Show More
@@ -1,480 +1,484 b'' | |||||
1 | """ |
|
1 | """ | |
2 | Module provides a class allowing to wrap communication over subprocess.Popen |
|
2 | Module provides a class allowing to wrap communication over subprocess.Popen | |
3 | input, output, error streams into a meaningfull, non-blocking, concurrent |
|
3 | input, output, error streams into a meaningfull, non-blocking, concurrent | |
4 | stream processor exposing the output data as an iterator fitting to be a |
|
4 | stream processor exposing the output data as an iterator fitting to be a | |
5 | return value passed by a WSGI applicaiton to a WSGI server per PEP 3333. |
|
5 | return value passed by a WSGI applicaiton to a WSGI server per PEP 3333. | |
6 |
|
6 | |||
7 | Copyright (c) 2011 Daniel Dotsenko <dotsa[at]hotmail.com> |
|
7 | Copyright (c) 2011 Daniel Dotsenko <dotsa[at]hotmail.com> | |
8 |
|
8 | |||
9 | This file is part of git_http_backend.py Project. |
|
9 | This file is part of git_http_backend.py Project. | |
10 |
|
10 | |||
11 | git_http_backend.py Project is free software: you can redistribute it and/or |
|
11 | git_http_backend.py Project is free software: you can redistribute it and/or | |
12 | modify it under the terms of the GNU Lesser General Public License as |
|
12 | modify it under the terms of the GNU Lesser General Public License as | |
13 | published by the Free Software Foundation, either version 2.1 of the License, |
|
13 | published by the Free Software Foundation, either version 2.1 of the License, | |
14 | or (at your option) any later version. |
|
14 | or (at your option) any later version. | |
15 |
|
15 | |||
16 | git_http_backend.py Project is distributed in the hope that it will be useful, |
|
16 | git_http_backend.py Project is distributed in the hope that it will be useful, | |
17 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
17 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
18 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
18 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
19 | GNU Lesser General Public License for more details. |
|
19 | GNU Lesser General Public License for more details. | |
20 |
|
20 | |||
21 | You should have received a copy of the GNU Lesser General Public License |
|
21 | You should have received a copy of the GNU Lesser General Public License | |
22 | along with git_http_backend.py Project. |
|
22 | along with git_http_backend.py Project. | |
23 | If not, see <http://www.gnu.org/licenses/>. |
|
23 | If not, see <http://www.gnu.org/licenses/>. | |
24 | """ |
|
24 | """ | |
25 | import os |
|
25 | import os | |
|
26 | import logging | |||
26 | import subprocess32 as subprocess |
|
27 | import subprocess32 as subprocess | |
27 | from collections import deque |
|
28 | from collections import deque | |
28 | from threading import Event, Thread |
|
29 | from threading import Event, Thread | |
29 |
|
30 | |||
|
31 | log = logging.getLogger(__name__) | |||
|
32 | ||||
30 |
|
33 | |||
31 | class StreamFeeder(Thread): |
|
34 | class StreamFeeder(Thread): | |
32 | """ |
|
35 | """ | |
33 | Normal writing into pipe-like is blocking once the buffer is filled. |
|
36 | Normal writing into pipe-like is blocking once the buffer is filled. | |
34 | This thread allows a thread to seep data from a file-like into a pipe |
|
37 | This thread allows a thread to seep data from a file-like into a pipe | |
35 | without blocking the main thread. |
|
38 | without blocking the main thread. | |
36 | We close inpipe once the end of the source stream is reached. |
|
39 | We close inpipe once the end of the source stream is reached. | |
37 | """ |
|
40 | """ | |
38 |
|
41 | |||
39 | def __init__(self, source): |
|
42 | def __init__(self, source): | |
40 | super(StreamFeeder, self).__init__() |
|
43 | super(StreamFeeder, self).__init__() | |
41 | self.daemon = True |
|
44 | self.daemon = True | |
42 | filelike = False |
|
45 | filelike = False | |
43 | self.bytes = bytes() |
|
46 | self.bytes = bytes() | |
44 | if type(source) in (type(''), bytes, bytearray): # string-like |
|
47 | if type(source) in (type(''), bytes, bytearray): # string-like | |
45 | self.bytes = bytes(source) |
|
48 | self.bytes = bytes(source) | |
46 | else: # can be either file pointer or file-like |
|
49 | else: # can be either file pointer or file-like | |
47 | if type(source) in (int, long): # file pointer it is |
|
50 | if type(source) in (int, long): # file pointer it is | |
48 | # converting file descriptor (int) stdin into file-like |
|
51 | # converting file descriptor (int) stdin into file-like | |
49 | try: |
|
52 | try: | |
50 | source = os.fdopen(source, 'rb', 16384) |
|
53 | source = os.fdopen(source, 'rb', 16384) | |
51 | except Exception: |
|
54 | except Exception: | |
52 | pass |
|
55 | pass | |
53 | # let's see if source is file-like by now |
|
56 | # let's see if source is file-like by now | |
54 | try: |
|
57 | try: | |
55 | filelike = source.read |
|
58 | filelike = source.read | |
56 | except Exception: |
|
59 | except Exception: | |
57 | pass |
|
60 | pass | |
58 | if not filelike and not self.bytes: |
|
61 | if not filelike and not self.bytes: | |
59 | raise TypeError("StreamFeeder's source object must be a readable " |
|
62 | raise TypeError("StreamFeeder's source object must be a readable " | |
60 | "file-like, a file descriptor, or a string-like.") |
|
63 | "file-like, a file descriptor, or a string-like.") | |
61 | self.source = source |
|
64 | self.source = source | |
62 | self.readiface, self.writeiface = os.pipe() |
|
65 | self.readiface, self.writeiface = os.pipe() | |
63 |
|
66 | |||
64 | def run(self): |
|
67 | def run(self): | |
65 | t = self.writeiface |
|
68 | t = self.writeiface | |
66 | if self.bytes: |
|
69 | if self.bytes: | |
67 | os.write(t, self.bytes) |
|
70 | os.write(t, self.bytes) | |
68 | else: |
|
71 | else: | |
69 | s = self.source |
|
72 | s = self.source | |
70 | b = s.read(4096) |
|
73 | b = s.read(4096) | |
71 | while b: |
|
74 | while b: | |
72 | os.write(t, b) |
|
75 | os.write(t, b) | |
73 | b = s.read(4096) |
|
76 | b = s.read(4096) | |
74 | os.close(t) |
|
77 | os.close(t) | |
75 |
|
78 | |||
76 | @property |
|
79 | @property | |
77 | def output(self): |
|
80 | def output(self): | |
78 | return self.readiface |
|
81 | return self.readiface | |
79 |
|
82 | |||
80 |
|
83 | |||
81 | class InputStreamChunker(Thread): |
|
84 | class InputStreamChunker(Thread): | |
82 | def __init__(self, source, target, buffer_size, chunk_size): |
|
85 | def __init__(self, source, target, buffer_size, chunk_size): | |
83 |
|
86 | |||
84 | super(InputStreamChunker, self).__init__() |
|
87 | super(InputStreamChunker, self).__init__() | |
85 |
|
88 | |||
86 | self.daemon = True # die die die. |
|
89 | self.daemon = True # die die die. | |
87 |
|
90 | |||
88 | self.source = source |
|
91 | self.source = source | |
89 | self.target = target |
|
92 | self.target = target | |
90 | self.chunk_count_max = int(buffer_size / chunk_size) + 1 |
|
93 | self.chunk_count_max = int(buffer_size / chunk_size) + 1 | |
91 | self.chunk_size = chunk_size |
|
94 | self.chunk_size = chunk_size | |
92 |
|
95 | |||
93 | self.data_added = Event() |
|
96 | self.data_added = Event() | |
94 | self.data_added.clear() |
|
97 | self.data_added.clear() | |
95 |
|
98 | |||
96 | self.keep_reading = Event() |
|
99 | self.keep_reading = Event() | |
97 | self.keep_reading.set() |
|
100 | self.keep_reading.set() | |
98 |
|
101 | |||
99 | self.EOF = Event() |
|
102 | self.EOF = Event() | |
100 | self.EOF.clear() |
|
103 | self.EOF.clear() | |
101 |
|
104 | |||
102 | self.go = Event() |
|
105 | self.go = Event() | |
103 | self.go.set() |
|
106 | self.go.set() | |
104 |
|
107 | |||
105 | def stop(self): |
|
108 | def stop(self): | |
106 | self.go.clear() |
|
109 | self.go.clear() | |
107 | self.EOF.set() |
|
110 | self.EOF.set() | |
108 | try: |
|
111 | try: | |
109 | # this is not proper, but is done to force the reader thread let |
|
112 | # this is not proper, but is done to force the reader thread let | |
110 | # go of the input because, if successful, .close() will send EOF |
|
113 | # go of the input because, if successful, .close() will send EOF | |
111 | # down the pipe. |
|
114 | # down the pipe. | |
112 | self.source.close() |
|
115 | self.source.close() | |
113 | except: |
|
116 | except: | |
114 | pass |
|
117 | pass | |
115 |
|
118 | |||
116 | def run(self): |
|
119 | def run(self): | |
117 | s = self.source |
|
120 | s = self.source | |
118 | t = self.target |
|
121 | t = self.target | |
119 | cs = self.chunk_size |
|
122 | cs = self.chunk_size | |
120 | chunk_count_max = self.chunk_count_max |
|
123 | chunk_count_max = self.chunk_count_max | |
121 | keep_reading = self.keep_reading |
|
124 | keep_reading = self.keep_reading | |
122 | da = self.data_added |
|
125 | da = self.data_added | |
123 | go = self.go |
|
126 | go = self.go | |
124 |
|
127 | |||
125 | try: |
|
128 | try: | |
126 | b = s.read(cs) |
|
129 | b = s.read(cs) | |
127 | except ValueError: |
|
130 | except ValueError: | |
128 | b = '' |
|
131 | b = '' | |
129 |
|
132 | |||
130 | timeout_input = 20 |
|
133 | timeout_input = 20 | |
131 | while b and go.is_set(): |
|
134 | while b and go.is_set(): | |
132 | if len(t) > chunk_count_max: |
|
135 | if len(t) > chunk_count_max: | |
133 | keep_reading.clear() |
|
136 | keep_reading.clear() | |
134 | keep_reading.wait(timeout_input) |
|
137 | keep_reading.wait(timeout_input) | |
135 | if len(t) > chunk_count_max + timeout_input: |
|
138 | if len(t) > chunk_count_max + timeout_input: | |
136 | raise IOError( |
|
139 | log.error("Timed out while waiting for input from subprocess.") | |
137 | "Timed out while waiting for input from subprocess.") |
|
140 | os._exit(-1) # this will cause the worker to recycle itself | |
|
141 | ||||
138 | t.append(b) |
|
142 | t.append(b) | |
139 | da.set() |
|
143 | da.set() | |
140 | b = s.read(cs) |
|
144 | b = s.read(cs) | |
141 | self.EOF.set() |
|
145 | self.EOF.set() | |
142 | da.set() # for cases when done but there was no input. |
|
146 | da.set() # for cases when done but there was no input. | |
143 |
|
147 | |||
144 |
|
148 | |||
145 | class BufferedGenerator(object): |
|
149 | class BufferedGenerator(object): | |
146 | """ |
|
150 | """ | |
147 | Class behaves as a non-blocking, buffered pipe reader. |
|
151 | Class behaves as a non-blocking, buffered pipe reader. | |
148 | Reads chunks of data (through a thread) |
|
152 | Reads chunks of data (through a thread) | |
149 | from a blocking pipe, and attaches these to an array (Deque) of chunks. |
|
153 | from a blocking pipe, and attaches these to an array (Deque) of chunks. | |
150 | Reading is halted in the thread when max chunks is internally buffered. |
|
154 | Reading is halted in the thread when max chunks is internally buffered. | |
151 | The .next() may operate in blocking or non-blocking fashion by yielding |
|
155 | The .next() may operate in blocking or non-blocking fashion by yielding | |
152 | '' if no data is ready |
|
156 | '' if no data is ready | |
153 | to be sent or by not returning until there is some data to send |
|
157 | to be sent or by not returning until there is some data to send | |
154 | When we get EOF from underlying source pipe we raise the marker to raise |
|
158 | When we get EOF from underlying source pipe we raise the marker to raise | |
155 | StopIteration after the last chunk of data is yielded. |
|
159 | StopIteration after the last chunk of data is yielded. | |
156 | """ |
|
160 | """ | |
157 |
|
161 | |||
158 | def __init__(self, source, buffer_size=65536, chunk_size=4096, |
|
162 | def __init__(self, source, buffer_size=65536, chunk_size=4096, | |
159 | starting_values=None, bottomless=False): |
|
163 | starting_values=None, bottomless=False): | |
160 | starting_values = starting_values or [] |
|
164 | starting_values = starting_values or [] | |
161 |
|
165 | |||
162 | if bottomless: |
|
166 | if bottomless: | |
163 | maxlen = int(buffer_size / chunk_size) |
|
167 | maxlen = int(buffer_size / chunk_size) | |
164 | else: |
|
168 | else: | |
165 | maxlen = None |
|
169 | maxlen = None | |
166 |
|
170 | |||
167 | self.data = deque(starting_values, maxlen) |
|
171 | self.data = deque(starting_values, maxlen) | |
168 | self.worker = InputStreamChunker(source, self.data, buffer_size, |
|
172 | self.worker = InputStreamChunker(source, self.data, buffer_size, | |
169 | chunk_size) |
|
173 | chunk_size) | |
170 | if starting_values: |
|
174 | if starting_values: | |
171 | self.worker.data_added.set() |
|
175 | self.worker.data_added.set() | |
172 | self.worker.start() |
|
176 | self.worker.start() | |
173 |
|
177 | |||
174 | #################### |
|
178 | #################### | |
175 | # Generator's methods |
|
179 | # Generator's methods | |
176 | #################### |
|
180 | #################### | |
177 |
|
181 | |||
178 | def __iter__(self): |
|
182 | def __iter__(self): | |
179 | return self |
|
183 | return self | |
180 |
|
184 | |||
181 | def next(self): |
|
185 | def next(self): | |
182 | while not len(self.data) and not self.worker.EOF.is_set(): |
|
186 | while not len(self.data) and not self.worker.EOF.is_set(): | |
183 | self.worker.data_added.clear() |
|
187 | self.worker.data_added.clear() | |
184 | self.worker.data_added.wait(0.2) |
|
188 | self.worker.data_added.wait(0.2) | |
185 | if len(self.data): |
|
189 | if len(self.data): | |
186 | self.worker.keep_reading.set() |
|
190 | self.worker.keep_reading.set() | |
187 | return bytes(self.data.popleft()) |
|
191 | return bytes(self.data.popleft()) | |
188 | elif self.worker.EOF.is_set(): |
|
192 | elif self.worker.EOF.is_set(): | |
189 | raise StopIteration |
|
193 | raise StopIteration | |
190 |
|
194 | |||
191 | def throw(self, exc_type, value=None, traceback=None): |
|
195 | def throw(self, exc_type, value=None, traceback=None): | |
192 | if not self.worker.EOF.is_set(): |
|
196 | if not self.worker.EOF.is_set(): | |
193 | raise exc_type(value) |
|
197 | raise exc_type(value) | |
194 |
|
198 | |||
195 | def start(self): |
|
199 | def start(self): | |
196 | self.worker.start() |
|
200 | self.worker.start() | |
197 |
|
201 | |||
198 | def stop(self): |
|
202 | def stop(self): | |
199 | self.worker.stop() |
|
203 | self.worker.stop() | |
200 |
|
204 | |||
201 | def close(self): |
|
205 | def close(self): | |
202 | try: |
|
206 | try: | |
203 | self.worker.stop() |
|
207 | self.worker.stop() | |
204 | self.throw(GeneratorExit) |
|
208 | self.throw(GeneratorExit) | |
205 | except (GeneratorExit, StopIteration): |
|
209 | except (GeneratorExit, StopIteration): | |
206 | pass |
|
210 | pass | |
207 |
|
211 | |||
208 | def __del__(self): |
|
212 | def __del__(self): | |
209 | self.close() |
|
213 | self.close() | |
210 |
|
214 | |||
211 | #################### |
|
215 | #################### | |
212 | # Threaded reader's infrastructure. |
|
216 | # Threaded reader's infrastructure. | |
213 | #################### |
|
217 | #################### | |
214 | @property |
|
218 | @property | |
215 | def input(self): |
|
219 | def input(self): | |
216 | return self.worker.w |
|
220 | return self.worker.w | |
217 |
|
221 | |||
218 | @property |
|
222 | @property | |
219 | def data_added_event(self): |
|
223 | def data_added_event(self): | |
220 | return self.worker.data_added |
|
224 | return self.worker.data_added | |
221 |
|
225 | |||
222 | @property |
|
226 | @property | |
223 | def data_added(self): |
|
227 | def data_added(self): | |
224 | return self.worker.data_added.is_set() |
|
228 | return self.worker.data_added.is_set() | |
225 |
|
229 | |||
226 | @property |
|
230 | @property | |
227 | def reading_paused(self): |
|
231 | def reading_paused(self): | |
228 | return not self.worker.keep_reading.is_set() |
|
232 | return not self.worker.keep_reading.is_set() | |
229 |
|
233 | |||
230 | @property |
|
234 | @property | |
231 | def done_reading_event(self): |
|
235 | def done_reading_event(self): | |
232 | """ |
|
236 | """ | |
233 | Done_reding does not mean that the iterator's buffer is empty. |
|
237 | Done_reding does not mean that the iterator's buffer is empty. | |
234 | Iterator might have done reading from underlying source, but the read |
|
238 | Iterator might have done reading from underlying source, but the read | |
235 | chunks might still be available for serving through .next() method. |
|
239 | chunks might still be available for serving through .next() method. | |
236 |
|
240 | |||
237 | :returns: An Event class instance. |
|
241 | :returns: An Event class instance. | |
238 | """ |
|
242 | """ | |
239 | return self.worker.EOF |
|
243 | return self.worker.EOF | |
240 |
|
244 | |||
241 | @property |
|
245 | @property | |
242 | def done_reading(self): |
|
246 | def done_reading(self): | |
243 | """ |
|
247 | """ | |
244 | Done_reding does not mean that the iterator's buffer is empty. |
|
248 | Done_reding does not mean that the iterator's buffer is empty. | |
245 | Iterator might have done reading from underlying source, but the read |
|
249 | Iterator might have done reading from underlying source, but the read | |
246 | chunks might still be available for serving through .next() method. |
|
250 | chunks might still be available for serving through .next() method. | |
247 |
|
251 | |||
248 | :returns: An Bool value. |
|
252 | :returns: An Bool value. | |
249 | """ |
|
253 | """ | |
250 | return self.worker.EOF.is_set() |
|
254 | return self.worker.EOF.is_set() | |
251 |
|
255 | |||
252 | @property |
|
256 | @property | |
253 | def length(self): |
|
257 | def length(self): | |
254 | """ |
|
258 | """ | |
255 | returns int. |
|
259 | returns int. | |
256 |
|
260 | |||
257 | This is the lenght of the que of chunks, not the length of |
