Mirror
/
cpython
mirror of https://github.com/python/cpython
4
1
Fork 0
Code Issues Releases Wiki Activity
cpython/Modules/unicodedata.c

1486 lines
44 KiB
C
Raw Blame History

/* ------------------------------------------------------------------------
unicodedata -- Provides access to the Unicode database.
Data was extracted from the UnicodeData.txt file.
The current version number is reported in the unidata_version constant.
Written by Marc-Andre Lemburg (mal@lemburg.com).
Modified for Python 2.0 by Fredrik Lundh (fredrik@pythonware.com)
Modified by Martin v. Löwis (martin@v.loewis.de)
Copyright (c) Corporation for National Research Initiatives.
------------------------------------------------------------------------ */
#define PY_SSIZE_T_CLEAN
#include "Python.h"
#include "ucnhash.h"
#include "structmember.h"
#include <stdbool.h>
_Py_IDENTIFIER(NFC);
_Py_IDENTIFIER(NFD);
_Py_IDENTIFIER(NFKC);
_Py_IDENTIFIER(NFKD);
/*[clinic input]
module unicodedata
class unicodedata.UCD 'PreviousDBVersion *' '&UCD_Type'
[clinic start generated code]*/
/*[clinic end generated code: output=da39a3ee5e6b4b0d input=6dac153082d150bc]*/
/* character properties */
typedef struct {
const unsigned char category; /* index into
_PyUnicode_CategoryNames */
const unsigned char combining; /* combining class value 0 - 255 */
const unsigned char bidirectional; /* index into
_PyUnicode_BidirectionalNames */
const unsigned char mirrored; /* true if mirrored in bidir mode */
const unsigned char east_asian_width; /* index into
_PyUnicode_EastAsianWidth */
const unsigned char normalization_quick_check; /* see is_normalized() */
} _PyUnicode_DatabaseRecord;
typedef struct change_record {
/* sequence of fields should be the same as in merge_old_version */
const unsigned char bidir_changed;
const unsigned char category_changed;
const unsigned char decimal_changed;
const unsigned char mirrored_changed;
const unsigned char east_asian_width_changed;
const double numeric_changed;
} change_record;
/* data file generated by Tools/unicode/makeunicodedata.py */
#include "unicodedata_db.h"
static const _PyUnicode_DatabaseRecord*
_getrecord_ex(Py_UCS4 code)
{
int index;
if (code >= 0x110000)
index = 0;
else {
index = index1[(code>>SHIFT)];
index = index2[(index<<SHIFT)+(code&((1<<SHIFT)-1))];
}
return &_PyUnicode_Database_Records[index];
}
/* ------------- Previous-version API ------------------------------------- */
typedef struct previous_version {
PyObject_HEAD
const char *name;
const change_record* (*getrecord)(Py_UCS4);
Py_UCS4 (*normalization)(Py_UCS4);
} PreviousDBVersion;
#include "clinic/unicodedata.c.h"
#define get_old_record(self, v) ((((PreviousDBVersion*)self)->getrecord)(v))
static PyMemberDef DB_members[] = {
{"unidata_version", T_STRING, offsetof(PreviousDBVersion, name), READONLY},
{NULL}
};
/* forward declaration */
static PyTypeObject UCD_Type;
#define UCD_Check(o) (Py_TYPE(o)==&UCD_Type)
static PyObject*
new_previous_version(const char*name, const change_record* (*getrecord)(Py_UCS4),
Py_UCS4 (*normalization)(Py_UCS4))
{
PreviousDBVersion *self;
self = PyObject_New(PreviousDBVersion, &UCD_Type);
if (self == NULL)
return NULL;
self->name = name;
self->getrecord = getrecord;
self->normalization = normalization;
return (PyObject*)self;
}
/* --- Module API --------------------------------------------------------- */
/*[clinic input]
unicodedata.UCD.decimal
self: self
chr: int(accept={str})
default: object=NULL
/
Converts a Unicode character into its equivalent decimal value.
Returns the decimal value assigned to the character chr as integer.
If no such value is defined, default is returned, or, if not given,
ValueError is raised.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_decimal_impl(PyObject *self, int chr,
PyObject *default_value)
/*[clinic end generated code: output=be23376e1a185231 input=933f8107993f23d0]*/
{
int have_old = 0;
long rc;
Py_UCS4 c = (Py_UCS4)chr;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0) {
/* unassigned */
have_old = 1;
rc = -1;
}
else if (old->decimal_changed != 0xFF) {
have_old = 1;
rc = old->decimal_changed;
}
}
if (!have_old)
rc = Py_UNICODE_TODECIMAL(c);
if (rc < 0) {
if (default_value == NULL) {
PyErr_SetString(PyExc_ValueError,
"not a decimal");
return NULL;
}
else {
Py_INCREF(default_value);
return default_value;
}
}
return PyLong_FromLong(rc);
}
/*[clinic input]
unicodedata.UCD.digit
self: self
chr: int(accept={str})
default: object=NULL
/
Converts a Unicode character into its equivalent digit value.
Returns the digit value assigned to the character chr as integer.
If no such value is defined, default is returned, or, if not given,
ValueError is raised.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_digit_impl(PyObject *self, int chr, PyObject *default_value)
/*[clinic end generated code: output=96e18c950171fd2f input=e27d6e4565cd29f2]*/
{
long rc;
Py_UCS4 c = (Py_UCS4)chr;
rc = Py_UNICODE_TODIGIT(c);
if (rc < 0) {
if (default_value == NULL) {
PyErr_SetString(PyExc_ValueError, "not a digit");
return NULL;
}
else {
Py_INCREF(default_value);
return default_value;
}
}
return PyLong_FromLong(rc);
}
/*[clinic input]
unicodedata.UCD.numeric
self: self
chr: int(accept={str})
default: object=NULL
/
Converts a Unicode character into its equivalent numeric value.
Returns the numeric value assigned to the character chr as float.
If no such value is defined, default is returned, or, if not given,
ValueError is raised.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_numeric_impl(PyObject *self, int chr,
PyObject *default_value)
/*[clinic end generated code: output=53ce281fe85b10c4 input=fdf5871a5542893c]*/
{
int have_old = 0;
double rc;
Py_UCS4 c = (Py_UCS4)chr;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0) {
/* unassigned */
have_old = 1;
rc = -1.0;
}
else if (old->decimal_changed != 0xFF) {
have_old = 1;
rc = old->decimal_changed;
}
}
if (!have_old)
rc = Py_UNICODE_TONUMERIC(c);
if (rc == -1.0) {
if (default_value == NULL) {
PyErr_SetString(PyExc_ValueError, "not a numeric character");
return NULL;
}
else {
Py_INCREF(default_value);
return default_value;
}
}
return PyFloat_FromDouble(rc);
}
/*[clinic input]
unicodedata.UCD.category
self: self
chr: int(accept={str})
/
Returns the general category assigned to the character chr as string.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_category_impl(PyObject *self, int chr)
/*[clinic end generated code: output=8571539ee2e6783a input=27d6f3d85050bc06]*/
{
int index;
Py_UCS4 c = (Py_UCS4)chr;
index = (int) _getrecord_ex(c)->category;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed != 0xFF)
index = old->category_changed;
}
return PyUnicode_FromString(_PyUnicode_CategoryNames[index]);
}
/*[clinic input]
unicodedata.UCD.bidirectional
self: self
chr: int(accept={str})
/
Returns the bidirectional class assigned to the character chr as string.
If no such value is defined, an empty string is returned.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_bidirectional_impl(PyObject *self, int chr)
/*[clinic end generated code: output=d36310ce2039bb92 input=b3d8f42cebfcf475]*/
{
int index;
Py_UCS4 c = (Py_UCS4)chr;
index = (int) _getrecord_ex(c)->bidirectional;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0)
index = 0; /* unassigned */
else if (old->bidir_changed != 0xFF)
index = old->bidir_changed;
}
return PyUnicode_FromString(_PyUnicode_BidirectionalNames[index]);
}
/*[clinic input]
unicodedata.UCD.combining -> int
self: self
chr: int(accept={str})
/
Returns the canonical combining class assigned to the character chr as integer.
Returns 0 if no combining class is defined.
[clinic start generated code]*/
static int
unicodedata_UCD_combining_impl(PyObject *self, int chr)
/*[clinic end generated code: output=cad056d0cb6a5920 input=9f2d6b2a95d0a22a]*/
{
int index;
Py_UCS4 c = (Py_UCS4)chr;
index = (int) _getrecord_ex(c)->combining;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0)
index = 0; /* unassigned */
}
return index;
}
/*[clinic input]
unicodedata.UCD.mirrored -> int
self: self
chr: int(accept={str})
/
Returns the mirrored property assigned to the character chr as integer.
Returns 1 if the character has been identified as a "mirrored"
character in bidirectional text, 0 otherwise.
[clinic start generated code]*/
static int
unicodedata_UCD_mirrored_impl(PyObject *self, int chr)
/*[clinic end generated code: output=2532dbf8121b50e6 input=5dd400d351ae6f3b]*/
{
int index;
Py_UCS4 c = (Py_UCS4)chr;
index = (int) _getrecord_ex(c)->mirrored;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0)
index = 0; /* unassigned */
else if (old->mirrored_changed != 0xFF)
index = old->mirrored_changed;
}
return index;
}
/*[clinic input]
unicodedata.UCD.east_asian_width
self: self
chr: int(accept={str})
/
Returns the east asian width assigned to the character chr as string.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_east_asian_width_impl(PyObject *self, int chr)
/*[clinic end generated code: output=484e8537d9ee8197 input=c4854798aab026e0]*/
{
int index;
Py_UCS4 c = (Py_UCS4)chr;
index = (int) _getrecord_ex(c)->east_asian_width;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0)
index = 0; /* unassigned */
else if (old->east_asian_width_changed != 0xFF)
index = old->east_asian_width_changed;
}
return PyUnicode_FromString(_PyUnicode_EastAsianWidthNames[index]);
}
/*[clinic input]
unicodedata.UCD.decomposition
self: self
chr: int(accept={str})
/
Returns the character decomposition mapping assigned to the character chr as string.
An empty string is returned in case no such mapping is defined.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_decomposition_impl(PyObject *self, int chr)
/*[clinic end generated code: output=7d699f3ec7565d27 input=e4c12459ad68507b]*/
{
char decomp[256];
int code, index, count;
size_t i;
unsigned int prefix_index;
Py_UCS4 c = (Py_UCS4)chr;
code = (int)c;
if (self && UCD_Check(self)) {
const change_record *old = get_old_record(self, c);
if (old->category_changed == 0)
return PyUnicode_FromString(""); /* unassigned */
}
if (code < 0 || code >= 0x110000)
index = 0;
else {
index = decomp_index1[(code>>DECOMP_SHIFT)];
index = decomp_index2[(index<<DECOMP_SHIFT)+
(code&((1<<DECOMP_SHIFT)-1))];
}
/* high byte is number of hex bytes (usually one or two), low byte
is prefix code (from*/
count = decomp_data[index] >> 8;
/* XXX: could allocate the PyString up front instead
(strlen(prefix) + 5 * count + 1 bytes) */
/* Based on how index is calculated above and decomp_data is generated
from Tools/unicode/makeunicodedata.py, it should not be possible
to overflow decomp_prefix. */
prefix_index = decomp_data[index] & 255;
assert(prefix_index < Py_ARRAY_LENGTH(decomp_prefix));
/* copy prefix */
i = strlen(decomp_prefix[prefix_index]);
memcpy(decomp, decomp_prefix[prefix_index], i);
while (count-- > 0) {
if (i)
decomp[i++] = ' ';
assert(i < sizeof(decomp));
PyOS_snprintf(decomp + i, sizeof(decomp) - i, "%04X",
decomp_data[++index]);
i += strlen(decomp + i);
}
return PyUnicode_FromStringAndSize(decomp, i);
}
static void
get_decomp_record(PyObject *self, Py_UCS4 code, int *index, int *prefix, int *count)
{
if (code >= 0x110000) {
*index = 0;
} else if (self && UCD_Check(self) &&
get_old_record(self, code)->category_changed==0) {
/* unassigned in old version */
*index = 0;
}
else {
*index = decomp_index1[(code>>DECOMP_SHIFT)];
*index = decomp_index2[(*index<<DECOMP_SHIFT)+
(code&((1<<DECOMP_SHIFT)-1))];
}
/* high byte is number of hex bytes (usually one or two), low byte
is prefix code (from*/
*count = decomp_data[*index] >> 8;
*prefix = decomp_data[*index] & 255;
(*index)++;
}
#define SBase 0xAC00
#define LBase 0x1100
#define VBase 0x1161
#define TBase 0x11A7
#define LCount 19
#define VCount 21
#define TCount 28
#define NCount (VCount*TCount)
#define SCount (LCount*NCount)
static PyObject*
nfd_nfkd(PyObject *self, PyObject *input, int k)
{
PyObject *result;
Py_UCS4 *output;
Py_ssize_t i, o, osize;
int kind;
void *data;
/* Longest decomposition in Unicode 3.2: U+FDFA */
Py_UCS4 stack[20];
Py_ssize_t space, isize;
int index, prefix, count, stackptr;
unsigned char prev, cur;
stackptr = 0;
isize = PyUnicode_GET_LENGTH(input);
space = isize;
/* Overallocate at most 10 characters. */
if (space > 10) {
if (space <= PY_SSIZE_T_MAX - 10)
space += 10;
}
else {
space *= 2;
}
osize = space;
output = PyMem_NEW(Py_UCS4, space);
if (!output) {
PyErr_NoMemory();
return NULL;
}
i = o = 0;
kind = PyUnicode_KIND(input);
data = PyUnicode_DATA(input);
while (i < isize) {
stack[stackptr++] = PyUnicode_READ(kind, data, i++);
while(stackptr) {
Py_UCS4 code = stack[--stackptr];
/* Hangul Decomposition adds three characters in
a single step, so we need at least that much room. */
if (space < 3) {
Py_UCS4 *new_output;
osize += 10;
space += 10;
new_output = PyMem_Realloc(output, osize*sizeof(Py_UCS4));
if (new_output == NULL) {
PyMem_Free(output);
PyErr_NoMemory();
return NULL;
}
output = new_output;
}
/* Hangul Decomposition. */
if (SBase <= code && code < (SBase+SCount)) {
int SIndex = code - SBase;
int L = LBase + SIndex / NCount;
int V = VBase + (SIndex % NCount) / TCount;
int T = TBase + SIndex % TCount;
output[o++] = L;
output[o++] = V;
space -= 2;
if (T != TBase) {
output[o++] = T;
space --;
}
continue;
}
/* normalization changes */
if (self && UCD_Check(self)) {
Py_UCS4 value = ((PreviousDBVersion*)self)->normalization(code);
if (value != 0) {
stack[stackptr++] = value;
continue;
}
}
/* Other decompositions. */
get_decomp_record(self, code, &index, &prefix, &count);
/* Copy character if it is not decomposable, or has a
compatibility decomposition, but we do NFD. */
if (!count || (prefix && !k)) {
output[o++] = code;
space--;
continue;
}
/* Copy decomposition onto the stack, in reverse
order. */
while(count) {
code = decomp_data[index + (--count)];
stack[stackptr++] = code;
}
}
}
result = PyUnicode_FromKindAndData(PyUnicode_4BYTE_KIND,
output, o);
PyMem_Free(output);
if (!result)
return NULL;
/* result is guaranteed to be ready, as it is compact. */
kind = PyUnicode_KIND(result);
data = PyUnicode_DATA(result);
/* Sort canonically. */
i = 0;
prev = _getrecord_ex(PyUnicode_READ(kind, data, i))->combining;
for (i++; i < PyUnicode_GET_LENGTH(result); i++) {
cur = _getrecord_ex(PyUnicode_READ(kind, data, i))->combining;
if (prev == 0 || cur == 0 || prev <= cur) {
prev = cur;
continue;
}
/* Non-canonical order. Need to switch *i with previous. */
o = i - 1;
while (1) {
Py_UCS4 tmp = PyUnicode_READ(kind, data, o+1);
PyUnicode_WRITE(kind, data, o+1,
PyUnicode_READ(kind, data, o));
PyUnicode_WRITE(kind, data, o, tmp);
o--;
if (o < 0)
break;
prev = _getrecord_ex(PyUnicode_READ(kind, data, o))->combining;
if (prev == 0 || prev <= cur)
break;
}
prev = _getrecord_ex(PyUnicode_READ(kind, data, i))->combining;
}
return result;
}
static int
find_nfc_index(PyObject *self, struct reindex* nfc, Py_UCS4 code)
{
unsigned int index;
for (index = 0; nfc[index].start; index++) {
unsigned int start = nfc[index].start;
if (code < start)
return -1;
if (code <= start + nfc[index].count) {
unsigned int delta = code - start;
return nfc[index].index + delta;
}
}
return -1;
}
static PyObject*
nfc_nfkc(PyObject *self, PyObject *input, int k)
{
PyObject *result;
int kind;
void *data;
Py_UCS4 *output;
Py_ssize_t i, i1, o, len;
int f,l,index,index1,comb;
Py_UCS4 code;
Py_ssize_t skipped[20];
int cskipped = 0;
result = nfd_nfkd(self, input, k);
if (!result)
return NULL;
/* result will be "ready". */
kind = PyUnicode_KIND(result);
data = PyUnicode_DATA(result);
len = PyUnicode_GET_LENGTH(result);
/* We allocate a buffer for the output.
If we find that we made no changes, we still return
the NFD result. */
output = PyMem_NEW(Py_UCS4, len);
if (!output) {
PyErr_NoMemory();
Py_DECREF(result);
return 0;
}
i = o = 0;
again:
while (i < len) {
for (index = 0; index < cskipped; index++) {
if (skipped[index] == i) {
/* *i character is skipped.
Remove from list. */
skipped[index] = skipped[cskipped-1];
cskipped--;
i++;
goto again; /* continue while */
}
}
/* Hangul Composition. We don't need to check for <LV,T>
pairs, since we always have decomposed data. */
code = PyUnicode_READ(kind, data, i);
if (LBase <= code && code < (LBase+LCount) &&
i + 1 < len &&
VBase <= PyUnicode_READ(kind, data, i+1) &&
PyUnicode_READ(kind, data, i+1) < (VBase+VCount)) {
/* check L character is a modern leading consonant (0x1100 ~ 0x1112)
and V character is a modern vowel (0x1161 ~ 0x1175). */
int LIndex, VIndex;
LIndex = code - LBase;
VIndex = PyUnicode_READ(kind, data, i+1) - VBase;
code = SBase + (LIndex*VCount+VIndex)*TCount;
i+=2;
if (i < len &&
TBase < PyUnicode_READ(kind, data, i) &&
PyUnicode_READ(kind, data, i) < (TBase+TCount)) {
/* check T character is a modern trailing consonant
(0x11A8 ~ 0x11C2). */
code += PyUnicode_READ(kind, data, i)-TBase;
i++;
}
output[o++] = code;
continue;
}
/* code is still input[i] here */
f = find_nfc_index(self, nfc_first, code);
if (f == -1) {
output[o++] = code;
i++;
continue;
}
/* Find next unblocked character. */
i1 = i+1;
comb = 0;
/* output base character for now; might be updated later. */
output[o] = PyUnicode_READ(kind, data, i);
while (i1 < len) {
Py_UCS4 code1 = PyUnicode_READ(kind, data, i1);
int comb1 = _getrecord_ex(code1)->combining;
if (comb) {
if (comb1 == 0)
break;
if (comb >= comb1) {
/* Character is blocked. */
i1++;
continue;
}
}
l = find_nfc_index(self, nfc_last, code1);
/* i1 cannot be combined with i. If i1
is a starter, we don't need to look further.
Otherwise, record the combining class. */
if (l == -1) {
not_combinable:
if (comb1 == 0)
break;
comb = comb1;
i1++;
continue;
}
index = f*TOTAL_LAST + l;
index1 = comp_index[index >> COMP_SHIFT];
code = comp_data[(index1<<COMP_SHIFT)+
(index&((1<<COMP_SHIFT)-1))];
if (code == 0)
goto not_combinable;
/* Replace the original character. */
output[o] = code;
/* Mark the second character unused. */
assert(cskipped < 20);
skipped[cskipped++] = i1;
i1++;
f = find_nfc_index(self, nfc_first, output[o]);
if (f == -1)
break;
}
/* Output character was already written.
Just advance the indices. */
o++; i++;
}
if (o == len) {
/* No changes. Return original string. */
PyMem_Free(output);
return result;
}
Py_DECREF(result);
result = PyUnicode_FromKindAndData(PyUnicode_4BYTE_KIND,
output, o);
PyMem_Free(output);
return result;
}
// This needs to match the logic in makeunicodedata.py
// which constructs the quickcheck data.
typedef enum {YES = 0, MAYBE = 1, NO = 2} QuickcheckResult;
/* Run the Unicode normalization "quickcheck" algorithm.
*
* Return YES or NO if quickcheck determines the input is certainly
* normalized or certainly not, and MAYBE if quickcheck is unable to
* tell.
*
* If `yes_only` is true, then return MAYBE as soon as we determine
* the answer is not YES.
*
* For background and details on the algorithm, see UAX #15:
* https://www.unicode.org/reports/tr15/#Detecting_Normalization_Forms
*/
static QuickcheckResult
is_normalized_quickcheck(PyObject *self, PyObject *input,
bool nfc, bool k, bool yes_only)
{
/* An older version of the database is requested, quickchecks must be
disabled. */
if (self && UCD_Check(self))
return NO;
Py_ssize_t i, len;
int kind;
void *data;
unsigned char prev_combining = 0;
/* The two quickcheck bits at this shift have type QuickcheckResult. */
int quickcheck_shift = (nfc ? 4 : 0) + (k ? 2 : 0);
QuickcheckResult result = YES; /* certainly normalized, unless we find something */
i = 0;
kind = PyUnicode_KIND(input);
data = PyUnicode_DATA(input);
len = PyUnicode_GET_LENGTH(input);
while (i < len) {
Py_UCS4 ch = PyUnicode_READ(kind, data, i++);
const _PyUnicode_DatabaseRecord *record = _getrecord_ex(ch);
unsigned char combining = record->combining;
if (combining && prev_combining > combining)
return NO; /* non-canonical sort order, not normalized */
prev_combining = combining;
unsigned char quickcheck_whole = record->normalization_quick_check;
if (yes_only) {
if (quickcheck_whole & (3 << quickcheck_shift))
return MAYBE;
} else {
switch ((quickcheck_whole >> quickcheck_shift) & 3) {
case NO:
return NO;
case MAYBE:
result = MAYBE; /* this string might need normalization */
}
}
}
return result;
}
/*[clinic input]
unicodedata.UCD.is_normalized
self: self
form: unicode
unistr as input: unicode
/
Return whether the Unicode string unistr is in the normal form 'form'.
Valid values for form are 'NFC', 'NFKC', 'NFD', and 'NFKD'.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_is_normalized_impl(PyObject *self, PyObject *form,
PyObject *input)
/*[clinic end generated code: output=11e5a3694e723ca5 input=a544f14cea79e508]*/
{
if (PyUnicode_READY(input) == -1) {
return NULL;
}
if (PyUnicode_GET_LENGTH(input) == 0) {
/* special case empty input strings. */
Py_RETURN_TRUE;
}
PyObject *result;
bool nfc = false;
bool k = false;
QuickcheckResult m;
PyObject *cmp;
int match = 0;
if (_PyUnicode_EqualToASCIIId(form, &PyId_NFC)) {
nfc = true;
}
else if (_PyUnicode_EqualToASCIIId(form, &PyId_NFKC)) {
nfc = true;
k = true;
}
else if (_PyUnicode_EqualToASCIIId(form, &PyId_NFD)) {
/* matches default values for `nfc` and `k` */
}
else if (_PyUnicode_EqualToASCIIId(form, &PyId_NFKD)) {
k = true;
}
else {
PyErr_SetString(PyExc_ValueError, "invalid normalization form");
return NULL;
}
m = is_normalized_quickcheck(self, input, nfc, k, false);
if (m == MAYBE) {
cmp = (nfc ? nfc_nfkc : nfd_nfkd)(self, input, k);
if (cmp == NULL) {
return NULL;
}
match = PyUnicode_Compare(input, cmp);
Py_DECREF(cmp);
result = (match == 0) ? Py_True : Py_False;
}
else {
result = (m == YES) ? Py_True : Py_False;
}
Py_INCREF(result);
return result;
}
/*[clinic input]
unicodedata.UCD.normalize
self: self
form: unicode
unistr as input: unicode
/
Return the normal form 'form' for the Unicode string unistr.
Valid values for form are 'NFC', 'NFKC', 'NFD', and 'NFKD'.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_normalize_impl(PyObject *self, PyObject *form,
PyObject *input)
/*[clinic end generated code: output=05ca4385a2ad6983 input=3a5206c0ad2833fb]*/
{
if (PyUnicode_GET_LENGTH(input) == 0) {
/* Special case empty input strings, since resizing
them later would cause internal errors. */
Py_INCREF(input);
return input;
}
if (_PyUnicode_EqualToASCIIId(form, &PyId_NFC)) {
if (is_normalized_quickcheck(self, input, true, false, true) == YES) {
Py_INCREF(input);
return input;
}
return nfc_nfkc(self, input, 0);
}
if (_PyUnicode_EqualToASCIIId(form, &PyId_NFKC)) {
if (is_normalized_quickcheck(self, input, true, true, true) == YES) {
Py_INCREF(input);
return input;
}
return nfc_nfkc(self, input, 1);
}
if (_PyUnicode_EqualToASCIIId(form, &PyId_NFD)) {
if (is_normalized_quickcheck(self, input, false, false, true) == YES) {
Py_INCREF(input);
return input;
}
return nfd_nfkd(self, input, 0);
}
if (_PyUnicode_EqualToASCIIId(form, &PyId_NFKD)) {
if (is_normalized_quickcheck(self, input, false, true, true) == YES) {
Py_INCREF(input);
return input;
}
return nfd_nfkd(self, input, 1);
}
PyErr_SetString(PyExc_ValueError, "invalid normalization form");
return NULL;
}
/* -------------------------------------------------------------------- */
/* unicode character name tables */
/* data file generated by Tools/unicode/makeunicodedata.py */
#include "unicodename_db.h"
/* -------------------------------------------------------------------- */
/* database code (cut and pasted from the unidb package) */
static unsigned long
_gethash(const char *s, int len, int scale)
{
int i;
unsigned long h = 0;
unsigned long ix;
for (i = 0; i < len; i++) {
h = (h * scale) + (unsigned char) Py_TOUPPER(Py_CHARMASK(s[i]));
ix = h & 0xff000000;
if (ix)
h = (h ^ ((ix>>24) & 0xff)) & 0x00ffffff;
}
return h;
}
static const char * const hangul_syllables[][3] = {
{ "G", "A", "" },
{ "GG", "AE", "G" },
{ "N", "YA", "GG" },
{ "D", "YAE", "GS" },
{ "DD", "EO", "N", },
{ "R", "E", "NJ" },
{ "M", "YEO", "NH" },
{ "B", "YE", "D" },
{ "BB", "O", "L" },
{ "S", "WA", "LG" },
{ "SS", "WAE", "LM" },
{ "", "OE", "LB" },
{ "J", "YO", "LS" },
{ "JJ", "U", "LT" },
{ "C", "WEO", "LP" },
{ "K", "WE", "LH" },
{ "T", "WI", "M" },
{ "P", "YU", "B" },
{ "H", "EU", "BS" },
{ 0, "YI", "S" },
{ 0, "I", "SS" },
{ 0, 0, "NG" },
{ 0, 0, "J" },
{ 0, 0, "C" },
{ 0, 0, "K" },
{ 0, 0, "T" },
{ 0, 0, "P" },
{ 0, 0, "H" }
};
/* These ranges need to match makeunicodedata.py:cjk_ranges. */
static int
is_unified_ideograph(Py_UCS4 code)
{
return
(0x3400 <= code && code <= 0x4DB5) || /* CJK Ideograph Extension A */
(0x4E00 <= code && code <= 0x9FEF) || /* CJK Ideograph */
(0x20000 <= code && code <= 0x2A6D6) || /* CJK Ideograph Extension B */
(0x2A700 <= code && code <= 0x2B734) || /* CJK Ideograph Extension C */
(0x2B740 <= code && code <= 0x2B81D) || /* CJK Ideograph Extension D */
(0x2B820 <= code && code <= 0x2CEA1) || /* CJK Ideograph Extension E */
(0x2CEB0 <= code && code <= 0x2EBEF); /* CJK Ideograph Extension F */
}
/* macros used to determine if the given code point is in the PUA range that
* we are using to store aliases and named sequences */
#define IS_ALIAS(cp) ((cp >= aliases_start) && (cp < aliases_end))
#define IS_NAMED_SEQ(cp) ((cp >= named_sequences_start) && \
(cp < named_sequences_end))
static int
_getucname(PyObject *self, Py_UCS4 code, char* buffer, int buflen,
int with_alias_and_seq)
{
/* Find the name associated with the given code point.
* If with_alias_and_seq is 1, check for names in the Private Use Area 15
* that we are using for aliases and named sequences. */
int offset;
int i;
int word;
const unsigned char* w;
if (code >= 0x110000)
return 0;
/* XXX should we just skip all the code points in the PUAs here? */
if (!with_alias_and_seq && (IS_ALIAS(code) || IS_NAMED_SEQ(code)))
return 0;
if (self && UCD_Check(self)) {
/* in 3.2.0 there are no aliases and named sequences */
const change_record *old;
if (IS_ALIAS(code) || IS_NAMED_SEQ(code))
return 0;
old = get_old_record(self, code);
if (old->category_changed == 0) {
/* unassigned */
return 0;
}
}
if (SBase <= code && code < SBase+SCount) {
/* Hangul syllable. */
int SIndex = code - SBase;
int L = SIndex / NCount;
int V = (SIndex % NCount) / TCount;
int T = SIndex % TCount;
if (buflen < 27)
/* Worst case: HANGUL SYLLABLE <10chars>. */
return 0;
strcpy(buffer, "HANGUL SYLLABLE ");
buffer += 16;
strcpy(buffer, hangul_syllables[L][0]);
buffer += strlen(hangul_syllables[L][0]);
strcpy(buffer, hangul_syllables[V][1]);
buffer += strlen(hangul_syllables[V][1]);
strcpy(buffer, hangul_syllables[T][2]);
buffer += strlen(hangul_syllables[T][2]);
*buffer = '\0';
return 1;
}
if (is_unified_ideograph(code)) {
if (buflen < 28)
/* Worst case: CJK UNIFIED IDEOGRAPH-20000 */
return 0;
sprintf(buffer, "CJK UNIFIED IDEOGRAPH-%X", code);
return 1;
}
/* get offset into phrasebook */
offset = phrasebook_offset1[(code>>phrasebook_shift)];
offset = phrasebook_offset2[(offset<<phrasebook_shift) +
(code&((1<<phrasebook_shift)-1))];
if (!offset)
return 0;
i = 0;
for (;;) {
/* get word index */
word = phrasebook[offset] - phrasebook_short;
if (word >= 0) {
word = (word << 8) + phrasebook[offset+1];
offset += 2;
} else
word = phrasebook[offset++];
if (i) {
if (i > buflen)
return 0; /* buffer overflow */
buffer[i++] = ' ';
}
/* copy word string from lexicon. the last character in the
word has bit 7 set. the last word in a string ends with
0x80 */
w = lexicon + lexicon_offset[word];
while (*w < 128) {
if (i >= buflen)
return 0; /* buffer overflow */
buffer[i++] = *w++;
}
if (i >= buflen)
return 0; /* buffer overflow */
buffer[i++] = *w & 127;
if (*w == 128)
break; /* end of word */
}
return 1;
}
static int
_cmpname(PyObject *self, int code, const char* name, int namelen)
{
/* check if code corresponds to the given name */
int i;
char buffer[NAME_MAXLEN+1];
if (!_getucname(self, code, buffer, NAME_MAXLEN, 1))
return 0;
for (i = 0; i < namelen; i++) {
if (Py_TOUPPER(Py_CHARMASK(name[i])) != buffer[i])
return 0;
}
return buffer[namelen] == '\0';
}
static void
find_syllable(const char *str, int *len, int *pos, int count, int column)
{
int i, len1;
*len = -1;
for (i = 0; i < count; i++) {
const char *s = hangul_syllables[i][column];
len1 = Py_SAFE_DOWNCAST(strlen(s), size_t, int);
if (len1 <= *len)
continue;
if (strncmp(str, s, len1) == 0) {
*len = len1;
*pos = i;
}
}
if (*len == -1) {
*len = 0;
}
}
static int
_check_alias_and_seq(unsigned int cp, Py_UCS4* code, int with_named_seq)
{
/* check if named sequences are allowed */
if (!with_named_seq && IS_NAMED_SEQ(cp))
return 0;
/* if the code point is in the PUA range that we use for aliases,
* convert it to obtain the right code point */
if (IS_ALIAS(cp))
*code = name_aliases[cp-aliases_start];
else
*code = cp;
return 1;
}
static int
_getcode(PyObject* self, const char* name, int namelen, Py_UCS4* code,
int with_named_seq)
{
/* Return the code point associated with the given name.
* Named aliases are resolved too (unless self != NULL (i.e. we are using
* 3.2.0)). If with_named_seq is 1, returns the PUA code point that we are
* using for the named sequence, and the caller must then convert it. */
unsigned int h, v;
unsigned int mask = code_size-1;
unsigned int i, incr;
/* Check for hangul syllables. */
if (strncmp(name, "HANGUL SYLLABLE ", 16) == 0) {
int len, L = -1, V = -1, T = -1;
const char *pos = name + 16;
find_syllable(pos, &len, &L, LCount, 0);
pos += len;
find_syllable(pos, &len, &V, VCount, 1);
pos += len;
find_syllable(pos, &len, &T, TCount, 2);
pos += len;
if (L != -1 && V != -1 && T != -1 && pos-name == namelen) {
*code = SBase + (L*VCount+V)*TCount + T;
return 1;
}
/* Otherwise, it's an illegal syllable name. */
return 0;
}
/* Check for unified ideographs. */
if (strncmp(name, "CJK UNIFIED IDEOGRAPH-", 22) == 0) {
/* Four or five hexdigits must follow. */
v = 0;
name += 22;
namelen -= 22;
if (namelen != 4 && namelen != 5)
return 0;
while (namelen--) {
v *= 16;
if (*name >= '0' && *name <= '9')
v += *name - '0';
else if (*name >= 'A' && *name <= 'F')
v += *name - 'A' + 10;
else
return 0;
name++;
}
if (!is_unified_ideograph(v))
return 0;
*code = v;
return 1;
}
/* the following is the same as python's dictionary lookup, with
only minor changes. see the makeunicodedata script for more
details */
h = (unsigned int) _gethash(name, namelen, code_magic);
i = (~h) & mask;
v = code_hash[i];
if (!v)
return 0;
if (_cmpname(self, v, name, namelen))
return _check_alias_and_seq(v, code, with_named_seq);
incr = (h ^ (h >> 3)) & mask;
if (!incr)
incr = mask;
for (;;) {
i = (i + incr) & mask;
v = code_hash[i];
if (!v)
return 0;
if (_cmpname(self, v, name, namelen))
return _check_alias_and_seq(v, code, with_named_seq);
incr = incr << 1;
if (incr > mask)
incr = incr ^ code_poly;
}
}
static const _PyUnicode_Name_CAPI hashAPI =
{
sizeof(_PyUnicode_Name_CAPI),
_getucname,
_getcode
};
/* -------------------------------------------------------------------- */
/* Python bindings */
/*[clinic input]
unicodedata.UCD.name
self: self
chr: int(accept={str})
default: object=NULL
/
Returns the name assigned to the character chr as a string.
If no name is defined, default is returned, or, if not given,
ValueError is raised.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_name_impl(PyObject *self, int chr, PyObject *default_value)
/*[clinic end generated code: output=6bbb37a326407707 input=3e0367f534de56d9]*/
{
char name[NAME_MAXLEN+1];
Py_UCS4 c = (Py_UCS4)chr;
if (!_getucname(self, c, name, NAME_MAXLEN, 0)) {
if (default_value == NULL) {
PyErr_SetString(PyExc_ValueError, "no such name");
return NULL;
}
else {
Py_INCREF(default_value);
return default_value;
}
}
return PyUnicode_FromString(name);
}
/*[clinic input]
unicodedata.UCD.lookup
self: self
name: str(accept={str, robuffer}, zeroes=True)
/
Look up character by name.
If a character with the given name is found, return the
corresponding character. If not found, KeyError is raised.
