrString(module, "HeadingEnd");

CommentStart = PyObject_GetAttrString(module, "CommentStart");
CommentEnd = PyObject_GetAttrString(module, "CommentEnd");

TagOpenOpen = PyObject_GetAttrString(module, "TagOpenOpen");
TagAttrStart = PyObject_GetAttrString(module, "TagAttrStart");
TagAttrEquals = PyObject_GetAttrString(module, "TagAttrEquals");
TagAttrQuote = PyObject_GetAttrString(module, "TagAttrQuote");
TagCloseOpen = PyObject_GetAttrString(module, "TagCloseOpen");
TagCloseSelfclose = PyObject_GetAttrString(module, "TagCloseSelfclose");
TagOpenClose = PyObject_GetAttrString(module, "TagOpenClose");
TagCloseClose = PyObject_GetAttrString(module, "TagCloseClose");
}

+ 69
- 0
mwparserfromhell/parser/ctokenizer/tokens.h View File

@@ -0,0 +1,69 @@
/*
Copyright (C) 2012-2015 Ben Kurtovic <ben.kurtovic@gmail.com>

Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/

#pragma once

#include "common.h"

/* Token globals */

extern PyObject* Text;

extern PyObject* TemplateOpen;
extern PyObject* TemplateParamSeparator;
extern PyObject* TemplateParamEquals;
extern PyObject* TemplateClose;

extern PyObject* ArgumentOpen;
extern PyObject* ArgumentSeparator;
extern PyObject* ArgumentClose;

extern PyObject* WikilinkOpen;
extern PyObject* WikilinkSeparator;
extern PyObject* WikilinkClose;

extern PyObject* ExternalLinkOpen;
extern PyObject* ExternalLinkSeparator;
extern PyObject* ExternalLinkClose;

extern PyObject* HTMLEntityStart;
extern PyObject* HTMLEntityNumeric;
extern PyObject* HTMLEntityHex;
extern PyObject* HTMLEntityEnd;
extern PyObject* HeadingStart;
extern PyObject* HeadingEnd;

extern PyObject* CommentStart;
extern PyObject* CommentEnd;

extern PyObject* TagOpenOpen;
extern PyObject* TagAttrStart;
extern PyObject* TagAttrEquals;
extern PyObject* TagAttrQuote;
extern PyObject* TagCloseOpen;
extern PyObject* TagCloseSelfclose;
extern PyObject* TagOpenClose;
extern PyObject* TagCloseClose;

/* Functions */

void load_tokens_from_module(PyObject*);

Write Preview
Loading…
Cancel
Save