|
261 | This is the lenght of the que of chunks, not the length of | |
258 | the combined contents in those chunks. |
|
262 | the combined contents in those chunks. | |
259 |
|
263 | |||
260 | __len__() cannot be meaningfully implemented because this |
|
264 | __len__() cannot be meaningfully implemented because this | |
261 | reader is just flying throuh a bottomless pit content and |
|
265 | reader is just flying throuh a bottomless pit content and | |
262 | can only know the lenght of what it already saw. |
|
266 | can only know the lenght of what it already saw. | |
263 |
|
267 | |||
264 | If __len__() on WSGI server per PEP 3333 returns a value, |
|
268 | If __len__() on WSGI server per PEP 3333 returns a value, | |
265 | the responce's length will be set to that. In order not to |
|
269 | the responce's length will be set to that. In order not to | |
266 | confuse WSGI PEP3333 servers, we will not implement __len__ |
|
270 | confuse WSGI PEP3333 servers, we will not implement __len__ | |
267 | at all. |
|
271 | at all. | |
268 | """ |
|
272 | """ | |
269 | return len(self.data) |
|
273 | return len(self.data) | |
270 |
|
274 | |||
271 | def prepend(self, x): |
|
275 | def prepend(self, x): | |
272 | self.data.appendleft(x) |
|
276 | self.data.appendleft(x) | |
273 |
|
277 | |||
274 | def append(self, x): |
|
278 | def append(self, x): | |
275 | self.data.append(x) |
|
279 | self.data.append(x) | |
276 |
|
280 | |||
277 | def extend(self, o): |
|
281 | def extend(self, o): | |
278 | self.data.extend(o) |
|
282 | self.data.extend(o) | |
279 |
|
283 | |||
280 | def __getitem__(self, i): |
|
284 | def __getitem__(self, i): | |
281 | return self.data[i] |
|
285 | return self.data[i] | |
282 |
|
286 | |||
283 |
|
287 | |||
284 | class SubprocessIOChunker(object): |
|
288 | class SubprocessIOChunker(object): | |
285 | """ |
|
289 | """ | |
286 | Processor class wrapping handling of subprocess IO. |
|
290 | Processor class wrapping handling of subprocess IO. | |
287 |
|
291 | |||
288 | .. important:: |
|
292 | .. important:: | |
289 |
|
293 | |||
290 | Watch out for the method `__del__` on this class. If this object |
|
294 | Watch out for the method `__del__` on this class. If this object | |
291 | is deleted, it will kill the subprocess, so avoid to |
|
295 | is deleted, it will kill the subprocess, so avoid to | |
292 | return the `output` attribute or usage of it like in the following |
|
296 | return the `output` attribute or usage of it like in the following | |
293 | example:: |
|
297 | example:: | |
294 |
|
298 | |||
295 | # `args` expected to run a program that produces a lot of output |
|
299 | # `args` expected to run a program that produces a lot of output | |
296 | output = ''.join(SubprocessIOChunker( |
|
300 | output = ''.join(SubprocessIOChunker( | |
297 | args, shell=False, inputstream=inputstream, env=environ).output) |
|
301 | args, shell=False, inputstream=inputstream, env=environ).output) | |
298 |
|
302 | |||
299 | # `output` will not contain all the data, because the __del__ method |
|
303 | # `output` will not contain all the data, because the __del__ method | |
300 | # has already killed the subprocess in this case before all output |
|
304 | # has already killed the subprocess in this case before all output | |
301 | # has been consumed. |
|
305 | # has been consumed. | |
302 |
|
306 | |||
303 |
|
307 | |||
304 |
|
308 | |||
305 | In a way, this is a "communicate()" replacement with a twist. |
|
309 | In a way, this is a "communicate()" replacement with a twist. | |
306 |
|
310 | |||
307 | - We are multithreaded. Writing in and reading out, err are all sep threads. |
|
311 | - We are multithreaded. Writing in and reading out, err are all sep threads. | |
308 | - We support concurrent (in and out) stream processing. |
|
312 | - We support concurrent (in and out) stream processing. | |
309 | - The output is not a stream. It's a queue of read string (bytes, not unicode) |
|
313 | - The output is not a stream. It's a queue of read string (bytes, not unicode) | |
310 | chunks. The object behaves as an iterable. You can "for chunk in obj:" us. |
|
314 | chunks. The object behaves as an iterable. You can "for chunk in obj:" us. | |
311 | - We are non-blocking in more respects than communicate() |
|
315 | - We are non-blocking in more respects than communicate() | |
312 | (reading from subprocess out pauses when internal buffer is full, but |
|
316 | (reading from subprocess out pauses when internal buffer is full, but | |
313 | does not block the parent calling code. On the flip side, reading from |
|
317 | does not block the parent calling code. On the flip side, reading from | |
314 | slow-yielding subprocess may block the iteration until data shows up. This |
|
318 | slow-yielding subprocess may block the iteration until data shows up. This | |
315 | does not block the parallel inpipe reading occurring parallel thread.) |
|
319 | does not block the parallel inpipe reading occurring parallel thread.) | |
316 |
|
320 | |||
317 | The purpose of the object is to allow us to wrap subprocess interactions into |
|
321 | The purpose of the object is to allow us to wrap subprocess interactions into | |
318 | and interable that can be passed to a WSGI server as the application's return |
|
322 | and interable that can be passed to a WSGI server as the application's return | |
319 | value. Because of stream-processing-ability, WSGI does not have to read ALL |
|
323 | value. Because of stream-processing-ability, WSGI does not have to read ALL | |
320 | of the subprocess's output and buffer it, before handing it to WSGI server for |
|
324 | of the subprocess's output and buffer it, before handing it to WSGI server for | |
321 | HTTP response. Instead, the class initializer reads just a bit of the stream |
|
325 | HTTP response. Instead, the class initializer reads just a bit of the stream | |
322 | to figure out if error ocurred or likely to occur and if not, just hands the |
|
326 | to figure out if error ocurred or likely to occur and if not, just hands the | |
323 | further iteration over subprocess output to the server for completion of HTTP |
|
327 | further iteration over subprocess output to the server for completion of HTTP | |
324 | response. |
|
328 | response. | |
325 |
|
329 | |||
326 | The real or perceived subprocess error is trapped and raised as one of |
|
330 | The real or perceived subprocess error is trapped and raised as one of | |
327 | EnvironmentError family of exceptions |
|
331 | EnvironmentError family of exceptions | |
328 |
|
332 | |||
329 | Example usage: |
|
333 | Example usage: | |
330 | # try: |
|
334 | # try: | |
331 | # answer = SubprocessIOChunker( |
|
335 | # answer = SubprocessIOChunker( | |
332 | # cmd, |
|
336 | # cmd, | |
333 | # input, |
|
337 | # input, | |
334 | # buffer_size = 65536, |
|
338 | # buffer_size = 65536, | |
335 | # chunk_size = 4096 |
|
339 | # chunk_size = 4096 | |
336 | # ) |
|
340 | # ) | |
337 | # except (EnvironmentError) as e: |
|
341 | # except (EnvironmentError) as e: | |
338 | # print str(e) |
|
342 | # print str(e) | |
339 | # raise e |
|
343 | # raise e | |
340 | # |
|
344 | # | |
341 | # return answer |
|
345 | # return answer | |
342 |
|
346 | |||
343 |
|
347 | |||
344 | """ |
|
348 | """ | |
345 |
|
349 | |||
346 | # TODO: johbo: This is used to make sure that the open end of the PIPE |
|
350 | # TODO: johbo: This is used to make sure that the open end of the PIPE | |
347 | # is closed in the end. It would be way better to wrap this into an |
|
351 | # is closed in the end. It would be way better to wrap this into an | |
348 | # object, so that it is closed automatically once it is consumed or |
|
352 | # object, so that it is closed automatically once it is consumed or | |
349 | # something similar. |
|
353 | # something similar. | |
350 | _close_input_fd = None |
|
354 | _close_input_fd = None | |
351 |
|
355 | |||
352 | _closed = False |
|
356 | _closed = False | |
353 |
|
357 | |||
354 | def __init__(self, cmd, inputstream=None, buffer_size=65536, |
|
358 | def __init__(self, cmd, inputstream=None, buffer_size=65536, | |
355 | chunk_size=4096, starting_values=None, fail_on_stderr=True, |
|
359 | chunk_size=4096, starting_values=None, fail_on_stderr=True, | |
356 | fail_on_return_code=True, **kwargs): |
|
360 | fail_on_return_code=True, **kwargs): | |
357 | """ |
|
361 | """ | |
358 | Initializes SubprocessIOChunker |
|
362 | Initializes SubprocessIOChunker | |
359 |
|
363 | |||
360 | :param cmd: A Subprocess.Popen style "cmd". Can be string or array of strings |
|
364 | :param cmd: A Subprocess.Popen style "cmd". Can be string or array of strings | |
361 | :param inputstream: (Default: None) A file-like, string, or file pointer. |
|
365 | :param inputstream: (Default: None) A file-like, string, or file pointer. | |