[clinic start generated code]*/
static PyObject *
unicodedata_UCD_lookup_impl(PyObject *self, const char *name,
Py_ssize_clean_t name_length)
/*[clinic end generated code: output=765cb8186788e6be input=a557be0f8607a0d6]*/
{
Py_UCS4 code;
unsigned int index;
if (name_length > NAME_MAXLEN) {
PyErr_SetString(PyExc_KeyError, "name too long");
return NULL;
}
if (!_getcode(self, name, (int)name_length, &code, 1)) {
PyErr_Format(PyExc_KeyError, "undefined character name '%s'", name);
return NULL;
}
/* check if code is in the PUA range that we use for named sequences
and convert it */
if (IS_NAMED_SEQ(code)) {
index = code-named_sequences_start;
return PyUnicode_FromKindAndData(PyUnicode_2BYTE_KIND,
named_sequences[index].seq,
named_sequences[index].seqlen);
}
return PyUnicode_FromOrdinal(code);
}
/* XXX Add doc strings. */
static PyMethodDef unicodedata_functions[] = {
UNICODEDATA_UCD_DECIMAL_METHODDEF
UNICODEDATA_UCD_DIGIT_METHODDEF
UNICODEDATA_UCD_NUMERIC_METHODDEF
UNICODEDATA_UCD_CATEGORY_METHODDEF
UNICODEDATA_UCD_BIDIRECTIONAL_METHODDEF
UNICODEDATA_UCD_COMBINING_METHODDEF
UNICODEDATA_UCD_MIRRORED_METHODDEF
UNICODEDATA_UCD_EAST_ASIAN_WIDTH_METHODDEF
UNICODEDATA_UCD_DECOMPOSITION_METHODDEF
UNICODEDATA_UCD_NAME_METHODDEF
UNICODEDATA_UCD_LOOKUP_METHODDEF
UNICODEDATA_UCD_IS_NORMALIZED_METHODDEF
UNICODEDATA_UCD_NORMALIZE_METHODDEF
{NULL, NULL} /* sentinel */
};
static PyTypeObject UCD_Type = {
/* The ob_type field must be initialized in the module init function
* to be portable to Windows without using C++. */
PyVarObject_HEAD_INIT(NULL, 0)
"unicodedata.UCD", /*tp_name*/
sizeof(PreviousDBVersion), /*tp_basicsize*/
0, /*tp_itemsize*/
/* methods */
(destructor)PyObject_Del, /*tp_dealloc*/
0, /*tp_vectorcall_offset*/
0, /*tp_getattr*/
0, /*tp_setattr*/
0, /*tp_as_async*/
0, /*tp_repr*/
0, /*tp_as_number*/
0, /*tp_as_sequence*/
0, /*tp_as_mapping*/
0, /*tp_hash*/
0, /*tp_call*/
0, /*tp_str*/
PyObject_GenericGetAttr,/*tp_getattro*/
0, /*tp_setattro*/
0, /*tp_as_buffer*/
Py_TPFLAGS_DEFAULT, /*tp_flags*/
0, /*tp_doc*/
0, /*tp_traverse*/
0, /*tp_clear*/
0, /*tp_richcompare*/
0, /*tp_weaklistoffset*/
0, /*tp_iter*/
0, /*tp_iternext*/
unicodedata_functions, /*tp_methods*/
DB_members, /*tp_members*/
0, /*tp_getset*/
0, /*tp_base*/
0, /*tp_dict*/
0, /*tp_descr_get*/
0, /*tp_descr_set*/
0, /*tp_dictoffset*/
0, /*tp_init*/
0, /*tp_alloc*/
0, /*tp_new*/
0, /*tp_free*/
0, /*tp_is_gc*/
};
PyDoc_STRVAR(unicodedata_docstring,
"This module provides access to the Unicode Character Database which\n\
defines character properties for all Unicode characters. The data in\n\
this database is based on the UnicodeData.txt file version\n\
" UNIDATA_VERSION " which is publicly available from ftp://ftp.unicode.org/.\n\
\n\
The module uses the same names and symbols as defined by the\n\
UnicodeData File Format " UNIDATA_VERSION ".");
static struct PyModuleDef unicodedatamodule = {
PyModuleDef_HEAD_INIT,
"unicodedata",
unicodedata_docstring,
-1,
unicodedata_functions,
NULL,
NULL,
NULL,
NULL
};
PyMODINIT_FUNC
PyInit_unicodedata(void)
{
PyObject *m, *v;
Py_TYPE(&UCD_Type) = &PyType_Type;
m = PyModule_Create(&unicodedatamodule);
if (!m)
return NULL;
PyModule_AddStringConstant(m, "unidata_version", UNIDATA_VERSION);
Py_INCREF(&UCD_Type);
PyModule_AddObject(m, "UCD", (PyObject*)&UCD_Type);
/* Previous versions */
v = new_previous_version("3.2.0", get_change_3_2_0, normalization_3_2_0);
if (v != NULL)
PyModule_AddObject(m, "ucd_3_2_0", v);
/* Export C API */
v = PyCapsule_New((void *)&hashAPI, PyUnicodeData_CAPSULE_NAME, NULL);
if (v != NULL)
PyModule_AddObject(m, "ucnhash_CAPI", v);
return m;
}
/*
Local variables:
c-basic-offset: 4
indent-tabs-mode: nil
End:
*/
Reference in New Issue View Git Blame Copy Permalink