1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
|
# hiredis-py
[](https://github.com/redis/hiredis-py/actions/workflows/integration.yaml)
[](https://opensource.org/licenses/BSD-3-Clause)
Python extension that wraps protocol parsing code in [hiredis][hiredis].
It primarily speeds up parsing of multi bulk replies.
[hiredis]: http://github.com/redis/hiredis
## How do I Redis?
[Learn for free at Redis University](https://university.redis.com/)
[Build faster with the Redis Launchpad](https://launchpad.redis.com/)
[Try the Redis Cloud](https://redis.com/try-free/)
[Dive in developer tutorials](https://developer.redis.com/)
[Join the Redis community](https://redis.com/community/)
[Work at Redis](https://redis.com/company/careers/jobs/)
## Install
hiredis-py is available on [PyPI](https://pypi.org/project/hiredis/), and can be installed via:
```bash
pip install hiredis
```
## Building and Testing
Building this repository requires a recursive checkout of submodules, and building hiredis. The following example shows how to clone, compile, and run tests. Please note - you will need the gcc installed.
```bash
git clone --recursse-submodules https://github.com/redis/hiredis-py
python setup.py build_ext --inplace
python -m pytest
```
### Requirements
hiredis-py requires **Python 3.7+**.
Make sure Python development headers are available when installing hiredis-py.
On Ubuntu/Debian systems, install them with `apt-get install python3-dev`.
## Usage
The `hiredis` module contains the `Reader` class. This class is responsible for
parsing replies from the stream of data that is read from a Redis connection.
It does not contain functionality to handle I/O.
### Reply parser
The `Reader` class has two methods that are used when parsing replies from a
stream of data. `Reader.feed` takes a string argument that is appended to the
internal buffer. `Reader.gets` reads this buffer and returns a reply when the
buffer contains a full reply. If a single call to `feed` contains multiple
replies, `gets` should be called multiple times to extract all replies.
Example:
```python
>>> reader = hiredis.Reader()
>>> reader.feed("$5\r\nhello\r\n")
>>> reader.gets()
b'hello'
```
When the buffer does not contain a full reply, `gets` returns `False`.
This means extra data is needed and `feed` should be called again before calling
`gets` again. Alternatively you could provide custom sentinel object via parameter,
which is useful for RESP3 protocol where native boolean types are supported:
Example:
```python
>>> reader.feed("*2\r\n$5\r\nhello\r\n")
>>> reader.gets()
False
>>> reader.feed("$5\r\nworld\r\n")
>>> reader.gets()
[b'hello', b'world']
>>> reader = hiredis.Reader(notEnoughData=Ellipsis)
>>> reader.gets()
Ellipsis
```
#### Unicode
`hiredis.Reader` is able to decode bulk data to any encoding Python supports.
To do so, specify the encoding you want to use for decoding replies when
initializing it:
```python
>>> reader = hiredis.Reader(encoding="utf-8", errors="strict")
>>> reader.feed(b"$3\r\n\xe2\x98\x83\r\n")
>>> reader.gets()
'☃'
```
Decoding of bulk data will be attempted using the specified encoding and
error handler. If the error handler is `'strict'` (the default), a
`UnicodeDecodeError` is raised when data cannot be dedcoded. This is identical
to Python's default behavior. Other valid values to `errors` include
`'replace'`, `'ignore'`, and `'backslashreplace'`. More information on the
behavior of these error handlers can be found
[here](https://docs.python.org/3/howto/unicode.html#the-string-type).
When the specified encoding cannot be found, a `LookupError` will be raised
when calling `gets` for the first reply with bulk data.
#### Error handling
When a protocol error occurs (because of multiple threads using the same
socket, or some other condition that causes a corrupt stream), the error
`hiredis.ProtocolError` is raised. Because the buffer is read in a lazy
fashion, it will only be raised when `gets` is called and the first reply in
the buffer contains an error. There is no way to recover from a faulty protocol
state, so when this happens, the I/O code feeding data to `Reader` should
probably reconnect.
Redis can reply with error replies (`-ERR ...`). For these replies, the custom
error class `hiredis.ReplyError` is returned, **but not raised**.
When other error types should be used (so existing code doesn't have to change
its `except` clauses), `Reader` can be initialized with the `protocolError` and
`replyError` keywords. These keywords should contain a *class* that is a
subclass of `Exception`. When not provided, `Reader` will use the default
error types.
## Benchmarks
The repository contains a benchmarking script in the `benchmark` directory,
which uses [gevent](http://gevent.org/) to have non-blocking I/O and redis-py
to handle connections. These benchmarks are done with a patched version of
redis-py that uses hiredis-py when it is available.
All benchmarks are done with 10 concurrent connections.
* SET key value + GET key
* redis-py: 11.76 Kops
* redis-py *with* hiredis-py: 13.40 Kops
* improvement: **1.1x**
List entries in the following tests are 5 bytes.
* LRANGE list 0 **9**:
* redis-py: 4.78 Kops
* redis-py *with* hiredis-py: 12.94 Kops
* improvement: **2.7x**
* LRANGE list 0 **99**:
* redis-py: 0.73 Kops
* redis-py *with* hiredis-py: 11.90 Kops
* improvement: **16.3x**
* LRANGE list 0 **999**:
* redis-py: 0.07 Kops
* redis-py *with* hiredis-py: 5.83 Kops
* improvement: **83.2x**
Throughput improvement for simple SET/GET is minimal, but the larger multi bulk replies
get, the larger the performance improvement is.
## License
This code is released under the BSD license, after the license of hiredis.
|