362 | :param buffer_size: (Default: 65536) A size of total buffer per stream in bytes. |
|
366 | :param buffer_size: (Default: 65536) A size of total buffer per stream in bytes. | |
363 | :param chunk_size: (Default: 4096) A max size of a chunk. Actual chunk may be smaller. |
|
367 | :param chunk_size: (Default: 4096) A max size of a chunk. Actual chunk may be smaller. | |
364 | :param starting_values: (Default: []) An array of strings to put in front of output que. |
|
368 | :param starting_values: (Default: []) An array of strings to put in front of output que. | |
365 | :param fail_on_stderr: (Default: True) Whether to raise an exception in |
|
369 | :param fail_on_stderr: (Default: True) Whether to raise an exception in | |
366 | case something is written to stderr. |
|
370 | case something is written to stderr. | |
367 | :param fail_on_return_code: (Default: True) Whether to raise an |
|
371 | :param fail_on_return_code: (Default: True) Whether to raise an | |
368 | exception if the return code is not 0. |
|
372 | exception if the return code is not 0. | |
369 | """ |
|
373 | """ | |
370 |
|
374 | |||
371 | starting_values = starting_values or [] |
|
375 | starting_values = starting_values or [] | |
372 | if inputstream: |
|
376 | if inputstream: | |
373 | input_streamer = StreamFeeder(inputstream) |
|
377 | input_streamer = StreamFeeder(inputstream) | |
374 | input_streamer.start() |
|
378 | input_streamer.start() | |
375 | inputstream = input_streamer.output |
|
379 | inputstream = input_streamer.output | |
376 | self._close_input_fd = inputstream |
|
380 | self._close_input_fd = inputstream | |
377 |
|
381 | |||
378 | self._fail_on_stderr = fail_on_stderr |
|
382 | self._fail_on_stderr = fail_on_stderr | |
379 | self._fail_on_return_code = fail_on_return_code |
|
383 | self._fail_on_return_code = fail_on_return_code | |
380 |
|
384 | |||
381 | _shell = kwargs.get('shell', True) |
|
385 | _shell = kwargs.get('shell', True) | |
382 | kwargs['shell'] = _shell |
|
386 | kwargs['shell'] = _shell | |
383 |
|
387 | |||
384 | _p = subprocess.Popen(cmd, bufsize=-1, |
|
388 | _p = subprocess.Popen(cmd, bufsize=-1, | |
385 | stdin=inputstream, |
|
389 | stdin=inputstream, | |
386 | stdout=subprocess.PIPE, |
|
390 | stdout=subprocess.PIPE, | |
387 | stderr=subprocess.PIPE, |
|
391 | stderr=subprocess.PIPE, | |
388 | **kwargs) |
|
392 | **kwargs) | |
389 |
|
393 | |||
390 | bg_out = BufferedGenerator(_p.stdout, buffer_size, chunk_size, |
|
394 | bg_out = BufferedGenerator(_p.stdout, buffer_size, chunk_size, | |
391 | starting_values) |
|
395 | starting_values) | |
392 | bg_err = BufferedGenerator(_p.stderr, 16000, 1, bottomless=True) |
|
396 | bg_err = BufferedGenerator(_p.stderr, 16000, 1, bottomless=True) | |
393 |
|
397 | |||
394 | while not bg_out.done_reading and not bg_out.reading_paused and not bg_err.length: |
|
398 | while not bg_out.done_reading and not bg_out.reading_paused and not bg_err.length: | |
395 | # doing this until we reach either end of file, or end of buffer. |
|
399 | # doing this until we reach either end of file, or end of buffer. | |
396 | bg_out.data_added_event.wait(1) |
|
400 | bg_out.data_added_event.wait(1) | |
397 | bg_out.data_added_event.clear() |
|
401 | bg_out.data_added_event.clear() | |
398 |
|
402 | |||
399 | # at this point it's still ambiguous if we are done reading or just full buffer. |
|
403 | # at this point it's still ambiguous if we are done reading or just full buffer. | |
400 | # Either way, if error (returned by ended process, or implied based on |
|
404 | # Either way, if error (returned by ended process, or implied based on | |
401 | # presence of stuff in stderr output) we error out. |
|
405 | # presence of stuff in stderr output) we error out. | |
402 | # Else, we are happy. |
|
406 | # Else, we are happy. | |
403 | _returncode = _p.poll() |
|
407 | _returncode = _p.poll() | |
404 |
|
408 | |||
405 | if ((_returncode and fail_on_return_code) or |
|
409 | if ((_returncode and fail_on_return_code) or | |
406 | (fail_on_stderr and _returncode is None and bg_err.length)): |
|
410 | (fail_on_stderr and _returncode is None and bg_err.length)): | |
407 | try: |
|
411 | try: | |
408 | _p.terminate() |
|
412 | _p.terminate() | |
409 | except Exception: |
|
413 | except Exception: | |
410 | pass |
|
414 | pass | |
411 | bg_out.stop() |
|
415 | bg_out.stop() | |
412 | bg_err.stop() |
|
416 | bg_err.stop() | |
