Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commitf9c969b

Browse files
committed
Handwritten char iter and _PyBytesWriter_
1 parent4c4eca9 commitf9c969b

File tree

1 file changed

+102
-4
lines changed

1 file changed

+102
-4
lines changed

‎Modules/main.c‎

Lines changed: 102 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -234,9 +234,105 @@ static PyObject*
234234
_utf_8_bytes_dedent(PyObject*bytes){
235235
char*input_data;
236236
Py_ssize_tnchars;
237+
boolcurr_reading_whitespace= true;
238+
intcurr_num_leading_spaces=0;
239+
intcurr_num_leading_tabs=0;
240+
intnum_common_leading_chars;
241+
charc;
237242

238243
PyBytes_AsStringAndSize(bytes,&input_data,&nchars);
239244

245+
intnum_common_leading_spaces=nchars+1;
246+
intnum_common_leading_tabs=nchars+1;
247+
248+
char*data_iter;
249+
250+
data_iter=input_data;
251+
while ( (c=*data_iter++) ){
252+
if (c=='\n') {
253+
// Finished reading the line
254+
if (!curr_reading_whitespace) {
255+
// If the line had some non-whitespace characters
256+
// update the current common leading tab/space count
257+
if (curr_num_leading_spaces<num_common_leading_spaces) {
258+
num_common_leading_spaces=curr_num_leading_spaces;
259+
}
260+
if (curr_num_leading_tabs<num_common_leading_tabs) {
261+
num_common_leading_tabs=curr_num_leading_tabs;
262+
}
263+
}
264+
// About to start reading a new line
265+
curr_reading_whitespace= true;
266+
curr_num_leading_spaces=0;
267+
curr_num_leading_tabs=0;
268+
}
269+
elseif (curr_reading_whitespace){
270+
if (c==' '&&curr_num_leading_tabs==0) {
271+
curr_num_leading_spaces++;
272+
}
273+
elseif (c=='\t'&&curr_num_leading_spaces==0) {
274+
curr_num_leading_tabs++;
275+
}
276+
else {
277+
// Encountered a non-whitespace character
278+
curr_reading_whitespace= false;
279+
}
280+
}
281+
}
282+
if (num_common_leading_spaces>num_common_leading_tabs){
283+
num_common_leading_chars=num_common_leading_spaces;
284+
}
285+
else {
286+
num_common_leading_chars=num_common_leading_tabs;
287+
}
288+
289+
if (num_common_leading_chars>0) {
290+
// We need to trigger a dedent
291+
char*new_data;
292+
char*curr_line_ptr=input_data;
293+
char*next_line_ptr;
294+
char*new_start_loc;
295+
Py_ssize_tnew_line_len;
296+
_PyBytesWriterwriter;
297+
_PyBytesWriter_Init(&writer);
298+
new_data=_PyBytesWriter_Alloc(&writer,nchars);
299+
if (new_data==NULL) {
300+
returnNULL;
301+
}
302+
303+
data_iter=input_data;
304+
c=*data_iter;
305+
while (c) {
306+
// Find the end of the current line.
307+
while ( (c=*data_iter++)!='\n' ){
308+
if (c==NULL) {
309+
break;
310+
}
311+
}
312+
next_line_ptr=data_iter;
313+
Py_ssize_tline_len=next_line_ptr-curr_line_ptr;
314+
if (line_len>num_common_leading_chars){
315+
new_start_loc=curr_line_ptr+num_common_leading_chars;
316+
new_line_len=line_len-num_common_leading_chars;
317+
}
318+
else {
319+
new_start_loc=curr_line_ptr;
320+
new_line_len=line_len;
321+
}
322+
// Copy this line over to the new buffer (removing common
323+
// leading chars)
324+
new_data=_PyBytesWriter_WriteBytes(&writer,new_data,new_start_loc,new_line_len);
325+
curr_line_ptr=next_line_ptr;
326+
}
327+
PyObject*new_bytes=_PyBytesWriter_Finish(&writer,new_data);
328+
returnnew_bytes;
329+
}
330+
else {
331+
// No leading chars, no work to be done.
332+
returnbytes;
333+
}
334+
335+
#if0
240336
// Allocate new data for the output as a copy of the input
241337
PyBytesObject*new_bytes=PyBytes_FromStringAndSize(input_data,nchars);
242338
if (new_bytes==NULL) {
@@ -245,8 +341,6 @@ _utf_8_bytes_dedent(PyObject *bytes){
245341
char*new_data=PyBytes_AsString(new_bytes);
246342

247343
// Step 1: Find N = the common number leading whitespace chars
248-
249-
// so we can use the descructive strtok to tokenize the input.
250344
Py_ssize_tnum_common_leading_spaces=nchars+1;
251345

252346
// Count the number of leading spaces on each line
@@ -309,8 +403,9 @@ _utf_8_bytes_dedent(PyObject *bytes){
309403
}
310404
// null terminate the string (is this sufficient?)
311405
(*curr_dst)=NULL;
312-
313406
returnnew_bytes;
407+
#endif
408+
314409
}
315410

316411

@@ -341,7 +436,10 @@ pymain_run_command(wchar_t *command)
341436
if (new_bytes==NULL) {
342437
gotoerror;
343438
}
344-
Py_DECREF(bytes);
439+
if (new_bytes!=bytes) {
440+
// dedent allocated new bytes, replace the old with the new
441+
Py_DECREF(bytes);
442+
}
345443
bytes=new_bytes;
346444
}
347445

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp