baidu-tr-free

Baidu translate for free -- local cache plus throttling. Let's hope it lasts.

Broken or not

: failing indicates broken. In case of failing, try the following workaround: log in to https://passport.baidu.com first. Then use the Chrome browser to visit https://fanyi.baidu.com/v2transapi?from=en&to=zh&query=test, press F12 and then ctr-R (or any method) to open devtools' Network tab and reload. Locate https://fanyi.baidu.com/v2transapi?from=en&to=zh and obtain the BAIDUID and token strings in the headers. Plug in the BAIDUID and token strings to lines 66-67 in the file bdtr.py.

Fixed

• Text longer than 30 characters can be handled now.
• Auto-fetching BAIDUID and TOKEN

Installation

pip install -U baidu-tr-free or

• install (pip or whatever) necessary requirements, e.g. pip install js2py requests_cache jsonpath_rw or pip install -r requirements.txt
• Drop the file bdtr.py in any folder in your PYTHONPATH (check with import sys; print(sys.path)
  • Several valid TOKEN/BAIDUID are included.
• or clone the repo (e.g., git clone https://github.com/ffreemt/baidu-tr-free.git or download https://github.com/ffreemt/baidu-tr-free/archive/master.zip and unzip) and change to the baidu-tr-free folder and do a python setup.py develop

Usage

Log in to https://passport.baidu.com using Chrome: bdtr needs the cookies info (BAIDUID) from the Chrome browser on baidu.com.

from bdtr import bdtr
print(bdtr('hello world'))  # -> '你好，世界'
print(bdtr('hello world', to_lang='de'))  # ->'Hallo Welt'
print(bdtr('hello world', to_lang='jp'))  # ->'ハローワールド'

to_lang='fr' does not seem to work tho.

Validation and Pressure Tests

• pip install pytest
• pytest
  • Average delay for throttling set to 0.6 s, ajdust as needed

Acknowledgments

• Thanks to everyone whose code was used
• JS_SIGN (javascript code for signing) can be found on the net. It's also not too difficult to obtain some similar code from Chrome's devtools.