413 | if fail_on_stderr: |
|
417 | if fail_on_stderr: | |
414 | err = ''.join(bg_err) |
|
418 | err = ''.join(bg_err) | |
415 | raise EnvironmentError( |
|
419 | raise EnvironmentError( | |
416 | "Subprocess exited due to an error:\n" + err) |
|
420 | "Subprocess exited due to an error:\n" + err) | |
417 | if _returncode and fail_on_return_code: |
|
421 | if _returncode and fail_on_return_code: | |
418 | err = ''.join(bg_err) |
|
422 | err = ''.join(bg_err) | |
419 | if not err: |
|
423 | if not err: | |
420 | # maybe get empty stderr, try stdout instead |
|
424 | # maybe get empty stderr, try stdout instead | |
421 | # in many cases git reports the errors on stdout too |
|
425 | # in many cases git reports the errors on stdout too | |
422 | err = ''.join(bg_out) |
|
426 | err = ''.join(bg_out) | |
423 | raise EnvironmentError( |
|
427 | raise EnvironmentError( | |
424 | "Subprocess exited with non 0 ret code:%s: stderr:%s" % ( |
|
428 | "Subprocess exited with non 0 ret code:%s: stderr:%s" % ( | |
425 | _returncode, err)) |
|
429 | _returncode, err)) | |
426 |
|
430 | |||
427 | self.process = _p |
|
431 | self.process = _p | |
428 | self.output = bg_out |
|
432 | self.output = bg_out | |
429 | self.error = bg_err |
|
433 | self.error = bg_err | |
430 |
|
434 | |||
431 | def __iter__(self): |
|
435 | def __iter__(self): | |
432 | return self |
|
436 | return self | |
433 |
|
437 | |||
434 | def next(self): |
|
438 | def next(self): | |
435 | # Note: mikhail: We need to be sure that we are checking the return |
|
439 | # Note: mikhail: We need to be sure that we are checking the return | |
436 | # code after the stdout stream is closed. Some processes, e.g. git |
|
440 | # code after the stdout stream is closed. Some processes, e.g. git | |
437 | # are doing some magic in between closing stdout and terminating the |
|
441 | # are doing some magic in between closing stdout and terminating the | |
438 | # process and, as a result, we are not getting return code on "slow" |
|
442 | # process and, as a result, we are not getting return code on "slow" | |
439 | # systems. |
|
443 | # systems. | |
440 | result = None |
|
444 | result = None | |
441 | stop_iteration = None |
|
445 | stop_iteration = None | |
442 | try: |
|
446 | try: | |
443 | result = self.output.next() |
|
447 | result = self.output.next() | |
444 | except StopIteration as e: |
|
448 | except StopIteration as e: | |
445 | stop_iteration = e |
|
449 | stop_iteration = e | |
446 |
|
450 | |||
447 | if self.process.poll() and self._fail_on_return_code: |
|
451 | if self.process.poll() and self._fail_on_return_code: | |
448 | err = '%s' % ''.join(self.error) |
|
452 | err = '%s' % ''.join(self.error) | |
449 | raise EnvironmentError( |
|
453 | raise EnvironmentError( | |
450 | "Subprocess exited due to an error:\n" + err) |
|
454 | "Subprocess exited due to an error:\n" + err) | |
451 |
|
455 | |||
452 | if stop_iteration: |
|
456 | if stop_iteration: | |
453 | raise stop_iteration |
|
457 | raise stop_iteration | |
454 | return result |
|
458 | return result | |
455 |
|
459 | |||
456 | def throw(self, type, value=None, traceback=None): |
|
460 | def throw(self, type, value=None, traceback=None): | |
457 | if self.output.length or not self.output.done_reading: |
|
461 | if self.output.length or not self.output.done_reading: | |
458 | raise type(value) |
|
462 | raise type(value) | |
459 |
|
463 | |||
460 | def close(self): |
|
464 | def close(self): | |
461 | if self._closed: |
|
465 | if self._closed: | |
462 | return |
|
466 | return | |
463 | self._closed = True |
|
467 | self._closed = True | |
464 | try: |
|
468 | try: | |
465 | self.process.terminate() |
|
469 | self.process.terminate() | |
466 | except: |
|
470 | except: | |
467 | pass |
|
471 | pass | |
468 | if self._close_input_fd: |
|
472 | if self._close_input_fd: | |
469 | os.close(self._close_input_fd) |
|
473 | os.close(self._close_input_fd) | |
470 | try: |
|
474 | try: | |
471 | self.output.close() |
|
475 | self.output.close() | |
472 | except: |
|
476 | except: | |
473 | pass |
|
477 | pass | |
474 | try: |
|
478 | try: | |
475 | self.error.close() |
|
479 | self.error.close() | |
476 | except: |
|
480 | except: | |
477 | pass |
|
481 | pass | |
478 |
|
482 | |||
479 | def __del__(self): |
|
483 | def __del__(self): | |
480 | self.close() |
|
484 | self.close() |
General Comments 0
You need to be logged in to leave comments.
Login now